Jan 11, 2022
Lars Schreiber is another one of the great examples of the human element of data that we love! He's an MVP, a long-standing member of the Power BI Community, and his passion for the Power Platform is quite catchy. Catch up with Lars at his website, Self Service BI Blog!
References in this episode:
Lars' Podcast with Jeffrey Wang
Lars' PowerPivotPro Blog Post from 2014
Episode transcript:
Rob Collie (00:00:00):
Hello friends. Today, we continue our tour to force for 2022 by
welcoming MVP, Lars Schreiber to the show. You might know him today
as a member of the Power BI community, and also a host of a
podcast, which we're working on backstage for me to be on his
podcast, because you know, fair is fair. But I've known Lars for a
very long time. He was actually one of the OG readers of the old
Power Pivot Pro blog. Not just a reader, he was actually also a
guest blogger. The fact that he wrote a guest article for our old
website eight years ago, led to a somewhat familiar conversation
about how as you get better and better with Power BI you're
increasingly embarrassed by your past work. That might sound
familiar to a number of you out there. In that vein, we also
discuss how a lot of the DAX in particular on the internet these
days, I can't imagine how intimidating that must be to a
beginner.
Rob Collie (00:00:53):
On the less technical front, we talk about how he went solo without
securing base of customers firsthand. He just sort of stepped off
that ledge, things turned out well for him, but didn't exactly go
the way originally that he expected. We talk about some of the
nuances of the German language. We also get on the way back
machine, and talk a little bit about how Microsoft Office
internationalized its product back in the late '90 and the very
special and funny role that German played in our lives back in that
point in time. We talk about the disturbing trend of Power BI being
used, not as an analysis tool, but as a PowerPoint replacement
where literally people are using Power BI to draw charts to display
conclusions that they've already reached using their gut. We talk a
lot of out the power of naming. He asked me a very interesting
question about the origin of the Power Pivot name. We talk a little
bit about the pros and cons and the strengths of that name and the
weaknesses. And in general, we had a blast. Tom might have gotten a
little bit over his 7% threshold in this episode, so we're going to
be keeping a close eye on that if you know what I mean. So without
further monologuing, let's get into it.
Announcer (00:02:04):
Ladies and gentlemen, may I have your attention please?
Announcer (00:02:11):
This is The Raw Data by P3 Adaptive Podcast with your host Rob
Collie and your co-host Thomas LaRock. Find out what the experts at
P3 Adaptive can for your business. Just go to P3adaptive.com. Raw
data by P3 Adaptive is data with the human element.
Rob Collie (00:02:35):
Welcome to the show. Lars Schreiber, is that how you would
pronounce it? How do you say it in Germany?
Lars Schreiber (00:02:40):
I say Lars Schreiber. See there's that extra syllable it, is a
three syllable word when you do it right. There's a little pause,
little breath in the middle there. Anyway, I'll never master that,
like I never mastered rolling my Rs in Spanish. Really, really
happy you're here. I know it's late for you. You're from Germany,
are you in Iceland right now?
Lars Schreiber (00:03:02):
No, I would love to be, but I'm not in Iceland. I'm in Hamburg,
where I live.
Rob Collie (00:03:06):
What does loving Iceland on your Twitter bio mean?
Lars Schreiber (00:03:10):
It means I'm loving Iceland. The country is amazing and the people
there are amazing too, so I've been there three times on vacation,
and this country is nothing but beautiful. We've
Rob Collie (00:03:21):
Been there once as a family a few years ago. It was just an epic
trip. I mean, holy cow, it is an amazing country, isn't it?
Lars Schreiber (00:03:29):
Yeah.
Rob Collie (00:03:30):
And apparently I know that, you know this because you've been
there. Here's a little unexpected Iceland trivia; some very, very
outsized percentage of the aluminum refining in Europe takes place
in Iceland, because it's a very, very electrical intensive process
and they just have free power geothermal and hydroelectric. They
just have endless electricity in Iceland. I guess we should also
expect them to be at the forefront of Bitcoin mining in Europe as
well, right? All right. So loving Iceland, but not living in
Iceland.
Lars Schreiber (00:04:05):
No, unfortunately not. My wife wouldn't go there, and honestly, I
wouldn't know how to earn my money there because I know nothing
about sheep, and they have so many of them and earn their money
with that.
Rob Collie (00:04:17):
But Lars, they also have horses.
Lars Schreiber (00:04:18):
Yeah. I know. I know nothing about horses as well.
Rob Collie (00:04:21):
They have both professions.
Lars Schreiber (00:04:23):
They do. Reykjavik is a beautiful town, but when I want to live in
Iceland, I don't want to go to a big city. I want to see the
countryside and see all the waterfalls and stuff, and I wouldn't
know how to earn my money there, so I stay in Hamburg.
Thomas LaRock (00:04:37):
Iceland is the one with the Blue Lagoon, right?
Lars Schreiber (00:04:40):
Yes.
Thomas LaRock (00:04:40):
Reykjavik?
Lars Schreiber (00:04:41):
Yeah, exactly.
Thomas LaRock (00:04:42):
I want to go there.
Lars Schreiber (00:04:43):
It's a beautiful place. But I guess the Blue Lagoon is the most
tourist made thing of Iceland. You don't see Iceland there
really.
Rob Collie (00:04:52):
It is by far the most tourist trappy place in the country. However,
you still got to do it. I mean, there's nothing quite like getting
off the airplane like we did, the overnight, you never really even
slept, you're just like blurry eyed. You drive in the cold because
you know, even though it's June, it's still cold in the morning,
you drive to this place, you don't speak the language. They speak
the language, they speak English. Still it's very disorienting when
you haven't slept. By the way, we changed in the airport, bathrooms
at RA Vic airport into our bathing suits, but then put winter coats
back over the top of them, went to this place. You go in, you
freeze to death, and then you get in the warm water, and then you
have cocktails. The weirdest sequence of events, one of the
weirdest anyway, end-to-end. The flight for you isn't quite as
long. I wouldn't think.
Lars Schreiber (00:05:35):
It's two and a half or three hours from here.
Rob Collie (00:05:37):
It was surprisingly short for us. Those polar routes like six or
seven hours, but it was still overnight and didn't sleep and all
that. So Lars, this is what we do on the podcast about data; we
talk about Iceland.
Thomas LaRock (00:05:49):
The Iceland podcast. Yeah.
Rob Collie (00:05:51):
Yeah. Raw Data, a show about Iceland. Have you snorkeled with the
dry suit in that river, essentially? The spring that runs between
the North American tectonic plate and the Eurasian plate?
Lars Schreiber (00:06:04):
No, I haven't. I've seen a lot in Iceland, but I didn't do
that.
Rob Collie (00:06:08):
Well, I mean here of Mr. Fancy pants, he's been to Iceland three
times, but he hasn't snorkeled the tectonic divide.
Lars Schreiber (00:06:16):
Did you you do that?
Rob Collie (00:06:17):
We did because my wife planned our trip and she's amazing. We did
do that. It's pretty awesome. Okay, so you said it'd just be so
hard to imagine earning a living, living in Iceland, even though
they offer sheep, and horses, and aluminum refining, and power, and
the Blue Lagoon. You could work there. So how do you earn a living
today? What would you be leaving behind professionally if you left
Hamburg?
Lars Schreiber (00:06:43):
Honestly, I'm living my dream right now. I'm a freelancer, earning
my money with Power BI. And because my background is business
administration, that was not really what I was supposed to do, but
I always liked to do this day data stuff. Starting with Excel, then
doing a lot of VBA then Power Pivot with and without space in the
middle.
Rob Collie (00:07:06):
Yeah. You were one of the OG, the originals.
Lars Schreiber (00:07:08):
And back in 2018, I decided to become a freelancer and it works
out. It's fun.
Rob Collie (00:07:13):
So you built a client base for yourself, geographically around you
in Germany. We've had a number of people on the show who are
similarly living this dream, who at some point, like 2018 for you,
went rogue, went solo. Not really rogue; solo, independent. And one
of the things I always like to ask people is, how did you know when
it was the right time? And it's a leap? Did you already kind of
have clients lined up? That's usually how people do it; they're
already trying to know they're going to have plenty of work to do.
How did that go for you?
Lars Schreiber (00:07:45):
I always had side projects next to my employment. And back then it
was VBA. I always had VBA projects for clients and I did a lot of
VBA for reporting. And when Power Pivot came up and Power Query and
of course, Power BI with its visualizations, all that. Yeah, I
could ignore VBI, didn't write email line of VBA anymore, and I did
it with this. And because you're asking for, "What was the initial
point where you decided to become a freelancer?" That was something
that was maybe not the best idea to use that as a starting point.
That was my MVP, and I was working in the community, and many
people knew me in the community, but I misunderstood being
recognized in the community, and having a client base. That is
something completely different. And when I became an MVP in 2017
and I always wanted to be a freelancer, I looked at my wife and she
said, "Do it." And I said, "Okay, if not now, then never." And I
thought, "Okay, now the clients will come and they didn't come in
the beginning."
