I am inspired to write this article from my my life long experiences working with companies that 1) embrace Self Service BI as a synergistic addition to Enterprise BI as well as 2) those that don’t embrace Self Service BI as anything other than a nuisance.  Of course there is another group that is full of people that “don’t even know what they are missing”, but I will leave that group is for another day.  When I say “those that don’t embrace Self Service BI”, I am not talking about the enlightened “Business People” of course.  I have never found 1 enlightened business person that failed to see the value in augmenting Enterprise BI with Self Service BI. I have found plenty of Business people that don’t know anything about Self Service BI however.  When I say “those that don’t embrace Self Service BI”, I am of course talking about a sub set of IT folk that for what ever reason will simply not open their eyes and ears (and minds) to understand or embrace the opportunity that presents.

First Some Definitions

Enterprise BI:  Traditional BI solutions for typically large organisations built on expensive hardware and software with specialist IT skilled individuals to build and operate the solution. The build projects tend to be long and expensive, and in some cases they deliver great outcomes for the business.  In other cases the projects fail to deliver on the promise.  Regardless of the measure of success, Enterprise BI has a long history and has in the past (and will continue in the future to) add a lot of value to many large enterprises.

No BI:  Some organisations don’t have BI for what ever reason.  The normal reason is the cost and/or risk of project failure can’t justify the journey.  These organisations typically rely on either transactional system reporting (usually limited), and/or Excel, or they survive (or don’t survive) with nothing but gut feeling and instinct.

Self Service BI:  When modern BI tools can be put directly into the hands of the end user.  The users can gather and connect to their own data if they want, and/or simply connect to existing Enterprise data sources to build new reports and analyses in new ways.  Self Service BI tools can be an affordable BI solution for organisations that can’t afford Enterprise BI, or they can be a supplement to Enterprise BI.

The Enterprise BI Success-Failure Continuum

I am not advocating that most businesses should replace Enterprise BI with Self Service BI.  However in some (many?) cases Enterprise BI has not met (and can not meet) all the needs of the business.  Regardless of how successful Enterprise BI is in an organisation, there is still normally a place for Self Service BI.  However the role of Self Service BI does depend on the success of the existing Enterprise BI solutions.

Success or failure of Enterprise BI can be shown as a continuum.

continuum

The 5 sample points I call out (from best to worst) are:

  1. It adds lots of value to lots of people.
  2. It’s OK, lots of “export to Excel”
  3. Some use, but not worth the cost
  4. It is a failure and it is written off
  5. It is a failure but you keep it.

Note what I list as the worst possible outcome.  The solution is no good, and no one does anything about it.  This is much worse than writing it off as a failure as you can’t move on if you don’t accept you have a problem.

Self Service to Augment Whatever

Self Service BI should be, in most cases, an augmentation to whatever situation you already have (in my opinion).  In other cases it can be an alternative to Enterprise BI.  I can think of a number of scenarios.

  • You are going to purchase Enterprise BI but it is 12 months away – get started with Self Service BI.
  • You don’t know what you want/need from Enterprise BI – start prototyping with Self Service BI.
  • You have successful Enterprise BI, but you can’t afford to solve every bespoke requirement (due to timeliness and/or cost) with Enterprise BI– Self Service BI to the rescue.
  • You have Enterprise BI, but not all required data is loaded  – Self Service BI can combine what you have loaded with what is not.
  • You have Enterprise BI and it is great.  Some people will find it faster and more convenient to directly connect to the data warehouse with Self Service BI tools.
  • You have a mediocre Enterprise BI solution that can’t be written off.  Augment it now with Self Service BI.
  • You have a terrible Enterprise BI solution.  Keep the data warehouse and write off the rest (or don’t).  Build new reporting and analyses with Self Service BI tools and use these as your new consumption approach.
  • You are never going to purchase Enterprise BI – Use Self Service BI as both an Enterprise BI alternative and a Self Service tool.

Notice the absence of the following 2 items in my list above

  • “You’ve got Enterprise BI – you don’t need Self Service BI”?!
  • Self Service BI is now here – everyone get rid of your Enterprise BI solution.

So why then do some businesses (IT actually) insist on resisting progress with Self Service BI?  Well I think it can be for many varied reasons, including:

  • fear of loss of control
  • fear that there may need to be an expensive write off of an “asset” that is not being used – if people start using something else.
  • fear and/or ignorance of the unknown
  • fear of multiple versions on the truth

I am sure there are many more.  Notice the common word there – fear!

