Individual Excellence vs. Organizational Impact

Guess Which One Grows Your Career (or Company) More? (Hint: It’s the One on the Right)
(But Individual Excellence is a Prerequisite for Org-Wide Impact)

“Cuz we need a little, Con-tro-versy…”

  -Marshall Mathers

Don't Mess with Ken PulsLast week’s post on learning DAX before learning M mostly met with positive reviews, but also drew some fire.  A few staunch M supporters showed up and voiced their disagreement – including the one and only Ken Puls.  Now I know from experience not to mess with Ken…  OK OK, I confess…  messing with Ken is a barrel of monkeys actually.  Put it on your bucket list.  That said, I have immense respect for his skills and perspective, and only enjoy messing with him out of friendship.  He’s an amazing human being and learns more things in a year than I could ever squeeze into my leaky head over a lifetime.

But I still firmly believe in what I said.  I’m not here to offer apologies – only clarification and justification.  There very much IS an olive branch in all of this, but again, that comes merely from clarification.

First clarification:  I love love LOVE Power Query and M!  They are a godsend!  I never said what some people thought I was saying, which was “meh, you can ignore/neglect that part of the platform.”  Nope, you absolutely benefit tremendously from both.

The minor tension from last week raised a MASSIVELY important point, one that transcends any technical debate and puts things in their proper perspective.  So I’m grateful for the opportunity that the misunderstanding provides us.  Let’s begin with…

What’s the Max Benefit of Automating Your Job?

The Max Benefit of Efficiency Gains in Your Job is Approximately "One Full You"

At first this sounds like an incredibly abstract question.  I mean, how can you put a dollar figure on massive gains in personal efficiency?  Sounds impossible, right?

But I’ve got a card up my sleeve:  how much is your salary?  It’s not terribly outlandish to say that the best you could EVER do, in terms of speeding up the tasks in your workday, is to completely make yourself redundant.

And the market has already put a dollar value on THAT, right?  It’s called your salary.  No, it’s not a perfect number, not at all.  Many of you reading this are criminally underpaid in some very real sense, for instance.  But this is what the market is saying today, and that’s a very real quantity.  Furthermore it’s not like it’s really possible to automate 100% of your duties using ANY of the tools available today (thankfully!), so it’s somewhat “gracious” to set the max at 100% of salary.

The folks applying for our Principal Consultant role lately have ranged in current salary from a definitely-criminal $40k on the low end to a damn-near-executive $160k on the high end.

So let’s continue with the “gracious” theme and go with the high end:  $160k per year is a serviceable maximum ROI for “making my current job run faster.”

You might be thinking, at this point, that I’m still not being Gracious Enough.  It’s possible, after all, for a single hyper-efficient individual to suddenly replace MULTIPLE other individuals right?  Setting aside the distasteful notion of those lost jobs for a moment, I still think my $160k figure isn’t that bad, given that it’s 100% of an entire individual on the high end of the range.  But fine, if you want to multiply it by 3 and make it $480k, I don’t think that necessarily undermines any of the points I want to make.  I’m in the business of adding more zeroes to the productivity multiplier, not linear multiplications.

What Does that Have to Do With DAX and M?

dax and modelingAt first, nothing.  Both DAX* and M, in the early going, are BOTH very much “speed up my current workflow” kind of things.  And that’s perfectly natural – what you’re currently doing is ALWAYS the best place to start, the best place to learn.

(* Remember, when I say “DAX,” I use that as shorthand for “DAX and Modeling,” where “Modeling” is best described as “figuring out how many tables you should have, what they should look like, and how they are related).

And that “improve what I’m currently doing” lens is why M/Power Query steals the show in the earliest demos – it’s easier to see how it’s going to change your life, because it automates/accelerates a larger percentage of what Excel Pros have traditionally done.

Hold this thought for a moment while I introduce something else…

Raining Money:  The 100 Million Dollar Workbook!

