Mar 30, 2021
Donald Farmer is a data artist and Jedi of sorts. His BI wisdom is unmatched and he shares this knowledge as we cover his early interest in computers (and the power that they gave the user), the history of PowerPivot from one of the former "faces of Microsoft BI", and so much more! It's so very easy to respect this venerable figure of the data world. Here's Donald's website, Treehive Strategy
References in this episode:
Thomas Davenport's Competing on Analytics
Rob's Blog Post Featuring Alison Farmer's Artwork
Episode Timeline:
- 3:45 - Donald has had the data itch for a very long time and discovered power at an early age. It's a heck of an origin story!
- 11:15- The guys share some Donald Farmer stories, Tom's Donald inspired epiphany, and the Data Mining Add-In
- 20:00 - The history of Project Gemini AKA Power Pivot, the death of ProClarity, and the importance of tools of choice
- 39:45 - Data Artistry, some great Bill Gates stories, remote working
- 58:45 - The Community of purpose, and Power BI's features vs the competition, and the Treehive
Episode Transcript:
Rob Collie (00:00:00):
Hello, friends. Today, we welcome Donald Farmer. I had the pleasure
of working with Donald way back when, during the earliest days of
Power Pivot, and even greater pleasure, the greater honor of
calling him an old friend. In the world of data, you're just not
going to find someone warmer, more sincere, funnier, or dare I say,
wiser, than Donald Farmer. He's also pretty humble, so he probably
wouldn't like me saying those things about him. But tough, I'm
recording his intro on my own. As is customary with our guests,
Donald's got a very interesting backstory path from then to now,
but I would dare say his backstory would take the Pepsi challenge
with anyone's.
Rob Collie (00:00:39):
We had a really free-flowing conversation. Didn't really set a lot
of agenda for this one. And I think it played out really well. We
just got into some really interesting corners, some really funny
vignettes, and of course, pearls of wisdom. That's what Donald's
here for. We talk about his brief involuntary stint as Power Pivot
Yoda, his more than passing resemblance Tommy Chong, and also how
his willingness to blend his creative talents, artistic side with
this world of data really inspired me and gave me the confidence to
lean into my own voice, which has been incredibly important to me
over the last decade. He's a great person. It was a great
conversation. I hope you enjoy it. So let's get into it.
Announcer (00:01:20):
Ladies and gentlemen, may I have your attention please?
Announcer (00:01:24):
This is the Raw Data by P3 Adaptive Podcast, with your host, Rob
Collie, and your cohost, Thomas Larock. Find out what the experts
at P3 Adaptive can do for your business, just go to p3adaptive.com.
Raw Data by P3 Adaptive is data with the human element.
Rob Collie (00:01:47):
Welcome to the show, Donald Farmer. How are you, old friend?
Donald Farmer (00:01:51):
I am great, thanks. Thank you for having me.
Rob Collie (00:01:53):
We invented this podcast so that we could have you on, and we
played it cool for a while. We didn't want to just go rushing right
in and have you be right out of the gate, but this is our
reason.
Donald Farmer (00:02:04):
No pressure then.
Rob Collie (00:02:06):
No pressure at all. Yeah. Just be yourself.
Donald Farmer (00:02:08):
Well, I'm delighted to be here, really. It's great.
Rob Collie (00:02:11):
Well, we're going to have fun. So I thought we'd start here. We've
had some very interesting people on this show who have careers in
data. We've had some very, very, very interesting backgrounds, some
very interesting origin stories. We've had horse trainers, we've
had NFL quarterbacks, we've had marine biologists. I think you're
also in contention. One of your former lives puts you on this map.
I'm fishing for it. What am I fishing for?
Donald Farmer (00:02:39):
I'm not even sure what you're fishing for, because I've done so
much in my so-called career, but I've worked in fish farming, I've
worked in archeology. I met my wife when we were working for an
archeological excavation unit. What else have I done?
Rob Collie (00:02:51):
It was the archeology that I was fishing for, but I hadn't thought
about fish farming as I was fishing. Was it medieval
archeology?
Donald Farmer (00:02:59):
It was. I was particularly interested in medieval cities and the
growth of medieval cities.
Rob Collie (00:03:04):
Not to be too specific. That is so awesome. So, which was first,
the fish farming or the medieval archeology?
Donald Farmer (00:03:14):
Oh, the medieval archeology. The fish farming was a kind of site
thing. I can tell you a little bit about that. It is interesting,
but the study of history, archeology, and languages were really
important to me. That was really where I started, and drifted into
doing all sorts of things, creating applications for the study of
archeology, creating applications, databases for managing
archeological sites, things like that. And then we lived very
remotely in Scotland. The nearest house was about two miles from
us, the nearest town was 15 miles from us. And so we lived fairly
remotely, and there's a relatively limited amount of things going
on out there. There's some forestry, there's some farming, and
there was a fish farm.
Donald Farmer (00:03:54):
And so I ended up helping them with their computer systems. And
then from there, I ended up developing a business doing fairly
complex data analysis for fish farming.
Rob Collie (00:04:02):
That sounds very believable, the sound of authenticity. And
somewhere in there though, there is evolutionary step change, where
you first encountered data software. Do you have any idea where
that was? Where did you first get that itch?
Donald Farmer (00:04:18):
Oh, so that's actually been there all along. We had computers in
the house very, very early. So when I was eight or nine, which
would be in the '70s, we had a computer in our house because my
father was an electronics designer for British Telecom, which was
the national telecom company. He designed the first digital
exchanges for example.
Rob Collie (00:04:37):
Oh my gosh.
Donald Farmer (00:04:38):
So he was really into this, and we had computers in the house
pretty early and really clunky hex programming that we had to learn
in order to do the basic thing. The real breakthrough for me was
when I was probably about 11 or 12 and we got our first computer
that I could actually play with. And I remember the first
experience that I had there, which was a program, the Sieve of
Eratosthenes, calculates all the prime numbers. And I calculated
all the prime numbers up to a million, and it just seemed
incredible. Then a couple of extra lines of code up to 10 million,
up to 100 million. It was just so exciting.
Donald Farmer (00:05:16):
And it was tremendously exciting for me. And I had this feeling,
this experience of power that was really remarkable. To come back
to my father, he also had a big old car because as an engineer, he
fiddled with everything. We this huge, big old car. And one day, he
had to change a tire or something, so he let me use a hydraulic
jack to pump up the car, to lift the car. And there I was, this
little kid like nine or 10 with a hydraulic jack. And with one
hand, I was lifting this like two-ton old British limousine. An
amazing feeling. I just felt so powerful, I felt like Superman.
Donald Farmer (00:05:46):
I had exactly the same feeling. First time I was able to program a
computer. I just have this power, it was like an intellectual
power. And then I immediately started doing things like I started
writing applications to handle my book collection and my collection
of wild flowers and things like that, so all the things I wanted to
do. And of course, I was data programming, I didn't know it. And
then I had this tremendous breakthrough, which was, I was using
arrays, I didn't know about databases, so I was using arrays and
matrices in order to store data in a grid format.
Donald Farmer (00:06:19):
And then I realized that a cell of the array could contain the
address of our cell in another array. It could contain a variable.
Well, you guys know this, but at the age of 12 or 13, that's an
incredible insight to have. I don't mean that I was genius, I don't
mean that, it's a very powerful insight to have, it changes your
life. At that point, when I discovered that you could navigate, you
could manage, you could make all these incredible connections.
That's when data really took off for me. And so all through my
career, what I've always been looking for is that feeling again of,
"Wow, that's incredible."
Rob Collie (00:06:58):
That leverage. I'm laughing here. So Luke and I, Luke, the producer
here, we were computer partners in computer class in middle school.
So around that same age 12, we were experiencing the tremendous
feeling of power of Apple II low resolution graphics, simulating
someone sneezing and a droplet of slim flying across the screen.
That didn't have the same feeling of power. Here you are at the
same age, but an earlier point in the world calculating primes,
cataloging your book and wildflower collection. We were little boys
at a computer, there was nothing going on that was remotely
serious.
Rob Collie (00:07:41):
I didn't know about pointers or arrays of pointers, hell, I'm not
sure I've ever programmed an array of pointers, but I didn't even
know that their existence until college. So yeah, you're right, it
was there for you from the very beginning. Holy cow.
Donald Farmer (00:07:55):
Yeah. That was just tremendously exciting. Now, that actually has
an important pattern then in my so-called career, because I never
thought of computer science as something that I wanted to study. I
never have formally studied computer science. To me, it was always
a means to an end, it was always something else. So I just felt it
was just this tremendous power that I had and that I enjoyed, but I
wanted to use it for something. So I've never really been very
interested in computers in themselves or computer science in
itself.
Donald Farmer (00:08:27):
Right now, for example, I can tell you that I'm running a Mac. I
have no idea what processor's in it, I have no idea how big the
hard disk is or the memory or look at that. None of that matters to
me compared to, what can I do with it? Sometimes you have to know
things in order to do things better, but I'm actually really not
interested in the science of computation and computer science in
its own right.
Rob Collie (00:08:50):
I think that's the way to be. That was one of my fields of study in
college and it didn't change my life really. I still wasn't going
out of my way to write arrays of pointers or anything like that,
there was just nothing to do with it. I think you and I both agree
100% on the role of technology, you should not strive to be a
technologist, you should strive to be someone who solves problems
and knows when and how to deploy technology. Yeah, the tech is
never the star of any story.