So What to Do

I don’t think there is one single answer, but there is a single common thread to every scenario.  A very smart guy once told me that “you get the IT department you deserve”.  What he meant by that is that the Business leaders have influence over what the IT department does for them.  There are normally (always?)  a lot more business leaders in an executive team than there are IT leaders. If the executive team accepts a sub standard result from IT, then it gets what it deserves.  If follows therefore that the secret to getting a great IT department (that supports Self Service BI and everything else) is that the executive must influence IT to deliver what it needs.  But it is unlikely that these B level managers will be the ones that discover Self Service BI, and hence it is up to the C and D level business managers to lobby and influence the executive business leaders for change.

org chart

So why don’t the B, C and D level managers influence the change?  I think there are 3 main reasons:

  1. They are not aware of the opportunity
  2. Fear created by IT that the sky will fall in
  3. They are not willing to tell another member of the executive how to do their job for fear it will backfire on them.

Where is the Motivation?

All of the reasons above for not influencing change can be overcome once there is motivation.

Let me talk about motivation for a moment. The fact is, us humans (and probably all living things actually) only do things we are motivated to do – really.

  • We go to work to earn money so we can live the life we want.  If we had a billion dollars, many of us wouldn’t go to work (certainly not our current job anyway).
  • We line up in a long queue to get a great coffee.  If the coffee wasn’t great, we wouldn’t line up.
  • We work late into the night for fear we will lose our job if we don’t get a task done on time.
  • We work late into the night because we want to get a promotion.
  • We work late into the night because we are inspired by our manager and we want to help them.

There are many reasons of motivation.  What will create motivation for an executive in a company to embrace Self Service BI is absolute clarity of the benefits.  

Become a Trim Tab

Stephen Covey in his book “The Eighth Habit” talks about the concept of a Trim Tab.  A Trim Tab is the little rudder that sits at the back of the big rudder on large ships.  These ships are so large, that they need very large rudders.  But the rudders are so large, that even the rudders can’t turn by themselves. A Trim Tab is the small rudder that turns the large rudder that turns the ship.   I remember the Trim Tab concept vividly as I was fortunate enough to meet Stephen Covey (one of my idols).  As Stephen was signing my book, he said “Matt, become a Trim Tab” – and that is exactly what I am trying to be.  You should too.

trim tab

You all, the loyal readers of PowerPivotPro.com need to become a Trim Tab – an evangelist, trail blazer and influencer of the larger rudders that steer the ship. If you can directly influence the Executive, then go ahead and do that.  If you can’t directly influence the Executive, then you need to influence someone that can.  The best way to influence C and D level managers is to show them what you can do with the Self Service BI tools.  Forget about the reasons why you are not allowed to do this or that and instead get on with using the Self Service BI tools you already have access to, to create value for your organisation.  Sooner or later someone is going to realise you are on to something.  The sooner everyone (even 1 person with influence) realises you are on to something, the sooner you will get the IT department you need.

Of course some of you can’t start using these tools because IT has locked down your computers so hard, that you can’t even install the Power Query plugin yourself (or Power Pivot if you are still on 2010).  But if that is the case, then getting this changed is a much easier battle to win – to get approval to install a standard Microsoft Plugin for Excel.  So fight and win that battle and then get on with showing the value to the C and D level managers, and have them influence the Executive.  Remember, most of the executives in most companies (if not all) are not in IT, and they will be motivated for change as long as they see the benefit.

Matt Allington

Matt Allington is a Microsoft MVP specalising in Power Pivot, Power BI and Power Query Consulting and Training based in Sydney Australia.

This Post Has 17 Comments

  1. Every time I read posts like this from you or Rob I feel like you’ve probably run into the same problems I run into every day at my job at every company you’ve ever worked with. Feels good at first, then bad again.

    But I will endeavor to become a trim tab.

