The previous article introduced the idea of the “Holy Trinity” : the three key characteristics of analytics sponsors. These go beyond having budget and mandate to perform analytics : while those two raise an individual to the title of “sponsor”, the Trinity determines whether the sponsor is a good one. The “goodness” of a sponsor is defined by their analytics function delivering actual and recognized value, and thriving on those terms.

The Trinity consists of Appropriate Understanding, Appropriate Empowerment and Appropriate Incentive. The current series of articles explores each of these. We will examine what success or failure of each element looks like. We will also explore the cases where only one element of the Trinity is present, and, the direst of all, which is total Trinity failure.

For each element, we first examine the case where the sponsor has the entire Trinity in place, but we focus our attention on the element in question. This will be referred to as the “Success Mode” of that element. It will describe why that element of the Trinity is so important, playing well with the other two. We then examine the “Failure Mode”, the situation where the element in question is missing, even as the other two are in place. We then switch to the element’s “Isolation Failure” mode, which is the case where this element is the only one present, and the other two absent. Finally, after listing these for all three elements, there will be an account of “Total Failure”, where all three elements are absent.

Trinity Element I : Appropriate Understanding

Success Mode

Successful understanding means that the sponsor knows what to do in order to create to create, support, protect, nurture and grow an effective analytics function. That sponsor can evaluate recommendations and pitches from consultants, vendors and internal stakeholders to the analytics function, and make effective decisions to further the growth and success of the function.

Such a sponsor understands the importance of both effective IT support and IT non-interference in the analytics function. He understands IT’s role in the provision of sandpit environments, and easy access to open source and commodity tools and all relevant data. He also understands that once data is provided and systems are in place, IT’s main role in analytics is to get the heck out of the way.

The understanding sponsor can manage their analytics team, understand issues raised and recommendations from analytics team leaders and can direct those team leaders effectively to achieve required results.

The understanding sponsor of a strategic analytics function is their number one client as well as a thoughtful, reflective and demanding consumer of their analytics product. He understands that decision support is not decision replacement, and that he has a vital value add to the process, which is to make raw information actionable. He understands that good BI makes decisions better, but not easier. Indeed, good BI is voraciously consumed by good decision makers, even as it is rejected by poor ones as “not actionable”. He actively builds growing support and demand for BI product among his peers, and drives a culture of objectivity, empiricism and accountability within the businesses.

The understanding sponsor of an operational analytics function realises that operational analytics is difficult, and that there are no shortcuts to key components, regardless of what software vendors may say as they beat at his door, and those of his superiors, as well as the CIO’s. He knows that data must be cleaned, processed, prepared and no magic tool does even 50% of that. He knows that there are human components to the operational value chain, from data collectors at the coalface, to IT/DWH as data providers / data bottlenecks, to human executors of analytics-driven operational directives. These people need to be won over or otherwise directed to operate as a smooth, flawless machine, otherwise the benefits are not realized and analytics often takes the blame. He realises the need for appropriate measurement of effectiveness, and the frequent absence of this as applied to the analytics-free status quo. He realises the need to decouple measurement of effectiveness from analytics itself in the eyes of less understanding executive peers and stakeholders.

Finally, the sponsor in the know understands the potential consequences of successful analytics. He knows that an objective performance management culture, and a strong decision support culture favours proven performers and intelligent decision makers, even as it exposes sophists, credit takers and artful persuaders. He realises the cascade effect this can have on the entire executive class, and spillover to the board, shareholders or equivalent stakeholders in government or NGOs. He also understands the expected subtle efforts to derail analytics for precisely these reasons, and knows ways to counter them.

This sponsor is a very rare beast to say the least, but they do exist, their teams thrive and their organisations reap the benefits of analytics.

Failure Mode

This is the case where the sponsor has all the best intentions, at least as far as he understands analytics, and the power to make the function work, if only he knew what that entailed. Unfortunately, in this case, it is lack of understanding which lets analytics down.

This failure mode is more common in tech startups and small privately owned companies where the sponsor is the owner, and thus has all the best incentives and mandate to act, but nevertheless gets lost as to where analytics actually fits, how it could help, and what might be required from the sponsor to make sure that analytics delivers value .
The most common gap in understanding in small owner-managed companies is the commonly held view that analytics is part of IT and resembles it in skills, focus and practice. The fallacy that analtyics is IT also helps in throwing analytics acquisition in with the broader IT acquisition stack, with strong influence from the CIO, resulting in unhelpful IT management and practice methods applied to analytics, usually staffed by people chosen for their IT-ish skills, and spending most of their time doing IT-ish things like coding. The analytics is IT fallacy is not helped by those software vendors who are all too happy to perpetuate it, the better to get people to spend money unwisely.

Even more fundamental problems can arise when executives or business owners cannot grasp the difference between “technical” (esoteric detail best left to specialists) and “strategic” (important issues for the executives themselves that cannot and should not be outsourced or delegated). All too often, anything that is not understood, and anything that required painfully rigorous thinking as analytics does, is relegated to the “technical” bucket, even when the issue is actually of utmost strategic importance. Important questions like “what kind of decisions do you want this report to support?” or “are you really asking for a forecast, or it is more like our agreed targets ?” or “what do you want to do with customer segments?” are often met with puzzled, impatient stares and the questioner relegated to the technical bucket along with the questions.