Donald Farmer (00:09:21):
It is for some people. Some people get really into it. We probably
couldn't do our jobs unless there were those people who are really
into it doing their jobs.
Rob Collie (00:09:29):
That's true.
Donald Farmer (00:09:30):
So I think there's room for everyone. Think of it in terms of
building a team, the team has to include people who, in a sense,
don't care about the technology but care about the problem. And if
you have people who don't care about the problem and care about the
technology, well, you need to find a role for them, but you have to
be pretty careful in managing that role. Otherwise, you end up with
the engineering led software, the danger of boys with toys, the
technology becomes its own excitement.
Rob Collie (00:09:57):
That's the thing that I've been steering away from now for more
than a decade, because I did experience exactly that, especially in
my early days at Microsoft. There were often entire teams organized
around tech for tech's sake. And so it requires an enthusiasm for
technology in many, many instances to get something done. And in
fact, I have enthusiasm for certain technologies, the ones that
I've found that are really effective. It's more just like a
strategic mindset to adopt, is that technology is a means to an end
rather than an end in its own, which should be obvious, but in some
cases isn't.
Donald Farmer (00:10:32):
Well, everyone, to be effective, needs a sense of purpose. And
there are of course different purposes. My purpose is always, what
problems can I solve? But more interestingly for me, I think, how
can I do things differently? I guess I'm always a contrarian. I'm
not that much interested in doing things better, so much as I'm
interested in doing things differently in order to get a better
result. Incremental improvements aren't super interesting to me.
I'm a radical, I want to burn the house down and build it again.
I'm not a version two guy of anything.
Rob Collie (00:11:00):
I'm a fan of that mindset. I don't know if you know that about me,
that burn the house down is... Or at least the houses that deserved
burning, but not every house. That's a perfectly good house over
there.
Donald Farmer (00:11:12):
And make sure you're not in it when you do it.
Rob Collie (00:11:14):
Yeah, that's right. So do you know Tom. Have the two of you crossed
paths before?
Donald Farmer (00:11:20):
Absolutely. Yeah, yeah. And it's been a while of course, because
nobody's seeing anybody on the circuit nowadays, so it's been a
while since we caught up. But absolutely, I know Tom.
Thomas LaRock (00:11:29):
In the time waiting for you to join today, Donald and I got caught
up, and I shared my earliest memories. He was at a past summit
sitting on the floor, teaching people how to use, I think he said
it was the data analysis add-in for Excel. Was that it, or?
Donald Farmer (00:11:44):
Yeah, the data mining data mining add-in.
Thomas LaRock (00:11:46):
Data mining data mining add-in for Excel, which was revolutionary
at the time. Of course you know, Rob, how exciting this was. But he
had about 20 people gathered around while he sat on the floor doing
an impromptu demo to make sure people understood. But the other
memory I have is when he dressed in costume to get on stage at the
past summit as part of the keynote, and he was tweeting to us from
backstage. So this had to be 2008-
Donald Farmer (00:12:15):
Sounds about right.
Thomas LaRock (00:12:16):
And I'm like, "I can't believe I'm tweeting to the person behind
the stage at the keynote right now." And then I remember I went and
I chatted with you and Buck Woody for you in the speaker room.
Anyway, I'm starstruck. I was like, "Oh my God, It's Donald Farmer.
I get to talk with Donald Farmer." And he never once mentioned you,
Rob.
Rob Collie (00:12:34):
I know, I know. I used to do that to Donald at conferences. If I'd
pass him in the hallway, I would immediately put on my fanboy
voice, "Oh my God, it's Donald Farmer. It's Donald Farmer," running
towards him.
Thomas LaRock (00:12:46):
Here, fans of the podcast won't be able to see this, but there's
this book, and I showed Don earlier, it's called Competing on
Analytics, and it's from 2007. And I asked Don for some advice,
like, "Hey, what should I do to get started in this weird world of
data analytics?" Because I was a guy that was only focused on that
little database engine. And he said, "Yeah, go buy that book,
Competing on Analytics, start there." And so I did immediately, and
my eyes were just... Of course I'm not computer science guy, I'm a
math degree, I have a master's in mathematics, actually.
Thomas LaRock (00:13:20):
So when I saw that data and math was basically being combined, it
really hit home for me. I'm like, "Oh, I get it now." When I was
getting my degree, the only thing you thought you might do with
math was become an actuary, which is also data and math. But when I
started seeing it getting applied in what this book was talking
about and I saw it happening in sports industries as well and
taking off, that's when I got excited. And actually, that's right
about when Rob came into my life.
Donald Farmer (00:13:49):
Well, that book Competing on Analytics by Thomas Davenport, one of
the things I really liked about that at the time, and I still like
about it now is it's not a technical book, but it's not only a
business book either. It takes you one step beyond saying, "Here
are the business environment that you're working in, here are the
business issues that you might come across, and here are ways to
address that analytically. And here's why it's better to address it
analytically." I think Thomas Davenport is a great thinker in terms
of business, but I like the way he lays out the path to success,
which is so different.
Donald Farmer (00:14:21):
You've got a degree in mathematics, and yeah, at one time, your
only option, if you like, would be to an actuary. Actuaries were
data scientists, actuaries still are data scientists. They were
data scientists before the term. And they were always data
scientists that were operational engineers who specialized in
optimizations who were doing data science. There's still a lot we
can learn from them, especially from... I look at people developing
all sorts of fancy neural net algorithms, and I sometimes sit back,
I say, "You could solve that problem quicker, more easily and more
reliably with simulated annealing or some such analysis from the
1970s or '80s." And sometimes, again, it's this tendency to get
caught up in technology for its own sake.
Rob Collie (00:15:05):
Yeah, the hot new hammer.
Donald Farmer (00:15:06):
Exactly.
Rob Collie (00:15:07):
Let's go back to the data mining add-in.
Donald Farmer (00:15:08):
Yeah, gosh.
Thomas LaRock (00:15:10):
Before we get too far off track, I do want to know your origin
story, the two of you. So I'm hoping, does the data mining path get
us there?
Rob Collie (00:15:18):
I think that almost gets us there. It gets us passing in the night,
which is a weird thing.
Donald Farmer (00:15:23):
Yeah, it does. And actually, I think it does get as there more
directly than even you might know, Rob. So data mining add-in,
there was a great data mining team at Microsoft. And they were
really, I would say along with Oracle, they were the first people
to really embed predictive analytics inside the database as a
platform. And so there was a team led by ZhaoHui Tang and Jamie
MacLennan, they took algorithms that had come out of Microsoft
Research and brought them directly into the OLAP engine, in fact,
into the OLAP framework. And so people were able to do real
predictive analytics and data mining within the database.
Donald Farmer (00:15:56):
That was great, but it didn't really help bring that into the hands
of business users because you had to be deep into the language and
the architecture of the OLAP engine in order to use them. So they
had this great idea of, "Well, why don't we surface this through an
Excel add-in and make it simple to use and find a way of showing
the results or various ways of visualizing the results from
predictive analytics that would make it easy to understand, or at
least simple to understand?" And so that was the data mining
add-in. There was a couple of things about that. You spend all this
time building this beautiful interface, and it really was very
effective as one-click, two-click data mining add-in that to my
mind could actually have changed the world of business analysis if
we'd had the support to do it.
Donald Farmer (00:16:40):
And then it turns out that, what are we allowed to call this? We
have to call it the Microsoft SQL Server 2008 Data Mining Add-ins
For Microsoft Office System 2007, or something. You ended up with
this name, which was literally 200 characters long because that's
the branding guidelines. I do can call it something like quick side
or something like that, which would have been much more impact.
Very often, I was trying to give presentations where I literally
couldn't fit the name of the product into the 200 characters of the
presentation title. So it was frustrating, but it was a great
product. A lot of fun.
Donald Farmer (00:17:17):
And when I presented it, your story, Thomas, it's actually really
true. People would come out up to the conference, want to see so
much more, I'd have to sit down in the corridor and demonstrate it
again and show people how to use it. What's relevant for that to my
meeting with Rob ultimately is the Excel part because that really
opened the eyes of a lot of people to what you could do in Excel in
terms of interface, in terms of usability, in terms of simplicity
and the fact that rather than making a data mining product which
brought users to data mining, we did it the other way around, we
built a data mining product that took data mining to the user and
met the user where they lived and worked, which is in Excel.
Donald Farmer (00:17:59):
And that really started to transform our thinking within the SQL
Server team about, how do we enable business analytics? Rather than
building a special tool, rather than requiring users to discover
rise and come to us, why don't we go to them and enable business
analytics where they already live and work, and that's the Excel
environment? And at that point, Rob can pick up that side of the
story because this was where you got super-involved.
Rob Collie (00:18:26):
That's right. The part where I had mentioned that I think we passed
in the night was that I had a couple of meetings when I still
worked on Excel with, I believe Bogdan and Jamie, a series of
meetings, actually. I don't remember you being in them.
Donald Farmer (00:18:43):
I would not have been in those, no. At that time, I was probably
still deep inside integration services.