Raining MoneyThis is a real thing, it belongs to one of our clients, and we helped them build it.  To call it a “workbook” is a bit of an insult of course, because it’s a modern marvel – an industrial-strength DAX model with a suite of complementary scorecards as a frontend.  But all built in Excel.  (Power Pivot, specifically).

And this model has provably returned about $25M a year to the bottom line for this client.  As in, profit.  Not revenue.  Pure sweet earnings.  This workbook is visible on their quarterly earnings reports to Wall Street.

And this wonder of the modern world is well into its fourth year of service now, bringing its lifetime “winnings” into the $100 Million range.  No lie.  This happened, and continues to happen.  This “workbook” is a tireless machine that makes it rain money.

Let’s do some math:  $25M per year vs. $160k per year is… 150x.

In other words, the ROI of this project went FAR beyond any amount of “accelerating what we already did.”  It was, instead, a MASSIVE dose of “doing something we’ve NEVER done before.”

 

“What We’ve Never Done Before” (WWNDB) Has Practically-Limitless ROI

What We've Never Done Before: A Limitless Landscape. No Ceiling.This may sound like a cheap verbal trick, but I sincerely think it is a weighty truth that everyone should internalize.  The workbook above, which now in some sense runs the show at this large client, had no predecessor whatsoever (its “forerunners” were a scattered collection of hundreds of distinct reports, each of which was just burying readers in borderline-raw data).  For an even bigger example, consider that Instagram started as a hybrid of Foursquare and Mafia Wars before deciding to go “all-in” on their most popular feature, photo sharing.  The blank canvas has no ceiling, if you permit me to mash-up some metaphors, and both of these success stories are rooted in a combination of analytics and courage.

What we’ve been doing traditionally, in both the traditional Excel and traditional BI worlds, is nothing to brag about.  Most of our reporting and analysis output has been, traditionally-speaking, designed by the path of least resistance – as opposed to defined by careful and creative thinking about what truly matters.  The president of one of our clients/partners’ told us last week, “people tend to measure what they can easily count, as opposed to what they SHOULD measure,” and I just about leapt out of my chair screaming “YES!  PRECISELY!”

If you want to learn more about this topic, start with Ten Things Data Can Do For You and We Have a “Crush” on Verblike Reports.  For now, it’s time to move on to Leverage.

Leverage is Everything.  It’s Being Bigger than Just You.

Leverage is Everything

You wanna know why Data is so “Hot” these days?  It’s because of Leverage.  Data is hot precisely because proper application of data can impact the behavior and productivity of MANY people simultaneously.  You can’t typically save or make millions of incremental dollars as an individual, but it’s “easy” to do if you can magnify benefits across dozens of other people – or hundreds, or even thousands (as is the case with the $100M workbook).

In fact, it’s worth considering that the $100M Workbook actually offers only modest benefit!  On a per-person basis, on a single day, you wouldn’t even notice the difference.  But multiply that modest, say, 3% benefit across tens of thousands of people and 365 days…  and you get $25M per year.  When you have the power of Leverage, you don’t even have to find something “big,” like the Instagram “pivot” from one mission to another, to get something BIG.

We are all, everyone reading this, INCREDIBY FORTUNATE to be working in data, because of its somewhat-unique capacity for leverage.  So many jobs, whether white- or blue-collar, are essentially cogs in the machine, and the top-end benefits they provide are limited to the “just you” size.  But WE have hit the jackpot.  WE have a job that is “unfairly” capable of leverage.  It just fell into our laps.  But then, the mind-numbing dosages of VLOOKUP (in the traditional Excel world) and endless documentation and miscommunication of requirements (in the traditional BI world) deflected us off into a relatively un-ambitious mindset.