Rob Collie (00:08:53):
I'm so sorry.
Lars Schreiber (00:08:54):
It was a tough time. It was a lot of learning, but it works out
now.
Rob Collie (00:08:59):
But I think the truth about the MVP program, like so many things,
is kind of in the middle, and in the middle of what, right? In the
middle of what non-Microsoft people think of it from the outside
looking in, which is this glowing thing on the hill. Like you
thought, "There you go. There's the badge. Now where's the money?"
Right? But the other endpoint is the one that I had, which is as an
insider. When I worked at Microsoft, I loved the MVPs, and I was
frankly even intimidated by them. But having that insider knowledge
of the program, it was like too familiar to me. And so, I might not
have valued it as much as it should have been. And I eventually
ended up bowing out of the program because I don't want to do the
paperwork anymore, to be perfectly honest.
Lars Schreiber (00:09:45):
That's the reason why you're not an MVP anymore?
Thomas LaRock (00:09:47):
I know it's sad.
Rob Collie (00:09:48):
I know it's so sad, right? And then there were a number of years in
there in between where I was like, "Ah, I should really get that
back," but I wasn't doing enough in the community to really make a
case for myself during those years. Now with the podcast and
everything, we probably are doing enough now, but it's got to
bubble to the top. So how did you and without giving away any
secrets, obviously, what was that process like? Okay, so you have
the MVP, you hang that on your website, and you just sit back and
go, "Okay, here they come." And then they don't. What do you do
next? How did you recover from that?
Lars Schreiber (00:10:20):
Honestly, I'd never do something special for becoming an MVP. I
blogged. I started podcasting a lot later. I'm creating videos now.
And I started doing it just for fun, because I didn't know anyone
who did Power BI back then. I had one friend that I could infect
somehow with this virus. Oh no, we shouldn't use that word
anymore.
Rob Collie (00:10:41):
It's okay. It's okay. We can talk about the R zero, the R not value
of Power BI, right?
Lars Schreiber (00:10:47):
No, I meant the word, "Virus."
Rob Collie (00:10:49):
I know, but still it's an appropriate metaphor. We talk about Kevin
Overstreet still proudly calls himself, "Patient zero for Power
BI," at Eli Lilly, which is a pharmaceutical company. Right? So
we're cool. "Virus" is not a canceled word. We're still okay with
it. I think.
Lars Schreiber (00:11:06):
Good. Good to know.
Rob Collie (00:11:09):
That's how my blog started too. There was a little bit of
professional interest in it. In the beginning, I needed a resume
outside of Microsoft. That isn't what powered me to write two post
a week for five years. It was excitement, it was enthusiasm. People
were always asking me like, "You're pouring all this effort into
this blog. Do you know how it's going to ever pay off for you?" No.
I got a vague sense that it probably will, but I don't know how.
That's fine.
Lars Schreiber (00:11:33):
And look how many people do you have influenced with what you did?
I know Matt Ellington is someone, I know he was influenced by you.
You already had [inaudible 00:11:43] on your podcast. She was
influenced by you. I read your blog posts back then in 2012, '13, I
guess, and I started thinking, "Why isn't something like this there
in German language?" And this is how I started my blog. So you
influence a lot of people by doing that, I guess, without having
the idea of doing that.
Rob Collie (00:12:05):
It did feel like a religion of sorts, right? But a good one, like a
really good fundamentals, like everyone should know about this. And
it is really cool to know that I did help ignite so many amazing
people. A lot of people who are working in Power BI today have no
idea, they don't even know who I am. It doesn't matter. And that's
good news. You want something to be so popular that what happened
10 years ago? "What? That was something going on 10 years ago? Most
people heard of Power BI five minutes ago. And you look at it, look
at the number of people who are producing amazing content and
essentially, discovering and inventing new cool techniques all the
time. This virus has really gone.
Thomas LaRock (00:12:48):
Pandemic.
Rob Collie (00:12:48):
Has gone pandemic. I don't know if we can say, "Pandemic."
"Pandemic" is probably a canceled word.
Lars Schreiber (00:12:53):
Yeah, but I get your point.
Thomas LaRock (00:12:55):
How else are you going to explain a virus that has left
containment?
Rob Collie (00:12:59):
Seems like, "Pandemic" might have a negative connotation. I know,
"Virus" does too, but, "Pandemic" is a bridge too far.
Thomas LaRock (00:13:04):
Cancer has a pretty negative connotation, but we haven't canceled
that word. I think pandemic will be safe.
Rob Collie (00:13:09):
Hey Luke, can you edit that out? Tom just said a bunch of really
dreadful stuff.
Thomas LaRock (00:13:14):
This is why I'm never on the podcast. Everything gets just cut.
Rob Collie (00:13:19):
You know those episodes where Luke's voice at the beginning says,
"With your host Rob Collie," and he doesn't say, "With your
co-host," well, Tom was there. We just edited the hell out of him.
No, he wasn't. We don't do that. So you didn't deliberately become
an MVP, but after you went solo, after you decided to turn
professional, to turn independent, you mentioned that at the
beginning, it didn't just happen. There weren't just clients. What
did you do then? How did you react to that? Because you're okay
now.
Lars Schreiber (00:13:52):
Yeah. I didn't change anything. It was just a matter of time. It's
necessary that people get to know you, and if it's a bigger
company, of course, different people from within the company are
asking for help. So, that's actually the case when I give
trainings. I'm a trainer and a developer for Power BI, and when I
give trainings for bigger companies afterwards, 50% of the
attendees ask questions that makes you more projects and more
money, of course. And that takes time. So I just decided to give me
that time. And the good thing was, my wife is working in IT, so she
learned something really important, not like me. And she said, "Do
it," and if you don't earn so much money in the meantime, we can
take mine." That was a big help, of course.
Rob Collie (00:14:36):
It's good to have the safety net.
Lars Schreiber (00:14:38):
Yep.
Rob Collie (00:14:38):
It's still a big leap. So you didn't have the portfolio of clients
waiting in the wings when you went independent.
Lars Schreiber (00:14:47):
Yep.
Rob Collie (00:14:48):
That just magnifies the size of the leap. So the amount of courage
and tenacity, you have to really stick to it in a situation like
that. So well done.
Lars Schreiber (00:14:57):
And I guess we shouldn't ignore one fact. In Germany, Power BI
wasn't a thing a couple of years ago.
Rob Collie (00:15:06):
Oh.
Lars Schreiber (00:15:06):
I guess when Power BI desktop or Power BI Designer Preview started
in the beginning of 2015, nobody in Germany knew what Power Pivot
was and even Power BI started a couple of years ago, maybe 2018,
maybe 2019. There were so many people who didn't know about it. And
I have no data, no good idea why it's like that. But I guess, it's
because in Germany, people don't like to read English stuff. And
there still wasn't so much good stuff in German. Now with Office or
Microsoft 365, many people just see, "Oh, there's Power BI, and it
is like Excel, so we all should be able to use it." We all know
it's not like that, but that's the marketing behind it.
Rob Collie (00:15:51):
One of the reasons, probably for the delay is that the German
translation of the applications just takes longer. They need more
space. The words are longer. This is an old joke; I was around at
Microsoft when they went to, they called it Worldwide EXE-
Lars Schreiber (00:16:07):
EXE?
Rob Collie (00:16:07):
E-X-E, like the executable files?
Lars Schreiber (00:16:10):
Ah, yeah.
Rob Collie (00:16:11):
Worldwide EXE, meaning in Excel.exe, going forward the worldwide
EXE campaign that swept through Microsoft office, the engineering
teams probably I think when the Office 2000 cycle. So basically the
idea was is that if you took Excel EXE under a microscope and
looked at it, you wouldn't find any English sentences in it
anywhere. Basically all of the strings that populate the UI, the
labels for buttons, the help, all of that would be moved out of
Excel ECE into separate DLLs. Separate delayed load library files
per language.
Rob Collie (00:16:54):
So Worldwide EXE meant, for the first time ever the Office apps
were going to be written in a way that was completely language
neutral. There'd be nothing in the exit files themselves, the
actual application files themselves, that was language specific.
And it would hydrate itself from its com companion DLL, for
whatever language you told it to be. And up until then, it wasn't
like that. There was a German version of Excel.exe, and an English
version of Excel.exe, and they decided to go this more modular way,
right? And because this was a big effort at the time, I spent a lot
of time with localization engineers, whose job is to man manage
this process, but also get all of the strings rewritten in every
different language, all bajillion languages that Office gets
delivered in.