Rob Collie (00:18:48):
That's right. And so I got to participate in the design of that
add-in. Looking back, those meetings were as much about getting
office buy-in as they were about getting office input, which is
obviously very clever and basically just me. I was a lead program
manager at the time and my team was all very, very busy with their
individual projects. I had also been reaching this conclusion that
I was like, as a lead, at that time, I had no direct
responsibilities, I wasn't doing any of this stuff that I used to
enjoy, which was actually designing software. So when this thing
came along, Jamie and Bogdan showed up and said, "Hey, we'd like
some help designing this add-in." And I was like, "Oh, yeah, Okay.
I'll take that. That'll be my thing to do for a couple of months
and stay interested."
Rob Collie (00:19:33):
We called one of the buttons, Key Factors. And I remember that was
my idea. That's something that came to me, was to call it Key
Factors. And I was very proud of it at the time. And now later on,
we see it in Power BI called Key Influencers. I'm like, "Oh, damn
it, that's so much better. That's such a better name." We were
close to working together there, but shortly thereafter is when I
disappeared over into the services side of the house, the Bing and
MSN. And it was like about a year later year and a half later that
project Gemini, Power Pivot, the Excel plan came knocking. And
that's when we really started to get to know each other. I ended up
going over there and reporting to Donald on Project Gemini.
Donald Farmer (00:20:15):
Right. I got involved in Gemini at a really interesting time
because it had been incubated by Thierry D'hers, who led the client
side of the incubation, the great work there. And he'd had just a
great experience of building business applications within analysis
services, within the Office team earlier in his career at Hyperion.
So he'd done some really interesting work. He had a great
understanding of business analysis at a depth that was really
remarkable. He understood all the calculations that people would
have to do. And there were other parts of the incubation that were
also very important. We had Cristian Petculescu developing the
Vertipaq Engine, which has become a really important part of the BI
infrastructure, Dave Wickert, doing all that crazy work that had to
happen to integrate the management and administration side of
Gemini into SharePoint, which was a thankless task.
Rob Collie (00:21:04):
Indeed, indeed.
Donald Farmer (00:21:05):
What a nightmare that was. And then there was the client side,
which was just a lot of fun. And we had DAX, we should talk about
DAX at some point as well, of course, close to your heart. But I
took over the Gemini team at that point and I just really, really
enjoy that process. I guess Gemini gave me part of that sense that
I'd had before as a kid. Look at the power of God, I don't just do
a million rows in Excel, I can do tens, hundreds of millions of
rows in Excel. And that was tremendously exciting. If you remember
the demo, we used to have the Moovly at demo, which was the movie
database, which was 50 million, I think it was. And that was really
intense. That was quite exciting.
Rob Collie (00:21:44):
To back up, Gemini was the code name for what became Power Pivot
part. We'll make that 100% clear to the listeners. And yeah, I
still tell people to this day, the Vertipaq Engine and DAX and all
of that, is the only time in my entire career at Microsoft that I
felt like I was witnessing science fiction. It was just insane what
that engine could do. We almost didn't believe it. And I agree with
you, it's like that moment of cosmic leverage and that really
hasn't worn off.
Donald Farmer (00:22:16):
Sure. I know the feeling.
Rob Collie (00:22:18):
Every time I go and load several gigabytes of text dump into a
Power BI file and I look at the resulting file and it's like 50
megabytes and it only eats about 100 megabytes of Ram, I'm just
jiggling. I'm just like, "This is crazy."
Donald Farmer (00:22:34):
This actually caused me a real problem, a huge problem once. I was
in Israel, visiting Microsoft in Israel and doing some
presentations about Power Pivot, that was our Gemini as it was
still in that stage. And I was demonstrating to the Microsoft
Israel team, but also to a lot of our customers. And as you know, a
lot of the Gemini team were Israeli originally. They came from an
acquisition. So I'm in Israel, I'm doing this presentation and I
have to come home and I get to the airport and the flight was
leaving at about two or three o'clock in the morning. And I don't
know if you've ever been to Israel. Have you ever flown there?
Rob Collie (00:23:04):
Yeah. I have, one time.
Donald Farmer (00:23:05):
So you've been through the airport?
Rob Collie (00:23:06):
Yes.
Donald Farmer (00:23:06):
They take security somewhat seriously.
Rob Collie (00:23:09):
Yes, they do.
Donald Farmer (00:23:10):
Really seriously, including the fact that if you've got a laptop,
they will ask you to boot it up to make sure it's actually working,
that its interior is working. They'll even ask you to click in a
couple of files and open them to prove that it's working. Of
course, what file does the security officer ask me to open? The
Moovly Excel file sitting on my desktop, which is 50 million rows
of data. It takes about five minutes to load at that point because
we were still in beta. And of course, I then have to explain, but
it's just an Excel file. Yes, but it's got 50 million rows of data.
No, it can't possibly have 50 million rows. Meanwhile, my flight is
leaving and I'm waiting for this damn [inaudible 00:23:50] to boot
up.
Rob Collie (00:23:51):
Yeah. That's the kind of security that you encounter in that
airport.
Donald Farmer (00:23:55):
I think they knew they were talking about.
Rob Collie (00:24:00):
You are interacting at that moment with the best and brightest.
Donald Farmer (00:24:05):
Yes. Not some poor soul and minimum wage.
Rob Collie (00:24:08):
It feels like our version of the Ivy League about to be road
scholar type people are interviewing you at that airport?
Donald Farmer (00:24:17):
Right. Psychologically profiling you as well.
Rob Collie (00:24:21):
And they're like, "Bullshit, Excel only holds a million rows. And
you know what, that's only in the most recent versions, it was 64K
just a moment ago." They're going to know their shit.
Donald Farmer (00:24:32):
Yeah, exactly. And the security is really serious unlike a lot of
countries. I remember going and getting on a plane once in Sea–Tac
or going through security at Sea–Tac, the guy in front of me had a
gun in his bag, and it set off the alarms and everybody is, "Gun,
gun, gun." And there was security running around. And then he gave
it back to him and told him to check it. He just tried to put a gun
on the plane in his hand and you are telling him to check. This is
not A1 security people.
Rob Collie (00:24:59):
No, no. It is really jarring flying back from there and landing in
the United States. It's literally like your very next interaction
with humanity is JFK. And you're just like, "Oh, things are just so
dumb here in comparison."
Donald Farmer (00:25:15):
Well, yeah. And it's that tone of seriousness and the fact that
they take it seriously. And to be fair also, many things the
application of technology and the fact that they hire great people,
it's actually a good example of how something that's very routine
in our lives can actually become a specialty if you focus on it
intensely enough, plus we have drifted.
Rob Collie (00:25:36):
Oh, that's the whole point. That's the whole point is to drift. We
were talking about almost an organizational epiphany on the SQL
side that we can bring the tools to the users. A lot of things in
SQL up until that point didn't even have an interface with their
own at all. They were just essentially an API like analysis
services on its own. All it was was it's API.
Donald Farmer (00:26:01):
Right. If you want it to do anything with it. There was this very
interesting third-party browser, which had been built specially for
browsing analysis services. The fact that at first it was the only
engine you could connect to as your analysis services engine was
built by some guy out of Stanford and he called it Tableau or
something. I don't know whatever happened to that.
Rob Collie (00:26:19):
Oh yeah, yeah. So it was just a little throwaway tool.
Donald Farmer (00:26:23):
Oh my God.
Rob Collie (00:26:26):
I thought we were setting up for a ProClarity joke or something,
like going acquire the number one front end for your flagship data
platform and then kill it.
Donald Farmer (00:26:36):
Well, yeah. This is really interesting. We did kill ProClarity and
we killed a number of products that we're actually very promising
by making one huge mistake, which was, we assumed that
standardization of the interface was critical to simplicity and
ease of use. So we had a ton of interesting applications that ended
up getting thrown into this SharePoint environment. So we had data
quality services, which ended up in SharePoint, hosted by the
Office team and a SharePoint environment. We had master data
services, which ended up plugged into the same SharePoint
environment.
Donald Farmer (00:27:16):
SharePoint is very limited in its user experience. So all sorts of
radio buttons and check boxes and endless forms that you had to
fill out in those applications to do anything. And one of the
things that we totally missed was the importance, and this is
something I just talk about all the time though, so I'm going to
rant, is the importance of a tool of choice. That is, it's one
thing to deliver capabilities to someone, but it's another thing to
have them have a tool of choice. I want to use this, not just I
want to solve this business problem and here's a platform which
does it, but this is something I want to use.
Donald Farmer (00:27:53):
Tableau did an amazing job, Christian at Tableau did an amazing job
of building a tool of choice that people wanted to use. They
identified themselves as Tableau users and they became what they
would call them, Zen Masters or whatever. But to them, that was
then part of their definition of their career, I'm a Tableau
expert, I'm Tableau person. And that's a tool of choice. And
Microsoft for many years was in the wilderness of pushing out
standardization, "We already own your platform, so we're going to
integrate everything into that platform." But Gemini was this big
breakthrough, it was one thing that really happened that people
could then adopt that and say, "This is our tool of choice to use."
And that was a breakthrough for Microsoft, I think.