  2. I think “fear” is too strong a word for ascribing failure in BI, and IT is probably the wrong target. You are making a very large assumption, which is that the information desired by the self-service people is available from the enterprise BI system or elsewhere with only a modicum of effort by the self-service users.
    Who do you think the self-service “trim tabs” are in the organization? Probably not IT. Getting the data to a state where self-service tools can pull it without in-depth knowledge of the underlying table structure is a massive undertaking beyond most businesspeople who simply want actionable data. Getting to that state is either the fruits of a lengthy ETL project, or collaboration at a very sophisticated level with IT in order to write accurate and precise queries.
    When the data is available, if you’re going against production data, the organization can’t run the risk of you dimming the lights with your runaway query so that critical transactions are slowed, but IT will take the heat if it happens. So your queries need to be reviewed by qualified staff and you can’t just wing it. Or you’re asking them to set up a reporting environment that mirrors production data. Either solution requires $$$ and resources – are you willing to pay your share of those? (If you’re writing dinky queries that don’t risk either of these situations then you don’t need BI!)
    The next problem is in getting your department’s data conversion on the priority list. Unless you are a customer-facing business unit with direct impact on revenue, or you are a central business service with clearly allocated costs, you’re going to be pretty far down the hit parade.

    In some cases your business unit (vs. IT resources) may have analysts who are capable of writing pretty good queries. Here I think Microsoft (and sometimes the PowerPivot team!) make erroneous assumptions, that those analysts want to move up the technology stack to SQL Server, SSA, and the whole alphabet soup of ETL to become data gurus. Nope. These analysts are on the business side and they want to move up to department administrators and managers, not learn the vagaries of Hadoop or delve into the Lawson or SAP data schema. They want and need business knowledge to ask the right questions and interpret the answers, not technical knowledge to extract the data. So when blogs tell them to move their data to ABC environment to take advantage of XYZ technology, they’re going to hit a wall. PowerPivot and PowerBI are great, but that’s where (I think) most of your self-service people are going to stop, especially the higher up you go in the food chain. So again I am concerned about your assumption that the data is available to these users at a skill level they can afford (in terms of time and their organizational goals) to attain.

    So by all means become a trim tab in your organization, but (to extend the metaphor) know where you need to point. No one group has the tools to complete the circuit from raw data to actionable intelligence, and IT is part of the solution. When you go to the managers and executives, show how you’re going to complete that circuit: prioritizing data extraction, selecting the correct reporting environment, managing queries appropriately, and which resources can do the job.

    1. What you describe here is exactly where good Enterprise BI does and should operate. i never said there wasn’t a place for this. I am just saying it is not the ONLY piece in the jigsaw. Once the data warehouse is up and running, there is added value that users can get by loading the cleansed and structured data into self service tools.

      1. I agree that our data utopias look similar. I don’t think that IT is the enemy to the extent you described. Lack of imagination afflicts all departments and it’s important to understand everyone’s motivations when trying to sell self-service.

        What I’ve found is that both IT and the responsible departments are focused on getting information IN to the transactional systems that have the data we want and the OUT portion has been limited to the canned reports necessary to satisfy basic management, auditors and stockholders. This will remain the primary responsibility of both areas – none of the A/B/C/D managers can afford to slow or interrupt the OLTP systems and any sales pitch that doesn’t take that into account will fail. The consequences for IT are especially severe so they’re just sensitive, not malicious.

        Readers of this blog, and your customers, probably have put more thought into what they want to accomplish with the data than their corporate peers including those in IT who may not have the bandwidth to consider advanced analytics when their antiquated servers are constantly throwing error messages or the network is hosed because staff are streaming the Olympics to their workstations.

        Our ERP team is very good – superb in fact – at bending the ERP to do almost anything we ask, and they often believe that if you give them requirements they can build an equally good solution within the tool. But they can miss the fact that we’re trying to reach people with the data who don’t generally use the ERP – why train them to use it and go thru multiple logins and menu options to get to the report when we could put a PowerBI report onto SharePoint? And as requirements change so frequently once you get people to see the data and ask the next question, and the next, then we’re involving 2 or 3 people in translating one set of requirements into analysis instead of 1 person making 3 sets of requirements come to life. That’s the kind of cost-effectiveness argument you can put in front of the team so that IT can put its effort into making more raw data available rather than tweaking the end result. It’s kind of RIcardo’s theory of competitive advantage within the teams . Where can people best contribute their skills?

        Your readers can probably come up with their own carrots instead of sticks for the groups they need to partner with. Make allies.

        1. I also think we are not that far apart, and given I have seen both sides, I can associate. I certainly didn’t use words like enemy or stick – I said influence, evangelist, demonstrate value, creating motivation etc. I have written a follow up article for next month where I expand on the “you get what you deserve” piece and what I call “battered IT Syndrome”, and that it is business’s responsibility to foster the right environment. Of course, it is just my opinion again, and others may disagree.