Rob Collie (00:17:42):
And there was a joke going around and I think you can still find it
on the internet today, it wasn't a Microsoft joke. What we would do
by default as program managers, we would say, "Okay, well here's
the English sentence or the English label that is required for this
error string," or there's an error message that pops up. We provide
this, however many characters, let's say it's a 200 character
English phrase. Okay? And then the developers would say, "All
right, well, we probably need to pad that a little bit; maybe 256
characters, because the different languages will be different
lengths, right?" But then German would come along and require 400
characters, so I don't think this'll be nearly as funny to a German
as it is to an American that worked in software. I think the joke
is something like the list of translations for English words. So
English word, "Dog," German translation for dog is, "Scratching,
sniffing, wolfing woofer," or something. It's a made up word,
right?
Lars Schreiber (00:18:37):
Yeah. Okay.
Rob Collie (00:18:37):
But it conveys the fact that the German nouns basically contain all
kinds of verb of what this thing does. It's a very long
description. It's very long-winded description of what this thing
does. "Panting, scratching, sniffing," whatever. Okay? And then,
"Dog catcher," the word for that in German is, "Panting,
scratching, sniffing, catcher."
Lars Schreiber (00:18:57):
Because we're combining all the words. Yeah.
Rob Collie (00:18:59):
Yeah. Right. The joke was you've really got to account for those
Germans. Over, and over, and over again, we'd make this mistake;
this string allocation is too small. You need to bump it up for
German. But then we also found out that Swedish was even worse.
So.
Lars Schreiber (00:19:12):
And try Finish.
Rob Collie (00:19:13):
We never did finish. All right. Well Luke, you can decide whether
that joke is worth leaving.
Thomas LaRock (00:19:23):
I want to hear the data origin story. I don't think we have that
yet.
Rob Collie (00:19:27):
No, we don't. We don't business background. Right?
Lars Schreiber (00:19:30):
Mm-hmm (affirmative).
Rob Collie (00:19:31):
What's your first collision with data? What was the first time you
encountered Excel or similar, and started getting that itch that
this might be something that appeals to you?
Lars Schreiber (00:19:40):
I started working as a business control and I always saying I was a
financial controller, but I guess that's not correct. I was a
business controller doing budgets, and forecasts and stuff.
Rob Collie (00:19:51):
You can call it whatever you want. People over here in particular,
right? But they might call it financial. We've had this on the
podcast before. If you work outside of finance, you call it,
"Finance," but if you work in finance, you call it, "Finance."
Lars Schreiber (00:20:04):
I will try to remember that. Thanks.
Thomas LaRock (00:20:06):
Nobody says that, I don't know what he's talking about.
Rob Collie (00:20:10):
We still call it, "Finance." That's fine. Business controller,
financial controller. I can't tell the difference.
Lars Schreiber (00:20:14):
Me neither. That's why I left that job. No, just kidding. I just
liked Excel. I liked helping people because it took just a couple
of months and people noticed that I know that software, and they
came to me and asked me questions and I could help them. And I
liked to help people, especially with data. And when you see people
sitting in front of Excel and trying to build a sum using their
calculator, they're having in their left hand. There's something
you've got to do
Rob Collie (00:20:41):
Call to action.
Lars Schreiber (00:20:43):
Yeah, exactly. And I replaced a guy who created a lot of code to
extract data from a mainframe, all these big pixels on a black
screen, and trying to read out the data. Something you would do
with a Power Query with a couple of clicks now. But back then, it
wasn't possible. And we read all the data from the screen into an
Excel file, to being able to analyze the data, and to reload it
into other systems. And I somehow was the guy who had to maintain
that file and that code without knowing anything about VBA. And I
just learned it because I was interested in and that's how all that
started.
Rob Collie (00:21:24):
Your description of the person you're looking or shoulder and
they're using Excel, and they've got a calculator on the other
hand, you know, when a dog or a child falls into a river, and
someone jumps in and saves them, and then they're interviewing the
hero on the news afterward, and the hero is just saying, "Ah, I
just did what anybody would do in my position. I had to do
something," right? That's sort of the same feeling you get when you
see that happening in Excel, right? It's like, "Well, I just have
to," you got to jump in and save the baby. We should have similar
shows where we entered. "Yeah, boy, that was a close one. They were
almost going to waste the rest of their life adding that up with
the calculator, but I rescued them."
Lars Schreiber (00:22:07):
The thing is usually people are sitting in an office for 20, 30
years, and nobody tells them how to use the software they put on
their PCs. Yeah, "Here's Excel. Here's Word. Here's PowerPoint."
I'm loving the Power tools we are having right now, but I never
knew how mighty or how powerful PowerPoint is. I never used it the
way you could do. And I have some friends who are deep into the
PowerPoint thing, and there's so much to learn. But if nobody shows
you what's possible, you will never learn. Especially if you have
no idea what you could look up, what you could solve with this
tool. This is why I always, when I give a training, especially in
Power Query and I have a basic training for as an introduction, at
the end, I always give a demo for what's possible too when you
learn a bit more about M or learn a bit more about how to use the
UI, because people need to know what's possible to invest the time,
and energy, and maybe money to get to that point. If they don't
know that they will never try to get there. I'm
Rob Collie (00:23:17):
So glad you do that.
Thomas LaRock (00:23:18):
I like that.
Rob Collie (00:23:19):
I'm partly glad you do that. Because it validates something that
I've been doing for a number of years. I don't really teach a whole
lot of classes anymore. I certainly got a lot of reps. I've taught
a lot of classes.
Lars Schreiber (00:23:28):
I can imagine.
Rob Collie (00:23:30):
Over time, what I started to do, in a typical two day training is I
started teaching a less and less ambitious curriculum over time.
When I first started out, I was trying to dump everything I know,
everything I'm capable of doing, I was trying to make everybody
capable of doing that in two days. Really not realistic. I couldn't
have been a student in my class and learned it in two days. It's
just no way, right? I had learned it myself over the course of
years at that point. So I started to do the same sort of thing; so
I started to not take people nearly as deep into DAX. I'd move more
slowly and I'd take them less distance total. However, I would do
something very similar because I felt, even though this was the
right thing to do for people, I was also short changing them, I was
doing them a disservice.
Rob Collie (00:24:17):
So then I would take a break and say, "Look, okay, everyone, you
can just take your hands off your keyboards. You don't need to try
to remember any of this. You don't need to learn how to do this;
I'm not even going to try to teach you. I want to show you just
some amazing capabilities that you know to look for." Oftentimes, I
give them a Google search so they could find the blog post that
explained it in detail when they needed it, they could write down
the Google search, things like that. But yeah, like the art of the
possible; you've got to stretch people. There've been so confined
in this little jail by the tools that they've had available to them
and that they've known how to use, right? It's natural, when you
take the walls down off of that jail, they just stand right there
on that square of concrete. And so it's like, "Hey, let's go run
around the yard. It's a big, big, big world." And I found that,
just showing them what was possible would people excited, and as
long as they knew it was there, they could go find it later. Is
that similar to what you do?
Lars Schreiber (00:25:12):
Yeah, exactly. I guess because we have this podcast episode today,
I was thinking about the past 10 years, because I started in 2011
with your blog posts and your DAX stuff, and that made me think of
the last 10 years a bit. And when I was at my first employer, that
was a supermarket company, I would've killed for what's possible
with Power BI desktop today; just putting the longitude and
latitude on a map and having all the stores on a map. I tried to
solve that back then with a VBA script and Google Earth, maybe you
remember that desktop top application that you could download. And
I didn't write that code myself, but I found it somewhere, and
someone created a code that in combination with Google Earth made
it possible to put little icons on a map where all those 300 stores
were and see it in Google Maps and a browser, that was so freaking
amazing. But if you see how easy it is, if you have longitude and
latitude, and probably a desktop, you can't believe it. But if you
have never seen that, this is possible with that tool, you will
never try to learn it. And this is why you need those ideas, those
demos.
Rob Collie (00:26:25):
Ambition expanding. Strengthen the ambition muscles, not just the
capability muscles.
Lars Schreiber (00:26:32):
And I think that's my personal experience. Not just with me, but
also many others; many people learn new stuff in their spare time,
not during the working time, because their employer doesn't give
them the time to do it ever, or they don't think it's time to
invest. So you have to tell those people why they should use their
spare time, which they usually should use for family or hobby to
invest it in something new. And they need to know why it is the
right thing to do.
Rob Collie (00:27:04):
Have you ever been hired for training twice, by essentially the
same people, to teach them the same thing again?
Lars Schreiber (00:27:14):
The same people are the same company?
Rob Collie (00:27:16):
Same people, same topic.
Lars Schreiber (00:27:19):
No.
Rob Collie (00:27:20):
What I'm really getting at is, I have. That happened to me once,
just once. And it was because the first time, even though they,
"Learned it," they didn't go back to their jobs and use it. You
said the spare time thing, right?
Lars Schreiber (00:27:36):
Yeah.
Rob Collie (00:27:37):
Well, when people go to training, it's sort of negative spare time.