Rob Collie (00:28:35):
That tool of choice, something that you actually love. And I've
heard the story, I have no idea if it's true, but it seems like
it's believable that Steve Jobs put the switch on the back of the
Apple II so that in order to turn it on, kids with short arms had
to lean forward and essentially hug the Apple II turn it on and
off. It doesn't even really matter whether that's true or not, that
story still illustrates the point. And we did, we hugged our Apple
IIs before we programmed them to have phlegm flying across the
screen.
Donald Farmer (00:29:10):
Well, one of the things that Apple have always done, which is
really great is they've never oversimplified everything. They have
had a design philosophy, which is based around simplicity, but
they've never oversimplified. In fact, if anything, they require
you to put some effort in. There's a point at which... I love this
idea of the iPod. We had iPods, we had phones, I had an iPod, I
later had a Zune, which I enjoyed actually, but I had an iPod, I
had a phone, I could travel anywhere in the world and make phone
calls. I had music with me wherever I went. I had two chargers, I
had two cables, I had headphones. It was a mess in some ways, but I
had these devices.
Donald Farmer (00:29:47):
And then what was super interesting about what Apple did was they
used to even meet this joke when they launched the iPhone. They
made a joke in the first keynote that what they were actually going
to do is just create an iPod that could make phone calls and they
had a dial on the front of it. They had this old telephone dial on
the front of an iPod, which they actually patented that as well. I
think perhaps deliberately leading people astray. And they made a
joke of that, but what the actually launched was the iPhone, which
didn't just solve the problem of I've got two devices, it was
beautiful and it was expensive, and it didn't work very well as a
phone, did other great things.
Donald Farmer (00:30:21):
And had all sorts of things in the user experience, which were
completely unnecessary, the little map pointer, which animated and
wobbled as it landed, there was all sorts of stuff there. And the
touch keyboard, the people were very skeptical of that, how could
you possibly type into thing I'm using this touch keyboard on the
class? They not only did he make all that work, but that's actually
a lot more effort than they had to do to solve the problem of
having too many devices. The actually took you on a journey, into a
new place that is a different environment altogether and requires a
commitment, financial and a commitment of usage from you as a
user.
Donald Farmer (00:30:59):
So rather than dumbing it down and let's just solve the problem,
they actually created a new world that they would take you on this
journey into the new world. That seems to me super interesting.
When I look at what we did with Power Pivot ultimately, and then
with Power BI, we actually took people into a new world. We
expected to learn new things. It was actually quite demanding in
some ways although it was a simple product, it was still demanding,
you had to learn DAX, you had to learn some new concepts. But that
in itself gave you a commitment, you now had a sense of a practice
that you had adopted in order to use that product.
Donald Farmer (00:31:35):
That's much more satisfying ultimately to a user than just simply
solving a problem. If all we done was extend Excel to be able to
handle 500 million rows, that wouldn't have been interesting, not
nearly as much as taking people into a new world.
Rob Collie (00:31:51):
There's a parallel there to what fraternities do to their recruits.
You beat them as a group and just abuse them for some number of
times and then they value being a member of the fraternity even
more than what they would otherwise because that's what forges the
bond.
Donald Farmer (00:32:08):
That explains so much about your specifications, but I can go with
this.
Rob Collie (00:32:12):
Hey, I was never in a fraternity, but I hear that's how it
works.
Donald Farmer (00:32:16):
People have told you.
Rob Collie (00:32:18):
Yeah. My friends would come back and go, "Oh man, they made us
drink a gallon of milk and then do setups and so we threw up." I'm
like, "Oh, that sounds great."
Donald Farmer (00:32:28):
So you weren't just taking people into this new world as you put it
because that almost to me implied that you had this group of users
that you were then dragging... When Gemini became a Power Pivot, it
opened up the door for people like me. So you didn't just take
people, you cast even the wider net and you dragged people along
into a much better future. I cannot imagine a world where you
didn't have Power Pivot or PivotTables. All the time, I'm talking
to my wife and she's working with real estate data and I'll look
out and go, "Oh, we'll just make a PivotTable." And she's working
saying Google Sheets or something that, but it's everywhere.
Donald Farmer (00:33:07):
I'm doing Python, they're like, "Oh, so we'll just make a
PivotTable of that data frame." They're everywhere. And I can
remember a time when they didn't exist in my life and now I can't
imagine going back to that. PivotTables are a great example of a
data structure, an interface, a user interface that we're very,
very familiar with, but we're so familiar with it that we overlook
the power of them. When they first came out and VisiCalc or Lotus
1, 2, 3, that was just a tremendous breakthrough, the ability to do
that work. And if you remember at the beginning of the Gemini
process, the Power Pivot process, we actually defined our potential
user as somebody who understood PivotTables and used VLOOKUP or
HLOOKUP.
Donald Farmer (00:33:49):
And if you did those two things, then you were potentially a Gemini
user. I thought that was fun. But one of the odd things about our
users, I don't know if you remember that our business intelligence
users hated Gemini at first. You brought in the MVPs, we brought in
people Chris Webb, who now works at Microsoft. I remember Chris
saying, "This is terrible. This is a terrible product because
you're killing our careers. Our whole career is based on consulting
with people who need our help in order to do these complex
scenarios. And now you've made it so simple that they won't need us
anymore. And you're undercutting your partner base by making
everything so simple."
Donald Farmer (00:34:26):
And then either luckily or unluckily, depending on how you look at
it about a year later, Chris came back and said, "Actually, it's
not so simple after all, we're going to be okay." But it reminds me
of that because Microsoft in some ways, I think a general story
about Microsoft back at that time, we had great understanding of
user experience in certain areas, but we had a relatively poor
understanding of how user experience and the daily life of our
users came together. And I think ProClarity is a great example of
that, just bringing on board the capabilities wasn't enough to
sustain the product because ultimately the product died because
we'd forgotten that is not just about what you can do, it's about
how you do it and the daily life of the user.
Donald Farmer (00:35:12):
ProClarity users were rightly furious with us for missing that
aspect of it. That we'd actually implemented all the features and
lost all the soul of the product. Is that fair? Am I being super
hash?
Rob Collie (00:35:25):
I think that's completely fair. And I'm not even sure that we
implemented all the features. The first move after an acquisition
like that used to be, "This was the playbook," was announced that
is not going to ship again, there's no new versions of it coming
out. So you burned your ships in the harbor to commit the team to
integration, because you know that if you leave that door open,
you're just going to keep shipping that product separately forever
and it'll never get integrated. So some executive will draw a line
in the sand and say, "No, it's not going to ship anymore."
Rob Collie (00:35:58):
Well, then all the regular software engineering practices kick in
and we say, "Okay, so how do we fit this into our existing
products? And the answer of course is, "Not very well." And it's
integration, especially if you go back to that point in time at
Microsoft, even if everyone's head was on completely straight and
everyone was onsite in a way that was just not the case back then,
and even much, much less so then than now. I don't think we could
have pulled it off. How would you bring something like that
experience you're talking about of ProClarity, the life of a
ProClarity user, how would you keep that alive inside something
Excel while at the same time, not turning Excel into the tourist
trap of billboards that are all garishly colored.
Rob Collie (00:36:51):
Bill and I talked about it is, why are these amazing things in
Excel still hidden? And it's because you can't turn Excel into the
tourist trap.
Donald Farmer (00:37:00):
By Bill you mean Bill Baker or Bill Gates?
Rob Collie (00:37:04):
No. Bill Jelen.
Donald Farmer (00:37:05):
Oh, Bill Jelen. Okay. Right. Another great-
Rob Collie (00:37:07):
MrExcel. He's slot somewhere in that Pantheon?
Donald Farmer (00:37:10):
That's a great example.
Rob Collie (00:37:12):
And so that was always one of the tensions for us on the Excel team
was that we'd be working on something new and exciting. And of
course, I would want it to be pulsating in the interfacing, click
me, click me. And as a steward of the overall product, you can't do
that because next thing you know, the entire screen is posting.
Donald Farmer (00:37:30):
Well, that's why I was asking about Bill Baker or Bill Gates
because Bill Baker who ran the SQL BI team used to say that what we
have to avoid is the Fisher-Price user interface, the big yellow
button.
Rob Collie (00:37:40):
I actually like Fisher-Price interface.
Donald Farmer (00:37:43):
That also explains a lot about your specifications back in the
day.
Rob Collie (00:37:49):
Hey, it was a simpler time.
Donald Farmer (00:37:51):
No, I think it's a really important point. When we were not
developing Power Pivot, but actually marketing Power Pivot, one of
the big breakthroughs that we had there, which ultimately led to me
coming on in costume at the past conference, but one of the
breakthroughs we had, there was Daniel Yu who currently runs
marketing for Azure sign ups at Microsoft. Daniel Yu and I were
sitting down, how do we explain the value proposition of this to a
business user who doesn't know anything about it? It's one thing
talking to an analysis services user, is one thing talking to Excel
MVPs who get this, but what about the business user? What sells it
to them?
Donald Farmer (00:38:27):
And in order to find a message, we actually built a little story,
which was the story of a day in the life of the user, literally
from what do they do when he get up? What do they do on the way to
the office? What they do when you come into the office. And then
what does that story look with, or without Power Pivot? And that
really gave us a breakthrough because we started to realize that
Power Pivot fitted into their daily life, into their working life
in a way that was actually distinctly different from Excel, but
still complimentary to it, that this wasn't just something that
rolled into the Excel world. And equally, it wasn't something that
took the place of BI.