Data has ALWAYS had the advantage of Leverage, but the traditional methodologies and tools brought tremendous friction and inertia to the table.  They wore us down – in terms of time, money, and psychic energy.  They “chokepointed” our potential.  They enforced a terribly-linear culture of thinking.  In short, the traditional tools took the potential 100x or even 1,000x leverage possibilities of Data and tamped them down to about 10x – still good!  But so much less than what we COULD do.

Well guess what?  No more chokepoint.  Whatever you want to call it – Power BI, Power Pivot, Modern Excel – the next-generation toolset from Microsoft gives us those extra zeroes of potential.

Why DAX is “Better Than” M

Note the quotation marks in that heading, because the next section is more conciliatory, but there IS something very important to bring home here.

If you MADE me choose one or the other, I’d definitely choose DAX, because I think it offers us the virtually-unlimited twin powers of  WWNDB and Leverage.  In fact, I don’t think that, I know that – I (and my companies) have been blowing people’s doors off with this new toolset since 2010.  We didn’t even get Power Query until what, 2014?  Fully half the lifetime of this revolution pre-dates M.  Even the $100M Workbook predates M!  Heck, until Power Update came along, you couldn’t even schedule refreshes of models that relied on M, which almost by definition “funneled” M usage down the “just for me” path – and to this day, Microsoft still hasn’t finally released a server that natively runs M.

I just don’t think it’s nearly as easy to explore/exploit WWNDB or Leverage via the M path.  Not impossible, because there are plenty of exceptions that prove the rule.  And to be clear, I think most of the exceptions will be in the WWNDB category, not the Leverage category.

And that was kinda my whole point in last week’s article – Power Query dramatically captures the attention of new converts to Modern Excel precisely because of how well it fits and improves What We’ve Already Been Doing, as Individuals.  This is a Good Thing!  No caveats needed.  I just don’t want anyone to become so distracted with it that we miss the Big Wins of WWNDB and Leverage.

How M/Power Query Fits into the “Big Wins”

The Big Wins of WWNDB and Leverage, Provided by DAX & Modeling in Power Pivot and Power BI

This is What We Can Do With “Just” DAX and Modeling

The picture above illustrates how a single individual (you, or a member of your team) can achieve wins MUCH bigger than just them.  And it’s my experience-powered belief that you cannot get a Win of that size without leveraging DAX and Modeling.

But what if you THEN take that single individual’s newfound powers of WWNDB and Leverage, provided by DAX and Modeling, and now make THAT person more efficient?  “Holy Additional Multiplier, Batman!”

The Big Wins of WWNDB and Leverage, Provided by DAX & Modeling, MULTIPLIED by Power Query and M

If Our DAX Modeler Superhero “Levels Up” with the Efficiency Gains of Power Query / M…  Look Out!

Yeah, if you take THAT person, and make THEM more efficient, WOW, you can do EVEN MORE of the amazing, transformational, WWNB-and-Leverage style work.

Which we can all agree…  is a Very Good Thing.  One of my favorite personal sayings is “the length of a rectangle is not more ‘responsible’ for the area of the rectangle than the width.”  Double either one, and you double the area.  But that’s essentially my point in a nutshell – you can 100x with DAX and modeling, AND you can double with M.  If you had to choose one, choose 100x.  But we don’t have to choose.  Adding Power Query and M to your org-wide-impact powers, even if it’s “just” 2x or 3x, delivers JUST AS MUCH, or more, incremental Big Win as the original 100x.

We can have our flagons full of mead and drink them too, as Lothar once said.

  Subscribe to PowerPivotPro!
X

Subscribe

Rob Collie

One of the founding engineers behind Power Pivot during his 14-year career at Microsoft, and creator of the world’s first cloud Power Pivot service, Rob is one of the foremost authorities on self-service business intelligence and next-generation spreadsheet technology. 