  3. “It is business’s responsibility to foster the right environment”

    Business managers operate within the framework of their KPI list, and this simply does not show up on their list. GMF may work in an innovative dream company, so his comments may apply to his environment, but not most others.

  4. I love the idea of being a trim tab in an enterprise company. I work for a medium size org that doesn’t have an ent solution, and is fully supporting what we are doing in PowerBI.. The only slowdown was being convinced of security in the cloud. The benefits are obvious right from the start. Even with power pivot reporting, i couldn’t keep a managers attention for more than a couple minutes. Frankly because 1. It was a spreadsheet, and 2. They needed me to slice and explain it (it wasn’t theres). PowerBI masks all that and hands it over to the end user. Being an Excel guy, its a relief to be able to smooth the road to insights and be able to adjust it by user.

    My advice would be to just start small.. Find that one or two insights a manager is looking at the hard way, and just hand that over as a dashboard (maybe just one table or matrix). Letting them know you aren’t trying to change how they do things, but you want to show them something that is an option, or accessible away from the office if they need it. Ive done this with the most old school sales person in our company.. And guess who now logs in to Power BI 😉

  5. Spin the wheel of time back 30 years and it is the same logical argument toward data center / IT applications versus new fangled PC applications. It is central versus decentral. In a way this is also the basis for much of the humor of the Dilbert cartoon……

  6. Trim Tab, no. A Servo Tab or Servo Rudder helps move the rudder to turn a vessel (aircraft or waterborne.)
    A trim tab sets a small correction to counteract other forces such as torque steer or cross wind/current and frees the pilot from having to hold such input manually.
    But this (incorrect definition of Trim Tab) demonstrates a difference often seen between Enterprise BI and Self Service BI. In Enterprise BI the builders and operators are often not the business minded individuals able to see ongoing faults or shortcomings in the BI system. IT doesn’t necessary understand the business jargon of finance; they have to keep up with enough knowledge already.
    With Self-service BI it is more likely a business minded person, more involved with the day-to-day transactions and needs of decision makers, is able to make ad-hoc BI pages or use other methods to point out anomalies, good, bad or indifferent.

    Ah, but the rub is communication and possession. The two “sides” remains the wrong way to look at it. Each side, IT and BPs (business people) both serve the company. Their specializations should be synergized (yuck, that word.)

    “I said influence, evangelist, demonstrate value, creating motivation etc. ” Exactly. Business people want information and with the volume growing they need help. PowerBI provides a far simpler method than EnterpriseBI. It represents an opportunity for IT to be more focused on the operating environments (OS, Hardware, network) and BP to take hold of the information and manipulate it as needed. Moreover, it represents an opportunity for these two “sides” to better understand the other.

    Approaching someone with the right attitude and you’ll be rewarded. Approach with the wrong attitude and you might get a “knuckle sandwich.”

  7. The simplest spectrum of SSBI is when BU can do some very simple adjustments to the dashboard delivered by IT. In that case with use of PowerBI, data can be easily questioned with use of natural language and dashboard layout can be redesigned to some extent. Yet, the value of such adjustments is very limited. The next advancement of SSBI will be when BU acquire PowerBI skills and start to build own dashboard based on own reports where advanced drill down options allow them to perform more detailed analytics without questioning IT for more detailed dashboard. Next BU can acquire DAX skills to build own measures and design them in PowerBI to ultimately become more IT advanced in BI. However, BU will be still business oriented and BU’s only goal would be to get the right conclusion from the data, and IT goal at the same time should be to deliver him the right data. All we have in between delivering right data and drawing right conclusions from the data is the scope of work to be a subject for self-service with use of PowerBI. p.s. Every BU has the possibility now to eat what is delivered by the cook or start cooking by himself and enjoy the work in kitchen. But is every kitchen open for new entrants?

  8. After 14 years dealing with data as a business user in a large corporation, I clearly can see that analytics are getting much better nowadays. While I still see new hires creating reports using MS Access end user aplications, I also started to see few people using PowerQuery and PowerPivot in the last 2 years but they will face the same issues I faced over 4 years which is standard Excel 32 bits. When more experienced users discovers they can get all data they need with PowerBI 64 bits they get exited again. Management get impressed with PowerBI visualizations and Q&A but now the problems are different as protection of information and lack of the enterprise gateway but the reality is that traditional BI is where the attention and a large budget goes. But as mentioned before, we have to keep going and advance over new concepts and eventually break the remaining barriers and effectvelly become a trim tab.

Leave a Comment or Question