Donald Farmer (00:39:05):
They still sat down first thing in the morning when they're in
their office and looked at their reports, for example, and we're
not going to take the place of your reports and your dashboards,
but we are going to take the place of what do you do once you find
something in the report that you need to take action on, or once
you notice something in the dashboard and now you're in Power Pivot
world. And that was a real breakthrough for us, and it came from
this idea of modeling the daily life of a user.
Rob Collie (00:39:28):
You're speaking of the silent film that you made, right?
Donald Farmer (00:39:30):
Oh, you're right, we made these little films. They were cool.
Rob Collie (00:39:33):
What was that called?
Donald Farmer (00:39:34):
Oh, I can't remember. We made a few of them.
Rob Collie (00:39:37):
It's like The trouble With Data or something that?
Donald Farmer (00:39:39):
That's right, The trouble With Data. And there was few, we made a
little brought voice over it, so I'm sure they're still out there
on YouTube.
Thomas LaRock (00:39:45):
I need to see these.
Rob Collie (00:39:46):
Yeah. This was eye opening for me. I think by the time this was out
there, I think I had already left Redmond and it really got my
attention. Seeing you do certain things, opened a door for me,
which is that you can bring an artistic side to this data world.
I'm not an artist. There's nothing I can do, I can't draw, I can't
compose music. I can see things in my head. And you putting
together this silent film, this spoof of a silent film that
paralleled a day in the life of a business user of a business
analyst, really lodged in my brain. So like a lot of the things
that I've done since then like with stick figures and all this
stuff, it were in large part inspired by having seen it executed so
well.
Rob Collie (00:40:35):
And knowing that you didn't make those film segments, that you
spliced together, I'm like, "Oh, I can do that. I can do that sort
of composition." In a funny way, have you ever heard the C
language, the C programming language was invented so that they
could write the Unix operating system? Seems really bizarre. Like,
wait a second, don't you need an operating system to run the
language on? Anyway. So in a way, it's kind of like the Gemini
experience, the Power Pivot experience. In a number of ways, I
participated in building a set of tools that allowed me to become a
different version of me. I kind of stepped through a door.
Rob Collie (00:41:12):
It's not just the technical capabilities of like DAX and data
modeling and all that, which was a crucial, crucial, crucial
component, but also the inspiration of seeing things like, I think
it was called the trouble with data. I hadn't really thought about
that until you just walked right past it. And I'm really deeply
appreciative of you having done that. It's had a profound impact on
me.
Donald Farmer (00:41:31):
Well, thank you. That's actually great to hear. And I think the
work you've done since then, now that you mentioned it, I think you
have brought that. You might not think of yourself as an artist,
but you do bring a lot of creativity. The thing that we did at
Microsoft that was always effective in some ways, although we
didn't always execute well on it, was we had the resources to go
out and do things that would be difficult to do in another
business. And that included experimentation, taking some risks on
that. And the leadership were more than capable of taking that
on.
Donald Farmer (00:42:03):
But the other thing was, we also had big purpose. I remember
perhaps when I interviewed at Microsoft being told that the Excel
has 500 million users. What does an interface look like for 500
million users as opposed to the target market of a few thousand
users for another desktop application? And thinking in that scale
and thinking that big picture is really important. Rob, when you
were at Microsoft, did you ever have to present to Gates in one of
his business reviews and things?
Rob Collie (00:42:30):
One time.
Donald Farmer (00:42:31):
One time. Never again?
Rob Collie (00:42:33):
One time only. It was brutal.
Donald Farmer (00:42:37):
It was brutal. Gosh, I once had to do a one-on-one with Bill Gates
and Ray Ozzie so I guess a one on two. Actually no, because
Caroline Chao came with me, so it's two and two. Gosh, that was
nerve wracking.
Rob Collie (00:42:48):
You're like, "Okay, look, you set the pick, I'll roll."
Donald Farmer (00:42:51):
Exactly. But the fascinating thing about Bill Gates, and it's still
fascinating, still, when I see them talking and maybe in meetings
where he's presenting or something, he still has the same quality.
You know the rabbit and duck picture, you can see the rabbit, you
can see the duck? Nobody can see both at the same time, you have to
switch between the two, see a rabbit, now you see a duck depending
on your perspective or your focus. Bill had this huge image, this
huge efficient of what computing can do, what computing can do for
individuals, what it can do for businesses, what it can do for the
world and society in general.
Donald Farmer (00:43:29):
This enormous vision, which was greater than any of us could really
take in. And that was his capability. But he could also focus
tremendously on, why is that button on the left rather than the
right? But he could also do both of these things at the same time.
He could keep these focuses in mind at the same time. And so
presenting, say something like Gemini to him, you never knew which
Bill is going to show up, you never knew if the question that comes
next is going to be about the impact of this on society and the
world and the future of education, or if it's going to be, why is
that not called key influencers rather than key factors? That's the
dumbest thing I've ever seen.
Donald Farmer (00:44:09):
And you never knew which one you're going to get, which made
talking with Bill very, very difficult and a real challenge. But it
was fascinating. You learn so much from it. This ability to see the
big picture and the detailed picture at the same time.
Rob Collie (00:44:23):
My experience of that was that the detailed picture had a 50-50
mix. 50% of the time when you dive into the details, you did learn
something, and the other 50% of the time, you dive into the details
and you'd feel like it's just because he couldn't help it. It was a
derailer which could derailed the whole conversation. Our BI review
with him, we lost... This was one I wasn't invited to, because of
my poor performance at the first one we'd done. The second one
devolved for like 15 minutes about how in one of the completely
graphically designed, it was like done in Photoshop mockups,
someone had used a pie chart to represent years.
Rob Collie (00:45:03):
And he was saying, "Listen, anyone that would do years on a pie
chart," he was basically saying, "Shouldn't be working here."
Donald Farmer (00:45:11):
I've got to agree with him.
Rob Collie (00:45:16):
Yes, yes. But also looking back years later, I'm like, "Bill, this
was a culture that you created. Hiring the computer scientists, the
people who could solve these brain teasers or these academic
riddles without ever having been polluted by the real world." Of
course, as a 25-year-old, I wouldn't have known any better either.
I would've drawn a pie chart and I would've just slapped years on.
I wouldn't have ever thought twice about that. I can also
understand the stupidity of it, but sympathize completely with the
mindset that would do it.
Donald Farmer (00:45:48):
For sure. But that's actually not a derailment in a sense for Bill,
that is part of that.
Rob Collie (00:45:53):
I agree.
Donald Farmer (00:45:54):
The detail and the big picture are a part of the same thing. It's
fascinating. It was always a complex process, because I say you
never really knew who was going to turn up. He'd have been
perfectly capable of overlooking that and going on to some
completely different topic, which could have been even greater in
scope. It was always difficult to bring your story to him because
he always brought his story, and that's a real challenge. But the
guy was amazing. A tremendous privilege to have any time to work
with him at all.
Rob Collie (00:46:22):
Yeah, I concur. It's completely not what I expected. I remember
being struck by just like, I think this goes hand in hand with what
you're saying, is he had a tremendous recall of basically
everything.
Donald Farmer (00:46:34):
Yeah, that's true.
Rob Collie (00:46:37):
The data structures that he uses to record things about the world,
he has this like lossless storage and the ability to index into any
particular thing no matter how detailed.
Donald Farmer (00:46:50):
His brain is an array of pointers.
Rob Collie (00:46:52):
Yeah, that's right. It wouldn't surprise me if he remembered what
shirt I was wearing that day that he chewed me out. That would be
completely believable 20 years later.
Donald Farmer (00:47:02):
Yeah, absolutely. Well, it's that connectivity, which a lot of
people find difficult, the fact that everything is connected for
him. The really creative people have that, and we don't all have it
in the same degree. But he has that extraordinary connectivity. And
the most interesting people I've worked with have it in all sorts
of ways, they're able to make those connections. And then they're
also able to make sense of new connections and to create
connections. You'll remember, Rob, when I was at Microsoft and I
was your manager, whenever I wanted to do a one-on-one or whenever
you wanted to do a one-on-one, I always wanted to go for a walk. I
always wanted to walk into campus walk, walk around the trails
around Microsoft.
Donald Farmer (00:47:37):
And one of the reasons I always enjoyed doing it is not just
because I like getting out, but because your conversation always go
somewhere else when you're walking. And there's two dynamics to it.
One is that you can't be face to face with someone when you're
walking unless you walk into each other. You have a lot of contact,
you have a lot of ability to read the other person, but you're not
constantly focused on their micro expressions and how they're
feeling, responding to you, whatever. But the other thing is that
new things happen when you walk, a squirrel runs across in front of
you, you see something, a beautiful flower, that you need to
describe. You see some crazy building work going on in campus and
you need to talk about that.
Donald Farmer (00:48:18):
And it takes you off subject, but it also provides you with a whole
new set of connections that you wouldn't have if you just sat down
in the office and speaking to each other. And I noticed this just
so much now that we're all connecting online, that we just don't
have that extra input. Our conversations, not this one, but our
conversations... Because we haven't spoken for some time, this is a
great conversation. But very often, my weekly conversations with
people or my daily conversations with people just the same old,
same old, because there's no new stimulus into that conversation
except what has happened in the business since the last time we
spoke, which gets old for pretty quickly.