Their regular is still continuing to pile up demands during those
days that they're in training. And so, it creates this debt. The
first instinct is, "Let me go clear out that debt and then I'll go
apply the new stuff." Slippery slope. One year later, you Rob
Collie's back teaching you the same thing again. That spare time
thing, it's the truth, right? The full time job, it looms large
over learning new things. It's the enemy, the opponent of learning
new things. I'll say it this way: you've never been called back to
teach the same thing to the same people yet. It still might
happen.
Lars Schreiber (00:28:23):
Yeah, it could be, but it wasn't the case so far. But because we
were talking about this spare time thing and learning thing, when I
saw Power Pivot the first time, I simply didn't understand what it
was. I didn't understand its analytical database. I thought it's a
bigger Pivot Table. But when I started learning what it really was,
I noticed that's different than learning, how to use the V look up
in Excel. And I also noticed when I don't get projects at my
current employer, and I didn't get any, I need more time to learn
it. So I decided to change my contract with my employer, so that I
didn't work on Fridays, so I could read those books and learn this
stuff.
Rob Collie (00:29:05):
Wow. You were the only person in Germany reading English books.
Lars Schreiber (00:29:10):
No, for sure not. But back then, there were only English books.
That's the case.
Rob Collie (00:29:15):
That's amazing. That's a soundbite. It was so compelling, even
though you didn't understand it all yet, it was so compelling that
you needed to restructure your employment agreement to give you 20%
of your work week back to learn it. That's deliberate. That is
deliberate AF, as the kids might say.
Lars Schreiber (00:29:38):
Yeah, the thing is, I wasn't happy with being in control because I
like the data stuff. And I was thinking about going back to
university and studying IT, but I wanted to do this BI thing and I
couldn't study business intelligence whatsoever. Then I just talked
to my wife and she said, "Okay, then do it on your own." But of
course, I needed the time for that. So then I decided to read the
right books and do it on my own.
Rob Collie (00:30:03):
When you say it so casually like that, it's one thing, but as an
outside observer, that's just a very conscious process. I don't
think I've ever had anything in my life, my career, play out that
consciously. It's just this bounce, bounce, bounce down the
mountain, a little bit of control, but not that much. I didn't even
know when I started blogging about Power Pivot that it was going to
turn into my next career. I discovered that.
Lars Schreiber (00:30:31):
But what you can't ignore is that I had you as an example. I could
see it works somewhere. It works in the US, it for Rob Collie and
for Matt [inaudible 00:30:44], and it will work in Germany and the
other parts of the world as well. And of course, I couldn't know
how many clients I could get in a certain amount of time, and I
also didn't know if I can make a living of it, but I knew I wanted
to do this thing maybe as an employee, maybe as a freelancer, but I
wanted to do that stuff and therefore I needed to learn it. And
that takes time if you have no clue about Star Schema, and what's a
proper ETL process, I did know all these things, but I was willing
to learn it.
Rob Collie (00:31:19):
Even though I worked on the product, Power Pivot, your experience
learning it, it was very similar to mine. I mean, we were coming at
it from the same backgrounds; the Excel and VBA. I'm sure you can
run circles around me in VBA. I know that you haven't been doing a
lot of it lately, neither of I, but peak Lars against peak Rob and
VBA, I'm definitely going to take you in that contest. We still
came from the same place. I'd heard about Star and Snowflake Schema
forever. It was one of those things that I just let wash over me. I
didn't didn't really know what the difference, didn't care. It
sounded cool, but I understand it. Now Power Pivot was the reason
to learn what those phrases meant.
Lars Schreiber (00:32:01):
There's a question I wanted to ask you forever; do you have a clue
who was responsible for giving the Verde pack engine, that name,
Power Pivot?
Rob Collie (00:32:12):
I do actually, or at least I know the person who is heading up the
effort, her name's going to escape me at the moment. I have to go
through some really, really old emails. Basically, the product
manager, marketing coordinator for Project Gemini was working with
the naming firm. Microsoft will bring in a branding firm that helps
them name things when they're actually going to name something;
they almost never name something. Especially in the old days, they
would give a really boring name like Microsoft SQL Servers,
Analysis Services, it was the German approach, right? It was the,
"Pant, scratch, sniff," right? But every now and then they'd want
to invest to actually name something. They'd bring in some pros.
And sometimes those pros, the, "Professionals" were awful. I think
I'd already moved to Cleveland at the time when they were naming
it, but I was still a Microsoft employee, and I know the product
manager for Project Gemini, she led the process. I ended up
collaborating with her a couple more of her roles that she had at
Microsoft afterwards, which is why I really regret that I don't
remember her name at the moment, but she came back with Power Pivot
from working with the naming firm.
Lars Schreiber (00:33:23):
The reason why I'm asking the following; when I was a VBA program,
a VBA developer, I had a problem and I talked to the IT, "Hey, can
you help me with VBA?" They looked at me and said, "Hey, VBA is not
really a programming language. We don't know anything about VBA."
And then, I talked to my other controlling colleagues and asked for
help. And they said, "Hey, I'm not an IT person. I'm not a
developer. I know nothing about VBA." So you were this island, you
were almost lonely with your topic.
Lars Schreiber (00:33:53):
And coming back to this Power Pivot thing, I think the name is
really confusing, because suggesting this thing has something to do
with the Pivot Table, and it does not really need to have
connection to it. But on the other hand, it's really smart because
imagine you go to a person and say, "I know you're a business
person, but please help me doing a business intelligence data
model, and help me creating such a report." Nobody would say, "No
problem. I'll start doing it." But you put this little icon into
Excel, simply completely different engine opens up, but it seems
like being in Excel. And it ha a name that sounds familiar with the
word. And suddenly people try to do things they would have never
tried in a different application. So, Power Pivot was a really good
name for making people using it. But for me, it took a while to
understand it's not the Pivot Table that is on Power.
Rob Collie (00:34:59):
When you're using Pivot Tables, like normal Pivot Tables, back in
the day, regular Pivot Tables. You're not thinking at all about the
engine behind the scenes that's crunching the numbers to produce
the Pivot Table. In fact, it was kind of a jaw dropping surprise to
me when I was on the team and someone explained to me that the
Pivot Table calculation engine had nothing in common with the
calculation engine for the sheet, like the normal Excel grid. They
don't share a single line of code; they've got nothing in
common.
Lars Schreiber (00:35:30):
Interesting.
Rob Collie (00:35:31):
It's a completely different beast. It's like this whole calculation
engine behind the scenes. And the really technical people at
Microsoft would refer to that, sometimes they'd call it the Pivot
Cache. Now I know the Pivot Cache is also... Oh, it's been a long
time since I've used that word. Do you remember that word? "Pivot
Cache?"
Lars Schreiber (00:35:46):
Sure.
Rob Collie (00:35:47):
Man. That's a blast from the past. This was the thing that made
your file more than twice the size if you duplicated the data,
right? So most people thought of the Pivot Cache is just the extra
copy of the data, but that Pivot engine, that's the thing that got
swapped out by Power Pivot. Take this really clunky one table at a
time, no relationships, no advanced formula language at all. Just
swap that out behind the scenes and plug in this other superpowered
one, give it the same interface. Really, the only way that Power
Pivot could surface its data in Excel, most of the time anyway, was
through a Pivot Table or a Pivot Chart. There were cube formulas as
well, obviously, but naming is so hard. Even Pivot Table is a
disaster. It was a terrible name.
Lars Schreiber (00:36:33):
Why?
Rob Collie (00:36:33):
Well, what is Pivot?
Thomas LaRock (00:36:36):
Yeah. You got to make me look it up. So SQL Server has Pivot and
Unpivot and I thought that was some antsy standard.
Rob Collie (00:36:42):
I don't know, but the average Excel person in 1998, who is supposed
to be adopting Pivot Tables, right? Has no what the SQL syntax of
Pivot and Unpivot. I tried to rename it Summary Tables. It speaks
to the novice better. Because I'm not a sequel person, I still to
this day do not equate Pivot with aggregate. They're not the same
noun at all. They're the same verb. They have nothing in
common.
Lars Schreiber (00:37:08):
I'm afraid I'm missing the correct words for it now, but a Pivot is
a point where you can circle things around. I guess it's a French
word. I'm a quite sure it's Pivo?
Rob Collie (00:37:19):
Oh, Power Pivo. A Pivo Table.
Lars Schreiber (00:37:22):
I might be wrong, but it's about this circling things around the
single point. And I guess that's the main criteria of a Pivot
Table; slicing, and dicing, and making columns rows, and row
columns.
Rob Collie (00:37:34):
Yeah. Swapping rows and columns on a Pivot Table is the only thing
that really remind me of pivoting. It's like a twist or a
transpose, right? Pivot and pace transpose seem to be similar
concepts to me. I just want to swap rows and columns, but the
number one reason to create a Pivot Table isn't to swap rows and
columns; it's to see what the data's telling you. It's to roll up
40,000 rows into grouped subtotals. To me, Tom, even in the SQL
world, it has more in common with group buy.