This Post Has 32 Comments

  1. I think we have the old adage of comparing Apples with Oranges. I currently barely use DAX and use “M” every day. Why no DAX? Because I am still learning and haven’t been forced to use it. And about 3 other people in the my building seem to have heard of it.
    Do I want to use DAX? Yes…. I mean YES !! Regular Pivot Tables are fantastic, but they still miss a couple of calculations.
    The poorer one’s data source is, makes “M”‘s importance increase. Not being able to connect to our DataBase means dealing with text and Excel reports so “M” is huge to me.
    It’s very easy to use “M” without going to DAX.
    It’s as easy (depending on your DB-admin) to use DAX without using “M”.

    “M” will be more impactful to individual data users, but DAX will create the (advanced) consumable content for distribution.

    1. Agreed. And this is why I wrote last week’s article in the first place. Too many people seem to missing the potential of DAX & Modeling, and I want to help make sure people know about it – and WHY they should know about it.

      Hopefully Brian you have found the two articles helpful, cuz that was the intent (rather than picking a fight).

  2. I would like to respectfully take issue with your math on the relative benefits, as from an organizational standpoint, a much larger number of people can/will use M than can/will use DAX. You don’t need to be a data person to get M, which throws open the gates to a lot more people. And at least in some industries (those of us not in tech, in particular), there’s still such a huge amount of work that hasn’t been properly digitized that the potential scope for M solutions is literally beyond my understanding. Sure, most if not all of that could have been done with a sufficiently good VBA wiz, but those are pretty rare on the ground, and with the ease of M, self-service ETL is already changing my office and starting to change my company from an operations perspective. DAX is changing our decision making (also in huge, amazing ways), but M is the savior of our desk-level agents.

    That said, I’m VERY much in the both camp, I just think the relationship is more like your legs (M) vs. your arms (DAX). You need your legs and feet to get you to places and do a lot of the heavy lifting, and you need your arms and hands to do the manipulation. Sure, you can use one without the other, but using them both together is a much, much better plan. Even without analyzing anything, my M solutions usually have a little bit of modeling in them, and I almost never do anything in DAX without pulling everything in via M in the first place.

    1. I disagree that there’s a much larger audience for M than DAX & modeling. I think there’s just a much-easier path to RECOGNIZING the benefits of M. Which is why I took the time to write last week’s article.

      Seriously, damn-near EVERYONE can benefit from portable formulas, where “everyone” means “anyone who creates pivots, writes macros, or uses VLOOKUP.” Everyone (with the same definition of everyone) can benefit from the CALCULATE function. And none of that is really at all difficult to learn. Easier than even the simplest hand-written M, for instance (but no, not easier than clicking buttons on Power Query’s ribbon). Easier than SUMIF, actually, if you were forced to learn both from scratch.

      And even if I conceded that the number of potential M users heavily outweighs the number of potential DAX users (which I am NOT conceding, heh heh), the potential power and career growth for any SINGLE individual makes DAX & modeling an absolute-must, and that absolute-must is getting neglected by a swath of our newly-arriving friends in the Modern Excel club.

      If you’re doing both, you’re already doing it right. Notice how many comments there have been on these past two posts where people are saying some flavor of “I use M everyday but have yet to see the value of DAX?” THOSE are the folks I have been seeing a lot of lately, and those are the folks I want to help. You weren’t the target, and neither was Ken/Miguel 🙂

      1. Fair point, I just looked through the number of people who were leaving DAX aside, and it is more than a little alarming. I’m in full agreement that ignoring DAX is like ignoring your hands. I just think that DAX and M (at least the button-clicking in Power Query/BI) should be learned simultaneously. Honestly, reading both your and Ken/Miguel’s books is the way to go, from my perspective, because I feel like both are absolutely crucial.

        1. These two are the Batman & Robin in data Gotham city. I’ll leave it to the imagination as to who assumes the Joker, Riddler & Penguin roles

  3. Rob is right that DAX is something new. It provides insights that would have been incredibly hard to achieve previously. On the other hand, M does not take me to a new place; it just gets me there faster and with less hassle. That means replacing my previous 20+ year data prepping workflow of Excel formulas, VBA, and Access. M and DAX are both no longer “nice to know”, but “need to know”. Thanks for the article.