Donald Farmer (00:48:55):
And I can't wait for the time that I have walking conversations
with people. I've had one of the companies I advise, the lady I'm
working with, she came up from Portland in order to have a socially
distant walk with me. So we were walking along the trail here,
eight feet apart, yelling at each other because she knew we had to
have a conversation while we were walking rather than sitting on
Zoom when we just had the same old conversation.
Thomas LaRock (00:49:19):
You mean yelling because you were eight feet apart and not because
you were angry with each other?
Donald Farmer (00:49:23):
Right, because we were eight feet apart. Yeah.
Thomas LaRock (00:49:24):
Yeah, okay. Because you basically just described every walk I have
with my wife, but that's fine.
Rob Collie (00:49:30):
Don, did I ever tell you about when I had a walking treadmill
desk?
Donald Farmer (00:49:34):
I know you had one. I could hear it in the background sometimes
when you called me.
Rob Collie (00:49:38):
Yeah. The idea was to get skinny while working. That didn't happen.
That low intensity exercise, I just would replace the calories.
Whatever calories I burned walking those eight miles that day, I
would just re-eat them and nothing changed. But I did find, even
just walking on a treadmill desk by myself, that I focused in a
different way, just the movement. So in addition to everything
thing that you're saying, there's also, I'd like to add one more
benefit, which is that just physically moving-
Donald Farmer (00:50:07):
It's that slight stimulation of the heartbeat and increase of
oxygen.
Rob Collie (00:50:09):
Yeah. And for whatever reason, I don't struggle with this as much
as I used to. But for a while there, like for a good decade of my
life, I would sit down in a chair to work, my ADD or whatever would
immediately have me opening a browser window and like navigating to
the ESPN homepage without even knowing that I'd done it. It was
like this craving of stimulation. And for whatever reason, when I
was walking, I was scratching that itch. That part of me, that
fidgety part of me had something to do, and that was walk. And then
my brain could actually go a lot more focused work. I missed the
treadmill desk because of that.
Rob Collie (00:50:47):
Eventually, the treadmill desk broke and I never bothered to
replace it. It didn't have its magical redefining shape properties
on my body that I was hoping for. But I do miss the focus, the
clarity, the calm that it brought, even though I'm just walking in
place, in a corner of a room. Maybe we should all get walking
treadmills and conduct our meetings that way. It would be like a
control. And then every now and then, we could just like release a
squirrel in the room.
Donald Farmer (00:51:15):
Well, it's going to be fascinating over the next year, probably two
years, as we start to look back on this very strange pandemic year
that we've had to see what working practices were effective and
which weren't, and then which ones we want to carry forward. I hear
a lot of people saying they don't want to go back to the office, I
hear other people saying that they're to go back to the office, and
it's going to be fascinating just over the next year to see how
that plays out.
Rob Collie (00:51:41):
Yeah. Once it's a choice, once it becomes optional, whether for the
individual or the company as a policy, that's when you're going to
find out for real. Right now, it's all what-if.
Donald Farmer (00:51:54):
Well, I'll tell you, I speak to quite a few startups and smaller
companies. And the managers of those companies were often telling
me that they really miss having people in the office. I speak to
the employees and they're not actually missing it so much. But what
people seem to miss, or they say that they miss when they talk
about, I wish we could all be in the office together is, "Oh, we
miss the creativity, we miss the interaction." And then I speak to
other teams and they say, "Actually, we don't miss that all. We're
doing fine without all that. Thanks very much."
Donald Farmer (00:52:21):
And I've done some analysis of it, not just making it up, I've done
some analysis of it for one company in particular, and we did some
surveys. And it turns out that, of course, the manager in some ways
is right, that there is a missing creativity and connection, but
it's their inability to scale the conversations. They'd walk around
the corridors, they'd see somebody at the water bottle and they'd
have a conversation about it, they'd stand having coffee. They're
not organizing a meeting, they're not scheduling that, they're not
sitting down with the person. In fact, it might be a very casual
conversation in which they can give a pointer, "Well, maybe you
should think about it this way. Maybe you should think about it
that way."
Donald Farmer (00:52:58):
It's the very light touch interaction. And to them as managers,
that's a critical part of their world, not sitting down having
regularly scheduled conversations, but the light touch interaction
that enables them to, as they see it, scale their interactivity and
creativity across an entire team, which you can't do when you're
all sitting on a Zoom meeting. On the other hand, it turns out that
a lot of the developers, especially, are actually very happy not to
have the manager giving their ad hoc direction, because they see
that as randomizing and distracting.
Donald Farmer (00:53:30):
So from the manager's point of view, it's, "Oh, I'm so creative, I
can have these light touch points and that's me adding value." And
to the developers, that's, "Oh my God, that guy's going on about
years on the pie chart again, just let me get on wiTh mu job."
Rob Collie (00:53:43):
You were there for no-meeting Wednesdays, right?
Donald Farmer (00:53:45):
Oh, no-meeting Wednesdays. Yeah, that was a great day for
meetings.
Rob Collie (00:53:46):
Yeah, that was great. And you remember me calling it snarkly
no-decision Wednesdays? Again, it's pros and cons. The developers
loved it, but I always felt like as a program manager that I might
as well not even come in that day.
Donald Farmer (00:54:01):
Yeah. Nothing to do.
Rob Collie (00:54:03):
We couldn't do anything without the people.
Donald Farmer (00:54:05):
Well, I don't know if you were there at a point where I gave some
of my team pedometers. Did I give you a pedometer?
Rob Collie (00:54:11):
I never got a pedometer. I demand a pedometer.
Donald Farmer (00:54:12):
Well, I'll send you one. No, the idea was that program managers
shouldn't be in their office. If you're in your office, you're not
doing your job. You need to be out there talking to people.
Rob Collie (00:54:20):
Yeah. Just blows my mind that there are program managers that can
work today remotely ignoring the pandemic. The program management
position has now become somewhat remote friendly at Microsoft. I
struggled with the phone, I really relied on being able to go and
talk to someone face to face.
Donald Farmer (00:54:41):
Absolutely. Yeah.
Rob Collie (00:54:44):
It would've been crippling for me to have had that taken away. And
so it made sense to me that when I moved away from Redmond, that I
wasn't going to be doing that job anymore. I look back now and,
again, I go, "Oh, there are a lot of people doing that job remotely
now, I wonder how that works."
Donald Farmer (00:54:57):
I can't remember exactly what it was, what the issue was, but there
was some issue in specification that had a dependency on another
team. And at that time, we were in Building 35 and the other team,
SQL Server core team was in 34, and they were not supporting us.
And we got these emails from their program manager saying, "We're
not going to do this feature. We're not going to do that feature."
And we really wanted this feature in order to enable our own work.
And the program manager who was reporting to me, she was kind of,
"I don't know what to do about this. They keep blowing me off. I
keep raising it in meetings, they don't do anything."
Donald Farmer (00:55:30):
And I said, "Well, let's just go talk to them." "Well, I don't have
a meeting arranged, I'll set something up." "No, let's just go and
camp out in their office." So we walked across to the building at,
across to 34, down into the basement, sat in the office. The guy
wasn't there, waited for him to come back. He came back in, and we
said, "So here's what we need and here's what you're not giving
us." And what I'd said to my program manager was, "It's going to be
much more difficult for him face to face, sitting office to say,
'No, you'll not get it. We're not doing that.'" And I was so
disappointed and I said, "Okay, well, we're going to push for this.
We're going to keep pushing for it."
Donald Farmer (00:56:06):
We came back, went back to our offices. I clearly had lost all
credibility as a manager and as a reader of human beings. And so
the next morning in ship room, guess what? They'd implemented the
feature overnight and done it. It was kind of funny. It was very
revealing that, yeah, he still was able to say no, but actually,
deep down, he understood our need at that point because we'd seen
face to face. It was kind of funny.
Rob Collie (00:56:32):
I mentioned earlier that you were in many ways, the inspiration for
some of the more creative efforts that came later in my career, but
you also killed one.
Donald Farmer (00:56:41):
Oh no.
Rob Collie (00:56:42):
You killed Power Pivot Yoda.
Donald Farmer (00:56:43):
Power Pivot Yoda. Wait, what?
Rob Collie (00:56:48):
Yeah. Power Pivot Yoda was going to have a bright future.
Donald Farmer (00:56:51):
Is this like a Clippy thing?
Rob Collie (00:56:53):
No, Power Pivot Yoda was a Twitter account. That was a photo morph
of Yoda and Donald. It was really kind of amazing. It was like one
of those one in a million things. Donald was the face of Power
Pivot, he was the community voice at Microsoft for all of this
stuff at the time. It was a funny little joke. And then like three
months later is when Donald left for Qlik and like, "Well, I can't
use Donald's likeness as Power Pivot Yoda anymore." And so that was
the end of one of my creations. But you miss 100% of the photo
morphs you don't make.
Thomas LaRock (00:57:27):
I don't understand why we can't just do that now.