Thomas LaRock (00:38:04):
So, no.
Rob Collie (00:38:06):
Oh shit. Now's where I actually have to learn something.
Thomas LaRock (00:38:09):
So I'm sorry, you said more in common. And so I don't want to say
no to that. To me, there's two different things. So Pivot is not an
antsy standard, but I want to say group buy roll up in cube are. I
think those are common aggregations, and part of [inaudible
00:38:25] standard as to how the engine should return the results.
The biggest thing is how data gets stored, and then you need to
transpose the rows into columns, and you want to perform an
aggregate on the column now, which used to be rows. So there was a
need for that, obviously, but that's not an antsy standard. It's
definitely in TSQL. So Microsoft saw that this was a need and
started building that out, but now just the concept of this for me,
that first of all, that it wasn't in Excel yet, or if it wasn't
Excel, it was all done by hand and macros and things like that. And
clearly, somebody said, "Wow, they're spending a lot of time
building these macros when we could just make magic happen for
them."
Thomas LaRock (00:39:06):
But I've written before, and I've talked to you about the idea that
you could just put your data in Excel and perform these operations
in the matter for a few clicks, as opposed to needing to know a
language, is just so powerful. If you said to make it a Summary
Table, I'd be like, I don't want to do some, I want to do an
average. I want to do all these other things.
Rob Collie (00:39:25):
Summary doesn't mean, "Sum." It just means-
Thomas LaRock (00:39:30):
But I think it's also a bad name.
Lars Schreiber (00:39:31):
Let's call it, "Aggregation Table."
Thomas LaRock (00:39:33):
Aggregation Table comes close, but I hate the way it sounds.
Rob Collie (00:39:36):
Yeah, it's a little, it's a little too nerdy. I'm going to have to
say that, nah, you're wrong, summary's better. Thought about this
for a long time. But here's the thing; it doesn't have to be
perfect to be infinitely better than Pivot Table.
Thomas LaRock (00:39:49):
See, I thought Pivot Table was fine. I don't have a problem with
Pivot Table.
Rob Collie (00:39:52):
We take it for granted. So, you said it was kind of mind blowing
that Pivot Table. Weren't always in there. You know what? They're
still not in there. Right? For the majority of Excel users, they're
not in there.
Thomas LaRock (00:40:05):
Oh, yeah, I know.
Rob Collie (00:40:07):
They're not using it. I mean, you wouldn't believe, sitting next to
an electrician on an airplane flight one time, he's describing to
me this Excel process that he has to hand off to some cell
consultant can pay him a lot of money and all this kind of stuff,
right? And I'm sitting there listening to him and going, "This
sounds like a Pivot Table." And I just, so I took my laptop made
some data for him, like, "Okay, many feet of conduit blah, blah,
blah, blah, blah, blah, blah." And then I create a Pivot Table, and
I went [inaudible 00:40:34], and he's like, "What?" And then, so
I'm talking to him about it, 20 minutes goes by and he leans over
to me and says, "Hey, can you show me that again?" He's like, "I
really want to remember it this time." Like Lars was saying the
beauty of the Power Pivot name was that it was building off of
something that people already knew. Now at one point, one of my
jobs at Microsoft was to try to get more people to know Pivot
Tables. When I worked on the Excel team, that was one of the things
that I was overseeing anyway. I've wrestled with this problem a
lot. And I kept coming back to, "Oh God, I wish we could change the
name." And we couldn't, so. Pivot, didn't speak to the Excel
audience.
Thomas LaRock (00:41:12):
Right, I get that.
Rob Collie (00:41:14):
From zero. But once they're on side with Pivot Table, you leveraged
it with Power Pivot.
Thomas LaRock (00:41:20):
Would it make you feel better to know that the concept of Pivot and
Unpivot is also in Python code?
Rob Collie (00:41:25):
I'm not surprised.
Thomas LaRock (00:41:27):
But, clearly there's a need for the concept of rows to columns.
Because that's what we're talking about, right?
Rob Collie (00:41:33):
Yeah. But those processes don't necessarily mandate
aggregation.
Thomas LaRock (00:41:38):
No.
Rob Collie (00:41:38):
And the Pivot Table is like Lars said, the aggregate table, even
though it doesn't roll off the tongue in English, right? That's
what it does. It's not to take your 400 rows of data and turn it
into a different 400 rows of data or in the a case of Unpivot, an
Unpivot operation actually you end up with more cells. You end up
with a larger square footage of data than you had beforehand.
Whereas a Pivot Table is a cruncher. Pivot Table's always going to
be smaller than the input. That was the mission, anyway. Naming,
I'm obsessed with naming. I don't know if you can tell. Let's name
some things. What have we got that we wish we could rename?
Thomas LaRock (00:42:20):
Perview.
Rob Collie (00:42:20):
We've talked about it on this show, right?
Thomas LaRock (00:42:22):
It's a fancy data catalog.
Rob Collie (00:42:24):
That's right. We were talking about this with John Hancock, where
his software that catches underage predators.
Thomas LaRock (00:42:29):
Perview.
Rob Collie (00:42:32):
Exactly.
Thomas LaRock (00:42:32):
It's Perview. See, I understand that purview is an actual English
word and what it means and all that. I just don't think in terms of
a service, it was likely the best choice. It is descriptive, but I
don't know.
Rob Collie (00:42:49):
In the spirit of international relations, we should probably
explain to Lars why this is funny.
Lars Schreiber (00:42:53):
That would be nice. Yes. Thanks.
Rob Collie (00:42:57):
All right. So the English word, "Purview," what is the definition
of that word?
Thomas LaRock (00:43:01):
Well, I'll just go look up the definition. It's so hard. It's hard
for me to explain.
Rob Collie (00:43:05):
See that Lars, even we don't know English,
Thomas LaRock (00:43:08):
"The purpose or scope range or limit."
Rob Collie (00:43:11):
But there's also the English word, "Pervert." So John Hancock's
company, they fight crime. And the free version for law enforcement
is catching child predators. And for good reason, they just can't
charge for that. That's free. And so, I think on that podcast, we
stumbled into the joke, we're talking about Azure Purview, and then
we realized that, "Hey, that'd be a name for John's software is
perv view." Lars looks so pained. I wish we had a screenshot of the
agony.
Thomas LaRock (00:43:45):
You ready for this? A story I heard years ago. Hyper-V, but when
they first wrote Hyperv somebody looked at and, "Hyperv? What is
this Hyperv thing you have?" And the guys at Microsoft, like, "Hi
Perv? How do you get Hi Perv?" And they're like, "No, seriously.
That's what that says." And they're like, "No, it's Hyper-V." So
they put the dash into Hyper-V.
Rob Collie (00:44:10):
Yeah. See, there you go. Naming, right? That's the naming
department. That'd be the naming person's purview. Putting the dash
in Hyper-V.
Thomas LaRock (00:44:16):
Make sure somebody says the word, "Hyper" and not, "Hi perv,"
right?
Rob Collie (00:44:20):
That's right, yeah.
Lars Schreiber (00:44:21):
Especially when you're an international worldwide company. Many
different cultures will interpret those things differently.
Thomas LaRock (00:44:29):
Well, that happens all the time. Yeah.
Lars Schreiber (00:44:30):
I can remember that.I talked to Michael Russo and they're doing
this, how do they say hello to all the people in their videos? They
they're using an Italian hello ciao or something. I'm pretty sure
it's something different. And he said, "There are many people who
don't understand that, don't understand it's hello in Italian." And
so you really have to do it waterproof or bulletproof what you're
doing there, and still people will misunderstand what you're
saying. And I guess naming is a very hard thing to do.
Rob Collie (00:44:59):
Yep, especially those Germans in there 80 character words as
we've
Thomas LaRock (00:45:05):
Discussed, that's one of the jokes, like, "Is there a long German
word for," and then say something, and then there's almost always
an answer. Usually, it's just something weird. Like, "What's the
long German word for when you are alone in stress eating all by
yourself," just something weird. And you're like, "Oh yeah, there's
a word for that." It's like a feeling emotion or whatever. It's
just, there's all these little things that you can always ask for
whatever scenario, "Is there a long German word," on Twitter and
then somebody will give you an answer.
Lars Schreiber (00:45:34):
Sometimes I have to laugh about it myself because when you read it,
it's funny. But I read once German is a difficult language. They
have one word for driving somebody over with a car, hurting someone
with a car is [foreign language 00:45:49] and [foreign language
00:45:49]. If I say [foreign language 00:45:52], then I drive
somebody over. And if I say, [foreign language 00:45:56], so I
pronounce the A more than everything else, then I circle around
him.
Rob Collie (00:46:00):
Oh, okay.
Lars Schreiber (00:46:03):
So it's the same word for two absolutely different things.