    1. Jeff, bingo. I wish I had your gift for brevity.

      You are “picking up what I’m putting down,” as they apparently used to say in the 70’s to indicate the state of deeply “digging it.” :p

    2. What a great way to put it. I could see that they were both very important to the new power of Excel, but this helps me understand why. I have been teaching workshops and doing presentations on both PQ and PP and “knowing” this without ever really knowing it. Now I get it!

  4. Fantastic summary Rob. Your enthusiasm for DAX is contagious.
    I still don’t agree with “Learn DAX first”, and my view is there will be way more individuals who use the Power Query interface than ever write a DAX formula.
    But I can whole heartedly agree that DAX is already providing massive business value.

  5. Rob, both DAX and M have completely failed because clearly neither of them have delivered the means for you to draw better stickmen!! There’s obviously a need for a CONCATENATESTICKMAN() function. Keep up the good work. Sorry I missed you when you were over this way. Oh, and for the record, if there’s a vote, I’m in the DAX corner!!

    1. Welcome back DC! Your posts still get a lot of traffic to this day.

      Definitely catch up next time sir 🙂

      (And we are approaching the point where I can commission custom art for all of this – it’s actually more the speed that keeps me doing it myself, rather than saving the expense. Communication cost – the dark matter of basically any endeavor.)

  6. OK – Enough of the Sizzle
    How about showing us some of the steak…?
    Redacted where necessary of course.
    DAX/M is still early adoption for the majority of excellent users – is it not?

    1. Hi Fran! What would qualify as “steak” for you? Serious question.

      And absolutely, I think we are still very early in the AWARENESS curve, much less ADOPTION or mastery.

      1. I feel there is a huge discrepancy between the truly invested experts and the general excel user. What is missing are some examples of real-world applications – not just rewrites of existing solutions – but some application examples that truly showcase PQ/PBI in the real world – something to strive for – some targets for excellence for the general user caught in an ever increasing maelstrom of data.
        -Fran

        1. Happy to showcase some of these to you Fran at a time to suit. It may even help with gainful employment opportunities ????

        2. So… in some sense Fran, we’ve shared tons of “steak” over the years on this site, it’s just that we aren’t overly-fond of the idea of “recycling” older posts. There are > 900 articles on this site, most of which are specific techniques. Lately I’ve personally writing about the less-technical aspects of all this, such as the human factors in the equation, strategic thinking, etc.

          What would you think about a “Best Of” compilation, where we compile links to some favorite (and powerful!) techniques we’ve covered over the years?

  7. I was fortunate enough to learn both M and DAX at the same time. Today I would never put any table into a powerpivot model wothout having it pass through powerquery first. Even if not much cleaning is required, I’m just doing it automatically. I rarely use normal Excel now. However I do agree with you. If I steal Matt’s metaphor above, M is my legs that take me to places, and DAX is my hand which I use to produce “Art”.

    1. Many thanks to Rob Collie and Ken Puls for leading us to the next industrial revolution. VBA and PQ are a blessing. I’m sure PP and Power BI desktop will jolt the current BI visualisation tools when these gains momentum in the small businesses. Keep up the good work, guys. Edmund

  8. Hi Rob! Your article is a complete masterpiece. I’ve been focusing on learning the M language since it captured my attention at first, but reading this article has changed my mind. Power Query is very useful to automate your daily tasks, but if you need and want to get deeper insights from your data, you definitively need to learn DAX.

  9. I am not sure there is even a need for this debate and I am not sitting on the fence either when I say that DAX cannot exist without M and M cannot exist without DAX. What is great is that both (the ability to extract or get data and the ability to analyse the data are in one suite.

      1. If anybody is really only using one they have missed the whole point about Self Service BI.I would also guess that most M over DAX supporters build more complex data models with simpler measures whilst the DAX over M supporters build simpler data models with more complex measures. As it makes no difference to the end result I guess it really comes down to where you feel more comfortable.