Donald Farmer (00:57:29):
Well, do you know the irony of that? As I went to Qlik and one of
the first things their marketing did was build an entire comic
strip around Qlik Yoda, which was me as Qlik Yoda, which I suspect
they got from Power Pivot Yoda. So your idea in a sense did live
on-
Rob Collie (00:57:44):
I didn't know that.
Donald Farmer (00:57:45):
... just in the service of a different business,
Rob Collie (00:57:47):
I feel redeemed hearing that later. I probably would've been upset
at the time, but now is a really good time to hear that. That's
great. The other things, Tom, that happened back in that day is...
Don is such a good sport, and is so well known and so loved, that
he is just a really easy figure to have some fun with. And so I
posted this photo gallery of like, "Is it Donald or is it Tommy
Chong?"
Donald Farmer (00:58:12):
I remember that. That was on camera.
Rob Collie (00:58:17):
And it was really close. And it was so funny that someone else,
about six months later wrote an article about Donald and used a
Tommy Chong picture on accident. So you know where he got it.
Donald Farmer (00:58:30):
That's great. Yeah, that was crazy.
Thomas LaRock (00:58:36):
I want to ask something slightly off topic, but I'm going to ask
anyway. I remember this event, I don't know who, I suspect it's the
SQL Server marketing team. I don't know what it is, but I believe
Donald, you hosted a wine pairing with SQL Server in some secret
room like the back cave or something inside of Microsoft. I
remember I was invited to fly to Redmond.
Donald Farmer (00:59:01):
That's right.
Thomas LaRock (00:59:03):
Was it for SQL Server 2008 R2? It had to be. There's this launch
event for what is essentially, let's call it a mid-major release of
SQL, and it's a wine pairing with Donald Farmer. So what was that
room first of all? What's that secret room on campus?
Donald Farmer (00:59:23):
That secret room was in the Executive Briefing Center and it was at
the back of the Executive Briefing Center where they had a section,
which was the Office of Tomorrow. And then at the back of that, it
was a set that they'd created for the Home of Tomorrow. And that
had actually gotten nowhere and they'd more or less abandoned it,
but it was beautifully equipped as this lovely home with big screen
televisions and couches and things out there. And that's where we
held that event.
Thomas LaRock (00:59:51):
Again, I was invited, but I couldn't make it. So I'm assuming
because I was still working for a previous employer. And I remember
photos from the event on social media where Donald was basically
introducing one of the features in 2008 R2 and he was pairing it
with a class of wine. Do I have this correct, sir?
Donald Farmer (01:00:12):
You do. Yeah, exactly. Yeah. And we had these funny things about IT
and business working together like wine and cheese. It sounds
really corny, but actually it is really corny. Of course, it was an
excuse to have good wine and nice cheese, but it was also a way of
getting people into a different minds set so they thought of this
less as a set of features and more as the creation of a community
of people who felt connected by a shared experience of this
product. So that R2 launch, we had a group of people who felt very
committed to each other and to us as a team. It was social
engineering.
Thomas LaRock (01:00:49):
Yeah. I am sorry I missed that, because as it was happening, I'm
like, "Oh, well, they'll do it again next release and maybe I'll
get to go to that one." It's never happened since. So it's been
like 12 years and there has yet to be a decent SQL Server launch
event with wine pairing. And I'm going to have to mention it to the
folks, MVP Summits coming up. I think I'm going to have to put the
bug in some of these ear about this because absolutely this should
be happening and I should be invited. That's how I feel.
Donald Farmer (01:01:17):
I'm a great believer in actually building community and in building
that sense of shared purpose. And far too often, we actually
present products to people as if they are passive receivers of our
products rather than part of our shared creative community. When I
joined Qlik, for example, Qlik at a really bad reputation for
talking down to industry analysts and for regarding industry
analysts as people who didn't really understand Qlik and didn't
understand its unique value proposition. And therefore, we always
mansplain to them about why we were so unique and different.
Donald Farmer (01:01:52):
One of the things we did was at a tech event in Vegas, where there
was a lot of analysts and influencers present, we did a little
event that I set up and the marketing team were great in supporting
this Lara Shackelford at the marketing team, did a great job of
pulling it together. We actually held it in the art gallery of the
Bellagio rather than the typical Vegas event we have in some kind
of meeting room, no. Scott Humphrey, my friend, who's an amazing PR
guy, we got the art gallery, we held it in the art gallery. We had
the nibbles, we had the wine, we had the pairing and we didn't talk
about the product, we talked about us.
Donald Farmer (01:02:28):
We talked about why we were at Qlik, what we were doing, what we
wanted to achieve. And we talked about that apart from a brief
introduction, we talked about one-on-one with people rather than
presenting to them. And what happened after that event was the
analysts were our friends, literally. There were people at Qlik who
had never sat down and had a meal or a drink with an analyst, and
the whole experience then changed because they now became part, of
course they weren't part of our community uniquely because they're
industry analysts, they're objective, but we did have this sense
that we're all part of a shared community of purpose, community of
practice that I felt was really important.
Donald Farmer (01:03:05):
And I think you're right, by the way, Thomas, I think Microsoft
have lost a lot of that. They're still doing this thing where
Microsoft is explaining their products to you rather than we are
part of a shared community of purpose. To be fair, I've been doing
a little work with the Azure signups team, where of course it's
been really difficult in the last year because we can't have those
events. We plan to have a great event in Woodville, pull people
together and it was going to happen last year, and of course it
couldn't. I think we need to do a lot more of that, of building,
not a community in a sense of a support group, a community, isn't a
support group.
Donald Farmer (01:03:39):
You don't need to join a community, you're part of a community
because that's who you are. I was brought up in a small village and
you don't join a village, you're in a village because you were born
there, because you live, because love them or hate them you're in a
shared community with other people. You join a user group, you join
a support group, you log onto a forum, but a community is much more
organic than that, and you have to do a lot of work to build that.
And I think be perfectly honest, I think Microsoft on the BI side,
on the SQL Server side, on the Azure side to a certain extent lost
that sense of shared community of purpose.
Rob Collie (01:04:12):
Yeah. I was thinking about Tom, as you were saying, I need to get
on it and let them know they need to do a wine event. I was
thinking about things Donald had said earlier like the Lampoon
version of this as Microsoft will say, "Oh right, yeah, we should
do that again." So they'll analyze the ingredients of that event.
So there was wine, there was cheese, and so they'll have like three
different, "Okay, we need two whites and two reds," and they'll
just take the first absolute vanilla versions of all of those, and
you'll get a handful of cheeses and think that that's the same
thing. That's the whole point is that it's not.
Donald Farmer (01:04:45):
Well, nowadays you would just do it online and send everybody a
coupon to go to Costco and that's the same thing.
Rob Collie (01:04:52):
Yeah. And I'd come home with hot sauce or something?
Thomas LaRock (01:04:55):
I think I have to agree with Donald, Microsoft may have lost a
little bit of that community focused. What I will say is they
continue to hire from their MVP community, which I think is
somewhat brilliant because that's instant credibility with your
best advocates, because now the people that you were sitting in
that room with, now one of you, they're the principal program
manager in charge of whatever. And it's really hard to look at
somebody that was a friend and a colleague, you only assume good
intentions. You know they're a good person, you know now they're
doing good work over at Microsoft.
Thomas LaRock (01:05:35):
And so I find that there's a nice, solid connection between people
building the product and their advocates, their community of MVPs.
But I think you're right in the sense that outside of that
community, that's still fairly closed, that there hasn't been that
outreach beyond that circle. I don't know how that's happened, but
I'm just reflecting back on the past, say 15 to 20 years of my
involvement in this community. And I think they have gotten away
from that. I'm not sure why or how.
Donald Farmer (01:06:09):
Well, I think it's partly to do with the success of the product.
I'm going to make a suggestion here which they may well dislike,
but I think it's true, is that earlier I was talking about the
importance of tools of choice as opposed tools which are
provisioned and Power BI is no longer a tool of choice. That's not
entirely true because of course there are people who choose to use
Power BI, but mostly it's just provision. It's provision by IT, it
comes, I don't know, we're not allowed to use the word bundling in
the Microsoft context, but it comes bundled with all fees or it
becomes bundled with an Azure deal in some way.
Donald Farmer (01:06:41):
Very few individuals sit down and say, "I want to use Power BI."
There are some who do and they're great at it, don't get me wrong.
But unlike the world of the newer tools, and I think Tableau again
is a great example of a company that has always done a good job of
this. People choose to use Tableau, people choose to use Power BI
in anything like the same way.
Rob Collie (01:07:01):
This jumps off the page to me in any interaction that I have with
Microsoft these days is that they're doing such a good job selling
the product at that top-down level. And I come at it from a
completely different perspective, which is, the bottom-up adoption.
I never had that phrase, tool of choice, but now I do. Excel is one
of Microsoft's only tools of choice, it's an accidental tool of
choice. The people who are good at Excel really love that product.
They love it. There aren't many things like that at Microsoft. And
Microsoft's inability so far to directly harness that audience
that's already theirs and turn them into passionate Power BI
developers still really bugs me, even as they've been really
successful.
Rob Collie (01:07:48):
Financially, the product is doing incredibly well. And look at the
Magic Quadrant, look at how well it's being received by those
analysts we talked about earlier, in some sense like on every
metric that matters, they're doing incredibly well. But it is hard
for me to connect with them on that bottom-up level because it
doesn't really seem to be hurting them.