Rob Collie (00:46:07):
Uh-huh (affirmative). Yeah.
Lars Schreiber (00:46:08):
But I had to read it on Twitter to recognize it myself.
Thomas LaRock (00:46:12):
It's the same word. So it's spelled the same?
Lars Schreiber (00:46:15):
It's spelled exactly the same way. Yeah.
Thomas LaRock (00:46:16):
So, you can only hear the difference? So for example, if this was
in a police report, This one word, "Did he drive over him or around
him? It's not clear to me."
Lars Schreiber (00:46:26):
You can get it from the context when you read it. It's better to
hear somebody saying it, yes.
Rob Collie (00:46:32):
There a command version of this word? Meaning like you tell someone
to drive around them, or you tell someone to drive over them?
Because this could get really, really dicey, like a written
order.
Thomas LaRock (00:46:42):
And you're what? In the mafia?
Rob Collie (00:46:47):
Yeah. Well, I mean, lives could hang on the balance here.
Thomas LaRock (00:46:50):
So we started talking about Iceland, and now we're talking about
long German words.
Rob Collie (00:46:54):
You're right, we should bring it back the real stuff. So in terms
of the work that you do, what can you tell us about it? Is it lots
of clients? Is it a handful, or one that you spend most of your
time with? Like Inca, do you find yourself staying close to home in
terms of what you originally were working on, even though you
didn't want to be a controller, are you mostly working with
controllers now? Or are you kind of all over the place in all kinds
of different data fields? Just give us some of the texture.
Lars Schreiber (00:47:18):
Of the texture. I guess my work is pretty close to what Inca does,
but I have more clients than her. By the way, she's a good friend
of mine. This is why I know that I have, I don't know, I guess 30
or 40 customers. And I remember you have written blog post about
how Power BI, Power Pivot will change the VI consulting, having
many clients, but just for a short period of time, just showing
them what's possible, and then maybe a bit of coaching, and then
they're doing it on their own. This is exactly what I'm trying to
do. My ideal customer, my typical customer comes from a technical,
non-IT department, so from marketing, or controlling or finance.
Finance?
Rob Collie (00:48:03):
Finance. But when you say that you have to take your pinky finger
off of whatever you're drinking.
Lars Schreiber (00:48:07):
Yeah, next time.
Rob Collie (00:48:09):
Finance.
Lars Schreiber (00:48:14):
And I'm trying to make their life easier. So I do three things; I
develop stuff and they're not interested in how I did it. That's
not the usual case. The other side is, I give trainings and show
them how they can do this stuff on own. That typically doesn't work
right from the beginning. Because when I give my basic trainings, I
always say, "I have a lot of experience in martial arts, but 10
years ago," and I say, "This is my one day women's self defense
course. After that, you won't be able to defend yourself properly,
but you know what you're dealing with." And so people know what's
ETL, what's a data model, what's a measure, what's a report, and
what comes after, because afterwards I show them, usually people
think Power BI is Power BI Desktop. Yeah? And they have heard
there's a free version of Power BI, which is called Power BI
Desktop. So you have to show them that there is another world
behind this published to web, or publish button. And so those
people usually come to me afterwards for consulting. And this is
what I do; I build solutions, but show them always how I do it. I
want to transfer my knowledge, because even if it doesn't sound
logical for a consultant, I don't want to see my clients again, for
the same reason.
Rob Collie (00:49:30):
Mm-hmm (affirmative). I completely understand that. The traditional
consulting move in certain situations is to hoard the knowledge
from your client, so that they're dependent on you. And they have
to come back to you for over, and over and over again. Well, even
if you set aside for a moment, the ethics of that business model,
or whether it feels fair, set that aside for a moment; it's boring
to be the one that has to do that same thing over, and over, and
over again. Why not show them how they can write the different
version of the calculate measure, right? If all they need to do is
just copy paste the formula and change the base measure, the year
to date, version of spend, you wrote that formula for them, but you
didn't write the year to date version of revenue. Well, come on,
they can do that, right?
Rob Collie (00:50:17):
And I know we've got calculation groups, and all kinds of whiz bang
things. Now we didn't have those back in the day. We hacked, we
hacked the Power Pivot files. We literally hacked them to duplicate
measures across families like that. So, I completely understand.
Have you ever found though that you run into clients that just
don't care? They don't want to learn? Even though you're dying to
transfer the knowledge, they're just sitting there going, "Oh, come
on. I love the report. I love it. Just keep more of this let's keep
going. Don't try to teach me. I don't want to learn. I don't want
to be you." Have you run into that at all?
Lars Schreiber (00:50:53):
That's a good question. I need to think about it a bit more, but I
don't think that I had people who didn't want to learn. I had one
scenario, one training I gave for a company and there was a person,
a woman who seemed to be good at Excel, that's what she said. And
in the end, I noticed she just knows how to use the SUM function
and nothing else. Which is okay, just tells you about how people
think about themselves.
Rob Collie (00:51:22):
That's right. Yeah. And also how they think about Excel. How deep
is it?
Lars Schreiber (00:51:26):
Yeah, exactly. And when I did one day of training in Power BI
Desktop, and everything was fine until we wrote the first DAX
measure. And then she said, "Oh, are there more formulas? Do I have
to remember all those formulas?" And I just did this some and the
SUM and a SUM X nothing else. "Oh, is there much more?" And in the
end, after two days, one day of training and one day of consulting,
I ask her, so what do you think about PA BI now? And she said
something that made me almost stumble out of the room, but then
made me think about it a bit deeper or a bit more. She said, "Power
BI is nice, but most of what you can do with it, I can do better in
PowerPoint." And I was, "Has she been in the same room with me
during the last two days?"
Lars Schreiber (00:52:13):
But then I noticed what they wanted to do was creating reports with
comments, what you usually do in a PowerPoint presentation. And it
took me a while to make her aware that what you do with Power BI is
one step before, or you create the comments, you find the
structures, the reasons why you have 1 million in turnover less
than last year. So the difference between having an analytical
report and a report you can print out as a PDF. And to notice that,
that changed the way I deal with people. When they say, "Hey, I
need this or that," I first of all, ask them, "Do you need an
analytical report? Or you need something you can print out
afterwards and put on the wall?" And I try to find that out right
at the beginning. And that was pretty interesting for me. Quite
lately, in what I do, I guess it was two years.
Rob Collie (00:53:05):
Ago. I'm going to make a guess that this particular training
client, that company is in the management consulting industry. It
sounds like a McKenzie, am I close.
Lars Schreiber (00:53:16):
No, they were in advertisement, but it could have been. Yeah,
sounds like that.
Rob Collie (00:53:20):
As you were saying, Power BI is the step before that. You're
finding out why, why that's like this. Right? I was just thinking
to myself, "Oh, this audience," the reason why I came to the
conclusion that this was a management, it was like, "But this
audience, the people you're talking to, they already know why. They
already know what they want to say." Your analysis might conflict
with that. They're not, "No, no, no. Don't let the facts get in the
way of a good narrative." Right? They already know what they want
to say. Now, they want to invent the data in the charts to back it
up. It's really hard in Power BI, because all it does is add up the
numbers that tell you the truth.
Rob Collie (00:54:00):
All right, advertising, you know what, for my purposes, I'm going
to round manage consultant to advertising agency. I'm going to call
myself correct. These two industries have that same thing in
common; they already know the narrative.
Lars Schreiber (00:54:14):
Yeah. Maybe.
Rob Collie (00:54:16):
Now they just need to package it.
Lars Schreiber (00:54:16):
It was a huge learning for me.
Thomas LaRock (00:54:18):
They're data riven as long as the data supports their preconceived
ideas.
Rob Collie (00:54:23):
Yeah. And once you've understood that way, why do you even need the
data? Just dispense with all that pointy headed geek work, and then
let's get down to the business of making money. That's so funny.
Well, we have a little bit of experience with this. Someone who
used to work with us, went to work for, let me be very vague, for a
very, very large consulting firm. General purpose, big, big, big
consulting firm. His reports back from the trenches sound like, "I
am really only using Power BI as a PowerPoint substitute." They
wanted a Power BI expert. They poached him from us, all's fair. And
now, he's the data driven PowerPoint guy.
Thomas LaRock (00:55:10):
Wow.
Lars Schreiber (00:55:11):
Power BI is PowerPoint for data.
Rob Collie (00:55:13):
Yeah. That's a good ambition. We've got a little ways to go yet, I
think. You mentioned PowerPoint being deep. How deep have you
gotten into PowerPoint? I had the same revelation somewhere around,
I don't know, 2014. "Oh my God, where have you been on my life,
PowerPoint. I didn't know all the things you could do for me." Are
you a big slide master? Do you put a lot of effort into slide
decks?
Lars Schreiber (00:55:36):
Not really, but I know how to use morphing. I like those effects a
lot. But the problem is how could you say how deep you're into a
tool, if you don't know the whole tool? That's why all people are
Excel masters, and they are not.