    1. Taking the PQ addin out and making the Get and Transform be part of core Excel in 2016 should do a lot for adoption of PQ / M!

      If only PP / DAX was not hidden in only the pro versions of Excel; it’s definitely a roadblock to getting more adoption.
      There’s still a distrust amongst finance people that if they can’t see the formula (looking into a cell and seeing VLOOKUPs), they can’t trust it!

  10. I like the comparison of M to legs and DAX to arms. I jumped all in to DAX mainly because I viewed it as an extention of what I was already doing in Excel but on steroids.. and I couldn’t learn enough at the time. My problem now is that I try to do everything that M should do with DAX. I’m walking with my arms.

    As a business user I believe DAX is really the key driver to the benefit. The flexibility, the time intelligence, the ability to crunch massive amounts of records should blow anyone away. However, DAX and M need to be used together to maximize the performance of the model.

  11. “Long-time listener, first-time caller” here. First,, thank you for the 900+ posts, the book, and the availability. That’s not just for Rob but for all of you “one percenters” of this world: The Italians,, The MVPs, the contributors to the community who get little attention or accolades, but who continue to stir that bubbly cauldron every day. All of my knowledge, going all the way back to the original Mr. Excel/ExcelISFun days of the early 2000s, comes from you guys and your willingness to share. You’ve done well for yourselves, I’m sure, but also helped an army of well equipped analysts/data specialists do better for themselves, too. I’ve been on my own since 2010 and it’s been a blast. Thank you all. Sincerely.

    Back to the discussion here, I remember back in 2008 when I was contemplating how Data, specifically Excel expertise/mastery, could be leveraged into a business in a world that was being taken over in a messy way by the big houses (Oracle, Cognos, Business Objects, etc). My feeling back then was: On the low hanging fruit alone that those companies provide (their inability to go the “last mile” and the constant budget overruns and bad implementations), I could be in business for a long time, coupled with “regular” advanced Excel work that provided value on its own. Simultaneous with that, I came across a presentation by Microsoft about the future: Workgroup BI. The non-klunky, non-insane take on BI and data analysis/architecture for people like me. DING! DING! DING!

    This entire site (PPP) is the intersection of that realization by Microsoft and its Magic Quadrant standing. Every time I read posts like this, I look back at the last ten years and go…WOW. It’s happened and it’s incredible.

    That’s why when I hear Rob paying homage to M/PQ but going all-in on DAX/Modeling, I totally get it. They are both extraordinary tools (I’ve only been sincerely using them with clients for a few years, myself, so very much a noob) that level up incredibly, but I see DAX/PowerPivot/PowerBI as THE thing that allows little ol’ me to go head-to-head with The Big Guys and offer clients an ability to do things they never could do before:

    * Circumvent IT
    * Play inside massive and complex data structures being maintained elsewhere
    * Provide similar value as the Big Guys for literally a fraction of the Time and Money
    * Working in tools they already have (no more big investment)
    * etc etc.

    The market for just low hanging scrubbing/cleanup work in PQ/M is no joke. It’s probably in the hundreds of millions of dollars when considering how so many companies manage/use their data (I.e, NOT like anyone who reads this blog). It’s everywhere, just hard to explain/sell at first.

    That said, the market for DAX/Modeling is probably easily into the billions/tens of billions when you think about what Microsoft has offered and the middlemen they’ve eliminated and who they are targeting (The Big Guys).

    It’s an extraordinary suite. I’m glad I didn’t go down the traditional BI/Tableau/whatever rabbit-hole. All super-useful tools, no doubt, but 500 million Excel installs by Microsoft is a force, man.

    I actually think we’re all on the same side here, but I do admit I’ve found these last two posts and their comments ridiculously awesome.

    Thanks, everyone!

Leave a Comment or Question