Donald Farmer (01:08:07):
No, no, it doesn't. But you can see it in the product, and this is
something that's super interesting to me. It does get reflected in
the product because what I see in the product is there are features
which have to my mind clearly been added almost as demo features.
There are very lightweight, not particularly well thought through
features. And in fact, very often I come back and talk to people a
year later, question and answer is a great example, the natural
language interface. How many people have actually been using that
for two years and are still using it as a daily practice after two
years?
Donald Farmer (01:08:38):
It's not a long term feature that. It's not a product, not a
feature that you use for two years. As soon as you've learned how
to use it, you abandon it and you go straight to the data again.
And there's a lot of features like that in Power BI that
continually pop up. And what I find when I look at those features
and when I start working with them is inconsistency of attention.
There are some aspects of Power BI where clearly people are giving
it long term focused attention on the quality of this feature. And
there's other features, which just feel like they've been hacked
together and pushed in order to check a box, "Oh, ThoughtSpot are
doing this so we'll do that and compete against them."
Donald Farmer (01:09:18):
"Oh, Tableau, I've got this, we've got that too." And that
inconsistency of attention, I think, is actually long run
potentially damaging. And the great success of Power BI is in some
ways also its undoing because it's going to be just a default
product, and at some point, because it's default, it'll no longer
move the needle.
Rob Collie (01:09:40):
Yeah. And when it's unenthusiastically adopted, it's adopted
incorrectly.
Donald Farmer (01:09:47):
A lot of that.
Rob Collie (01:09:48):
And it doesn't work well. It gets a negative reputation amongst
that workgroup because they watched those demos and everything was
so easy, but they also completely miss that it is a world-beater.
When properly utilized it is a world-beater of a tool that should
have been charged ridiculous amounts of money for. But you're
right, now it's just part of the E whatever, skew, and it's issued
to you. And my conversation with Microsoft, for good reason, you
can understand this, it's not a bad thing, their overwhelming
concern is what the CIO thinks of the tool. And the thing that
makes it truly great, the thing that makes it the tool of choice,
that people get excited about has nothing to do with what the CIO
thinks.
Rob Collie (01:10:38):
You're right, it probably does long term, and it is, even medium
term, short term. This is causing them trouble, adoption after the
sale is now something that's a concern for them.
Donald Farmer (01:10:49):
For sure. There's a lot of shelfware out there. We used to call it
shelfware, I don't know what you call it when it's on the
cloud.
Rob Collie (01:10:55):
You might as well call it shelfware, it's a virtual shelf.
Donald Farmer (01:10:57):
And what's super interesting, of course, is the CIO may well be
adopting Power BI for all the right reasons, but the tools of
choice in the organization are other tools of choice and are
starting to have that impact.
Rob Collie (01:11:08):
Yeah. What's TreeHive Strategy? What are you up to these days?
Donald Farmer (01:11:10):
Well, TreeHive Strategy is just a cute name for my consulting
practice. And what I primarily do nowadays is strategic advising to
vendors, to investors, and to enterprises about data and analytics
strategy. We have a simple mantra that any software company is a
data company, no matter what you do. If you're a data company, you
need to be an analytics company, otherwise your data's just sitting
there and not taking action requires analysis. And if you're an
analytics company, you probably need to be an advanced analytics
company sooner than you imagine.
Donald Farmer (01:11:42):
And therefore, what does that map look like to get from being a
company that has data to being a company that actually is actively
and innovatively using it? I have a lot of fun, I speak with a lot
of investors, which is great, because I can go into their portfolio
companies and then I start to deal with a whole lot of software
vendors doing everything from manufacturing and retail, restaurant
software, all sorts of things that I otherwise might not get to do.
So I enjoy that, and I help vendors large and small from the
largest vendors to small startups. And I work with enterprises
ranging from manufacturing and robotics.
Donald Farmer (01:12:15):
So I have a lot of fun and I'm lucky in the sense that I still have
to work, but I can choose what I do, and that's a great
pleasure.
Rob Collie (01:12:22):
And just so people know, where is the name TreeHive come from?
Donald Farmer (01:12:26):
Well, TreeHive comes from our treehouse. There's a great TV
program. Well, I think it stopped now, but by Pete Nelson, who's a
treehouse builder based in Fall City, which is about 15 miles from
us. And he built a treehouse for us and it was on the TV program
and my wife built a sketch of it. It looked like a beehive in a
tree, and Pete Nelson came along and said, "Well, that's
impossible, let's try it," which is a great attitude, and built it.
And it's beautiful. If you search for beehive treehouse, you'll
always certainly find it. It's a beautiful place. And that's my
thinking space.
Donald Farmer (01:12:57):
I retreat up there in the afternoon, no Wi-Fi up there, just the
ability to sit there and be on my own and think. It's great to have
a thinking space. So that's TreeHive Strategy.
Rob Collie (01:13:06):
Yeah. I remember there not being Wi-Fi in the TreeHive, but I was
wondering all these years later, if you broke it down and hooked it
up.
Donald Farmer (01:13:14):
No. Absolutely not, you've got to have somewhere that you can't get
Wi-Fi.
Rob Collie (01:13:18):
Yeah. The pictures online and the videos, they actually
surprisingly do it justice. I've been in the TreeHive, I've made
that pilgrimage and it's awesome. It is just so cool.
Donald Farmer (01:13:28):
It is very beautiful, it's just a beautiful place and we're
tremendously lucky to have it. And so it's become almost a defining
feature of our work. My wife's an artist and a wonderful artist and
I do my creative thing and it's become an inspiration for us in
many ways.
Rob Collie (01:13:43):
Do you still have the Goblin Hut?
Donald Farmer (01:13:45):
We still have the Goblin Hut, which is a little octagonal building
that we can go into and escape from as well. That's fun. We have an
art studio. We have, what's called a House of Doors, which is a
building made entirely out of doors, which is very interesting. And
we have a little log cabin codding shade because our builder, our
contractor is a log cabin builder and so he built us a porting
shade, which is actually a log cabin. And we have all sorts of
stuff and everything's painted as well. And that's the other thing,
my wife's an artist so anything that's doesn't move gets
painted.
Rob Collie (01:14:14):
Back in the day, there was even one of your wife's works, put a
picture of it on the blog, the Three Seconds of Now. That painting
that was in your office at Microsoft for so long. I've always been
captivated by that work.
Donald Farmer (01:14:26):
Yeah. A big six-foot square painting.
Rob Collie (01:14:28):
It makes an impression and it is, it's just called Now, right?
Donald Farmer (01:14:31):
It's called Now, yeah. Well, there's this psychological thing that
when we talk about the period of now, when you talk about something
happening now, we're actually talking about a three-second period
of time typically. In all the languages of the world, a line of
poetry typically takes three seconds to say. So three seconds
becomes this point at which a thought is focused. And if something
takes more than three seconds, your mind drifts away from it. So I
actually use it a lot. It's analytical, it's in our user
experiences, in our thinking, we've got three seconds to hold
someone's attention and have them move on to the next thing.
Rob Collie (01:15:04):
Yeah. If you click a filter and the results paint within three
seconds, your mind doesn't really wander. On the other side of that
three-second boundary, you've become asynchronous, you're no longer
engaged. And that was the point of that blog post all those years
ago was to try to get your slicer click, refresh in your Power
Pivot reports, three seconds or less. It's like the biological
quanta for time.
Donald Farmer (01:15:28):
It is, exactly. And of course, nowadays we expect the results to
come back in millisecond, but the point being if you've actually
overloaded the user with information so it takes some more than
three seconds to take in all the data and information that you are
showing them, you've still lost them. So even if your user
interface is highly responsive, there's still three seconds of
thinking and absorbing that you need to do.
Rob Collie (01:15:51):
All right. I'm looking forward to getting back out and hanging out
in the Goblin Hut or in the TreeHive. The House of Doors was under
construction last time I was there. That's how long ago?
Donald Farmer (01:16:00):
Wow. Guess what, that's all completed. Well, we need to get you out
here, we need to get Thomas out here so we can do some wine
tasting.
Rob Collie (01:16:05):
That's right. And we'll do it right.
Donald Farmer (01:16:07):
We'll do it right.
Thomas LaRock (01:16:08):
And we'll discuss Tableau, I guess.
Donald Farmer (01:16:15):
We can do that.
Thomas LaRock (01:16:15):
I'm sorry, no, we won't discuss Tableau, we'll discuss why we
wanted to work for Tableau or something.
Rob Collie (01:16:21):
We'll discuss why the Microsoft SQL Server data mining tools for
Office System 2000, whatever, pairs so well with this Kranti.
Donald Farmer (01:16:33):
That's exciting.
Rob Collie (01:16:35):
Well, Donald, thank you so much. I appreciate you coming down out
of the TreeHive to where the Wi-Fi reaches to have a conversation
with us.
Donald Farmer (01:16:42):
It's been a tremendous pleasure. Thank you Rob and Tom.
Announcer (01:16:45):
Thanks for listening to the Raw Data by P3 Adaptive Podcast. Let
the experts at P3 Adaptive help your business. Just go to
p3adaptive.com. Have a data day!