Rob Collie (00:55:50):
There's a joke that I always used to tell in my training classes,
which is: you ask someone how good they are at Excel and whatever
answer they give you, invert it to get the truth. If they say
they're good at it, they don't understand what they're signing up
for. They don't know how deep it is. Because someone who's actually
good at it is going to hedge. Lars, you're really good at Excel.
You might be a little rusty like me, but you're really good at
Excel. Certified good. But if I ask you on the street, you don't
know me, whatever, how good you are at Excel, you're not going to
say, "I'm in the 99th percentile."
Lars Schreiber (00:56:23):
For sure not, no.
Rob Collie (00:56:24):
You are not going to say that. That would be true, you are easily
in the 99th percentile in the world on Excel, easily. But you're
not going to say that. You're going to say, "Well, Hmm. Yeah. I can
do some things." Because you don't know what's coming next, right?
If you say you're good at it. And I say, "Oh man, those database
functions. Right? Huh? You know the database functions?" You'd be
like, "Um, maybe." If I say I'm really good at Excel and the very
next thing that comes out is array formulas, I know I'm in
trouble.
Lars Schreiber (00:56:57):
Yeah. Me too.
Rob Collie (00:56:59):
I'm like, "Control, shift, enter, right? That's yeah."
Lars Schreiber (00:57:02):
It's a good book.
Rob Collie (00:57:03):
Yeah. It's a really good book. And those are the formulas you can't
debug. It's like the DAX of Excel, but harder and less capable in
my opinion. Yeah, so someone who tells you they're really good at
Excel and then you find that they know the SUM function, well,
they're the one that graduated from using the calculator 12 months
before next to the Excel. And that's a huge change, right? It's an
infinite change in expertise. So how could you perceive it as
anything but, "I'm really good." Are you branching out into the
broader Power platform very much? Staying closer to home and
primarily Power BI, what do you find yourself getting into these
days? Have you tried Perview? I hear it's really good.
Lars Schreiber (00:57:47):
No, I haven't tried that. The thing is I work from home, also
before the pandemic started because I have three and a half year
old twin boys here and yeah, my wife and they need me at home. So
today's world with all teams and Zoom make it possible to earn the
money from home. And, the thing is, time is limited and you know
how fast the Power BI platform grows. And honestly, I always think
I know too less. That's a personal problem, I know, but that makes
me think I haven't got the time and energy to dip my toes into
Power Automat, or Power app or something like this. No, I haven't
tried that yet. I am willing to do so when it makes sense, but I
didn't do it until now.
Rob Collie (00:58:34):
I have that in common with you. Our company does all these things,
but I personally am still very close to home. If I'm using the
tools myself, I'm really just using Power BI. But in terms of our
client engagements, we've really diversified quite a bit. And it's
not because we wanted to diversify; it's that our clients' needs
started to mature. What we found is that, once they start to get
the BI problem mostly under control for the first time ever,
they've never have been anywhere close to that, they take a step
back and their focus zooms out a little bit. And that's when it
dawns on them that, it sounds like a cliche, but BI is really just
one part of the overall improvement system.
Lars Schreiber (00:59:20):
Sure.
Rob Collie (00:59:21):
So they start to focus more on the inputs and outputs, the human
processes that the dashboard, the report might recommend or
suggest. How do you make that more convenient? How do you tie it
closer into the workflow, essentially? Right? And so it tends to
grow outward as the BI problem starts to stabilize a little
bit.
Lars Schreiber (00:59:41):
There is no doubt that the Power platform is a huge thing and that
this no code, low code environment is something that will Be
successful in the future. For sure. But I'll learn something from a
good friend of both of us, Matt Ellington. You have to know when
you have to ask for help, and you can't know everything on your
own, or for yourself, or what's the right English at this point?
And this is what I do; I try to put my strengths together and just
consult for Power BI. That's why I don't offer VBA consulting
anymore. I just don't do it. I didn't do it for the last three
years or four years. And this is why I don't offer service at this
level.
Rob Collie (01:00:20):
And that's wise. Super wise. We got dragged into a VBA project
recently, and I was telling everybody ahead of time, "I don't know
folks. I think we should pass." And sure enough, you just get into
this really obscure stuff. People come to me like I'm like, "I have
no idea. You're asking me about some really, really obscure nuance
of the object model and how it behaves. Now we need to go find the
three people in the world know the answer to this." Used to be
five, but now we're down to three. Specialization is amazing,
right? I wouldn't be doing things any differently than you if it
was still just, me for sure. I mean, there's no two ways about it.
I've really only learned three or four really hard skills in my
whole life. It's a very short list of things that I've learned to
the level where I can claim competency. Excel is one of them. DAX
and simple data modeling is another. I can't even really claim
Power Query. I'm not very good at it. You talk about knowing when
you need to ask for help; whenever I write Power Query, I need to
ask for help. I know that it's coming, just a question of when.
Lars Schreiber (01:01:29):
Yeah, but there is so many amazing people in the community right
now. Every time when I think my Power Query solution is good. I
talk to [inaudible 01:01:37], I recognize, no, it's shitty.
Rob Collie (01:01:38):
Yeah.
Lars Schreiber (01:01:39):
And when you think you have master your DAX measure, and then you
read the next article by the Italians who tell you something about
shadow filter context that I still don't understand. There's no way
you can think you mastered something.
Rob Collie (01:01:52):
Shadow filter context. I never even heard of it. This is a rumor
it's like Bigfoot. It's a Loch Ness Monster. There's no such thing.
It just blows me a way that I can look at people's DAX today and
have no idea what it's doing. Not just the daks language, but the
dialects that people have adopted, the style of DAX that a lot of
people now write. I get the sense that the things that are being
written, I can write formulas that do the same thing. My formulas
will be about 1/10 the character count. It's just that at scale, my
formula's not going to perform very well, compared to this
multi-paragraph thing. I like to tell myself that. It might be that
I couldn't have written the formula at all. I don't know, but every
now and just for grins, I just crack open to the random DAXs
article on the internet, just so I can laugh at how mind blowingly
out there it can be, relative to what I understand. If that had
been the way DAX was when I was getting started, I would've never
approached it. I would've never learned it. I would've never gotten
in. If it looked to then the way it does now, I would've said, "No,
not me."
Lars Schreiber (01:03:00):
But, what you did for the community, at least from my point of
view, was a huge deal because I can remember that I bought the
first book of our Italian friends, and I simply didn't understand
what they were talking about. The first book I ever read about
Power Pivot was from Mr. Excel, Bill Jelen, and it's basically a
number of formulas explained what they're doing, but without
concepts and data modeling whatsoever. And the second book I read
was from the Italians, and I had no clue what they were talking
about. And you explained it for the Excel user, and that made a
huge difference. The audience that should use Power Pivot. So that
made it possible for me to dive into this topic and then read
further books.
Rob Collie (01:03:48):
I have this visual in my head of I'm down on my hands and knees,
and you're using me as a stepping stone. You're stepping up on my
back to get to the Italians. So the book by Mr. Excel, Bill Jelen,
I read that same book. He and I both acknowledged that it was
essentially a tour of the UI. "Here's what every button does," and
some examples and things like that, but he took a pass just
completely didn't address the DAX language. It was too ambitious of
a project at the time he wrote that book, I didn't know DAX.
Lars Schreiber (01:04:22):
Right?
Rob Collie (01:04:23):
There's no way he could have known DAX. Talk about naming, and
naming mistakes and things that you regret. The reason my first
book was called DAX Formulas for Power Pivot, we envisioned it as a
companion to his book.
Lars Schreiber (01:04:36):
Ah, interesting.
Rob Collie (01:04:38):
"You've had the UI tour, now let's talk about the formulas." It
ended up being more than that by the time I was done, but the name
was stuck. We'd already put it in the catalog and I didn't even
think we should change it.
Lars Schreiber (01:04:50):
How can it be that you didn't know DAX at that point?
Rob Collie (01:04:55):
Easy. Because I don't learn things. It's a better question, how did
I ever come to learn it? So I had sort of two separate dates of
leaving Microsoft. I had one in April, 2009. I found out that I was
basically losing my court case, and I had to move to Cleveland to
be with my kids. I wasn't going to be able to keep them in Seattle
with me; their mom was going to take them to Cleveland. So I had a
lot of things to get in order, and it wasn't until February of 2010
that I officially left Microsoft. But in the intervening 10 months,
other than getting relocated and moving to Cleveland and all of
that, and selling my house in Seattle, during that time between
April of '09, after that is when the DAX language first started to
truly appear, even in the internal builds of Project Gemini at
Microsoft. The calculate function didn't exist, I don't think. It
didn't exist or it just recently been conjured about the time I
stopped going into the office.