My analogy here is cars, particularly taxis. The construction and repair of a car is clearly technical. What about driving skills ? These a higher order skills, but still, these can be outsourced to a taxi driver. Now consider the situation where the executive climbs into a taxi, and the driver asks “where do you want to go?”. Now imagine an incredulous executive saying “how would I know ? I know nothing about cars. don’t bother me with technical detail. This is something that you should be taking care of. And above all, make sure you make me look good”.

Ridiculous as this analogy sounds, it is a good picture of what happens when the sponsor of analytics suffers a catastrophic failure in understanding. In this case, they “make analytics happen”, but aren’t entirely clear why or how.They put the people and software in place, perhaps with some very vague directives, and expect the ill-defined “analytics thing” to happen, whatever that may be. The failure of understating goes beyond not knowing what the “analytics thing” is, to not realizing that that knowing this could perhaps be useful, let alone vital. Most vital knowledge that the sponsor should have is an “unknown unknown”. The only upside in his case is that the sponsor is happy, confident and unperturbed, unaware that anything should be wrong. If you count that as upside.

Another symptom of a failure in understanding is an eagerness to reach for magic solutions and “best practice”, as promised by certain software vendors and consultants. The belief that analytics is IT helps vendor business models that prey on waste and ingnorance. If an executive, unaware of what they really need, is willing to spend millions on “analytics in a box”, that is just fine with the software company. If an executive wants “analytics best practices” put on place by junior process workers, or predictive modelling offshored to Cheapworkerstan, there is always a vendor ready to collect the money. Such a vendor may be quite indifferent to any debacle of error, waste, stagnation and failure that may emerge years later. Even more likely, the vendor is mot concerned that money would have been better spent on good people, that much difficult data plumbing work is in any case unavoidable and not helped by million dollar software and that free software would have been good enough to begin with. The understanding gap is certainly helped by an incentive gap when it comes to spending money on all the wrong things.

Extending the taxi analogy, failure in understanding often reaches reflexively for “best practice”. Not many people catch taxis asking to be taken to a “best practice” destination. Doing this with analytics is usually just as inappropriate and downright surreal, although it happens much more commonly. It helps that taxi drivers don’t usually encourage this kind of behaviour. Consultants and vendors however are often less shy.

Isolation Failure

The remaining issue to consider is the opposite failure mode. This is the situation where understanding is present, while incentive and empowerment are not. What happens if the sponsor has a very good idea of how to make analytics work, but no real interest in doing so, and no real mandate even if they did ?

Often, the lack of mandate is the very thing driving the lack of incentive. Sometimes there are other agendas – understanding analytics can be precisely the reason to undermine or derail it : after all, analytics makes people accountable, possibly obsolete and forces them to operate in a complex, ever-changing world. Some might think that it is best to kill it, and most likely this is done by the one person that knows that analytics is more than some ill-defined buzzword. Often, killing or derailing something like analytics is far easier than nurturing and growing it, so all it takes is a bit of understanding of what analytics can bring to people’s careers and accountability, along with very little empowerment and all the wrong incentives, arising from being the kind of worker that would not cheer for an analytics-empowered world. It would be naive and false to say that there aren’t such people or roles within organisations, however “negative” this truth may be.

Other than that, what is most likely to happen when understanding in supply but incentive and empowerment are not ? The answer is, usually nothing at all. Lack of incentive need not mean a destructive attitude to analytics, it merely means that the are other priorities, and given no empowerment, there is little bandwidth to meet them. So analytics languishes, if it exists at all. Perhaps a single analyst or small team is hired as an afterthought, their activities uncertain and their morale low. Data acquisition has to be painfully negotiated with IT and other stakeholders on an ongoing basis. IT has a very unhelpful say in what systems, tools, process and skills are in place. The team performs at best a rudimentary ad-hoc BI function, at those rare moments when someone actually cares about what is in the data. Most such reports are generated for compliance and similar external reporting. The one upside is that usually when the is no empowerment or incentive, the team finds itself using open source tools. This is not really an upside, nor anything resembling an Analyst First operation. Such functions can sometimes be found in smaller government agencies or NGO. They are particularly common in QUANGOs. Sometimes they are surprisingly well funded too. Interestingly, these functions can survive for years. These are often the people telling me sob stories at conferences. I usually tell them to get a new job.

The greatest opportunity for analytics is in privately owned, owner-managed organisations where understanding is the one missing element of the Trinity, and great value can be realized once this gap is closed. Even better, this sector is not, as great an opportunity to those software vendors who prey on ignorance, which is just the absence of appropriate understanding.

As a final note, it pays to remember that failure in sponsor understanding is the one easiest to fix, although “easiest” is not the same as “easy”. Perhaps a better word is “feasible”, whereas failures in incentive are impossible to fix, and failures in empowerment practically so as well. A sufficiently incentivised and empowered sponsor can and should educate themselves, and make that education a key part of the creation of the new analytics function. Hopefully, they understand at least enough to prioritise improving their understanding. I have been privileged to assist a number of sponsors in precisely this activity, with very satisfying results. Indeed, the bridging of the sponsor’s understanding gap can and should be the first step of any new analytics function.

This was but the first of three essays, the next one will explore failures in Incentive, which are the most damaging and irreversible of all.