Rob Collie (01:06:00):
So I learned DAX from the outside from a great distance, and I
might as well have not have been a Microsoft employee when I was
learning it. And I had nothing to do with it. I had nothing to do
with it's design. I wish I could claim I did, because it's so
amazing, but I didn't, I had nothing to do with the DAX language. I
had originally been involved, but none of the work that I did ever
made it into the product, because it wasn't as good. What they came
up with later was far, far better. My thinking on the topic wasn't
very good.
Lars Schreiber (01:06:32):
I had Jeffrey Wong on my podcast to that topic.
Rob Collie (01:06:35):
Yeah. And you know what? I think I actually listened to that
one.
Lars Schreiber (01:06:37):
Okay.
Rob Collie (01:06:38):
And I don't consume anything. So that's where the first time I ever
heard Jeffrey make the distinction, and it's an obvious distinction
once you hear it, but the distinction of Power BI being a
model-centric BI tool, whereas all the competitors are
report-centric. Heard that on the Lars three syllable Schreiber
podcast. And I've been using that a lot since then, so I appreciate
you having Jeffrey on the show so I could have that lingo. Yeah. So
Jeffrey's one of them, for sure. Jeffrey, Marius, Amir, Howie. Sure
there are a couple others in there that I'm forgetting, but this
was the task force, and they worked long nights just arguing over
the design of the language. Amir eventually had to start using a
trick, and he would tell me this during the day. "So what I do is I
just tell everyone we're not going home until we decide this."
Lars Schreiber (01:07:33):
Wow.
Rob Collie (01:07:35):
That was the only way that they could reach consensus sometimes. No
one gets to go home to their families until we're... People would
tend to abandon their positions a little bit more gracefully under
those circumstances. It's a very, very argumentative team in a
mostly good way.
Rob Collie (01:07:53):
The huge advantage I had, I had two big advantages. Number one was
I had nothing but spare time. We talk about the tyranny of the
full-time job, and how it's the enemy of learning new things. I
didn't have a full-time job then. My full-time job was really to
learn about and blog about... That was sort of the deal that I'd
kind of handshake made with Microsoft was, I'm not in the office,
I'm not doing my normal job, but in this interim period, I'll
advocate for you. And I did. And along the way I discovered, "Oh my
God, I can actually learn this. It's actually really, really good.
It's way better than I thought it was going to be. Oh my God, I'm
going to stick with this even after I leave Microsoft, this is
going to be the thing." I had no idea that was going to be the
outcome.
Lars Schreiber (01:08:34):
But you had a good idea of what it would become in the future, that
it would revolutionize how people deal with data. And that was
well, the big thing?
Rob Collie (01:08:46):
After a few months I did, but I promise you when I launched the
blog, I did not know that. Not at all.
Lars Schreiber (01:08:53):
When the idea of Power Pivot came up, did Microsoft have an idea of
the Power platform or at least Power BI platform? Was it planned
back then?
Rob Collie (01:09:02):
No. No, not that I was aware of at all. Have you managed to get
Amir on your podcast?
Lars Schreiber (01:09:06):
No. I have planned that for a longer time, but [inaudible 01:09:11]
did a very good episode with him and he asked a lot of questions I
already had. So.
Rob Collie (01:09:15):
Yeah. The format of this podcast, we're lucky. It's basically just
get together and chat. If you don't prepare questions, you don't
get concerned about what other people have already asked on other
shows. Our lack of preparation is a feature. It's genius.
Lars Schreiber (01:09:35):
That's something cannot do with my podcast, because English is not
my native language, so I really prepare the questions, not only
what I ask, but also how I will ask it, so that the person I'm
asking will definitely understand what I'm talking about. But yeah,
if you can do it in this relaxed form like we are doing it right
now, it feels comfortable, that makes fun. It's entertaining. I
like it.
Rob Collie (01:09:57):
Well, Lars, I have really enjoyed this. Is there anything that you
were expecting me to ask you, or hoping that I was going to ask you
that I haven't?
Thomas LaRock (01:10:06):
Or wishing we didn't ask you?
Rob Collie (01:10:09):
There's always that.
Lars Schreiber (01:10:10):
Honestly, no. I was just very excited for this episode because you
were my starting into this journey. It's definitely like that. And
I can remember in 2014, I published my first English speaking blog
post on your block.
Rob Collie (01:10:28):
That's right.
Lars Schreiber (01:10:29):
That was a very special point in time for me. And since then I
thought, "Okay, nothing is impossible. Just try it."
Rob Collie (01:10:37):
Hey Luke, let's make sure we look that one up and link it in the
description.
Lars Schreiber (01:10:41):
No, don't do so. I read the article today. It's shit.
Thomas LaRock (01:10:47):
Oh, we are keeping that.
Rob Collie (01:10:47):
Hey. Oh, come on, why not? I that's part of it, right? That's the
other thing I always tell people in my classes back when I taught
them was, "Your first go goal is to be bad at this because when
you're bad at this, you're amazing. The version of you that is bad
at these tools is going to be so freaking capable, relative to what
you're used to. Yeah, but you're going to keep getting better,
which is going to shock you. Every three to six months, you're
going to look back at things you did three to six months ago and
go, "Oh, I was so cute back then. Look how adorable I was. I had no
idea what I was doing. Oh, it's embarrassing. The way I did that.""
Then I would tell them most of the time, when you are embarrassed
by something you did in the past, you will not go back and fix it,
because that thing that you built is already working and working so
well, and everyone that's using it is loving it. And there's more
ROI in doing new magic, in new places than going back, correcting
something that only you care about."
Rob Collie (01:11:48):
Think about it. You know this Lars, right? You've been around the
website long enough that I carried around an extra [inaudible
01:11:56] in the greatest formula in the world pattern in my first
book, and defended why it needed to be there, and explained in
great detail why it needed to be there. On Twitter recently,
someone showed me a picture of their original DAX Formulas for
Power Pivot Book, and I'm like, "Oh, does that copy the extra
[inaudible 01:12:14]?" They're like, "Sure does." How about this?
If what you wrote in 2014, if you weren't embarrassed by it, if you
didn't think it was very good, if you still thought it was good,
that meant that you hadn't evolved in the intervening seven
years.
Lars Schreiber (01:12:31):
Exactly.
Rob Collie (01:12:33):
Of course you've evolved.
Lars Schreiber (01:12:34):
That's how I like to see it. It's good if what you did in the past
is not completely shit, but that you make an improvement is
hopefully what happens.
Rob Collie (01:12:47):
All right. I'm going to go look at your 2014 article. I'm going to
make a decision. Come on, you can't get any worse then publishing,
printing 10,000 plus copies of a bound volume that dedicates an
entire page to explaining why this all is necessary when it's not.
One of my earliest blog articles, I didn't realize yet that I
didn't know the words filter context yet, but I didn't know that
filters didn't travel both directions over a relationship. I didn't
understand that yet. I'm really grateful now that they, by default,
only travel one direction makes things a lot more predictable. But
I didn't know that then. And I published an article where, "These
are the right numbers," and Casper who wasn't at Microsoft yet, he
wrote in some comments and said, "Ah, I'm not getting the same, I
don't think that's the right answer." And I'm like, "Of course it
is." And then I go looking at it. I'm like, "No, you're right."
Lars Schreiber (01:13:48):
He was on your show as well, right? I have heard this story
before.
Rob Collie (01:13:53):
I was dividing by the whole calendar table, and thinking I was
dividing by the number of days that were active in the fact table.
Nope.
Lars Schreiber (01:14:04):
But on the other hand, because you learned it that way, you know
how to teach it to people who have never heard about filter
context, row context, because you have been there.
Rob Collie (01:14:15):
The jargon free zone.
Lars Schreiber (01:14:16):
Yeah.
Rob Collie (01:14:16):
Being a bad learner, I think that's a fair label to attach to me,
being a bad learner. Made me a better teacher.
Lars Schreiber (01:14:24):
Are you kidding me? That's my motto for last 10 years. Yes.
Rob Collie (01:14:28):
I'll leave you with a thought; you pointed out that now it's a
remote work culture, and you're working from home and everything.
Iceland calls. Oh, but your wife has a job still tied to the
region, so.
Lars Schreiber (01:14:41):
The thing is, if I go to Iceland, I have not only remote
relationship to my customers, but also to my wife and my boys.
Rob Collie (01:14:49):
I'm super happy where you found yourself; living the dream. Not
many people get to say that. So congratulations. Long road.
Lars Schreiber (01:14:56):
Same to you. It was a pleasure being here, thanks for the invite.
Thanks for the work you did. You influenced many, many people, one
of them. I have been, thanks to you.
Rob Collie (01:15:05):
Talking to people like you with stories like these, it's the
coolest. To be perfectly honest, it's even hard for me to
emotionally believe that I played a role like that for someone like
you. It's hard for me to believe that you didn't emerge fully
formed, you know? It's just so authentic. It's just a really
touching thought. It's a good thing what we've done.
Announcer (01:15:26):
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