The Politics of Analytics
Things have and haven’t changed.
Background
In a recent conversation with an old friend and analytics leader, I was reminded how much politics still exists in getting Data Analytics “functionalised” in an organisation. Her stakeholders were using their own “analytics” resources as the first-on-the-scene troops to extract, massage and report key metrics, instead of the official analytics resources from her team. Finance was at odds trying to justify the contributions of her data analytics team. While the data analytics community has obviously made very significant progress in getting C-suite leaders to recognise data analytics as an essential capability to compete in the modern economy, a fair many still give it lip service (so as not to appear as stone-age morons), while others try to bend the data analytics function to their will, seeing the obvious benefits of controlling information … and misinformation.
If you are in a management role in the data analytics function (this includes data science teams and various permutations of centres-of-excellence), how should you navigate this politics? And so I dedicate my 77th article to discussing this age-old issue, hoping to shed some light on possible navigation pathways.
(I write a weekly series of articles where I call out bad thinking and bad practices in data analytics / data science which you can find here.)
It’s All About Control
The desire to control is ubiquitous. With data analytics, it’s the desire to (a) control information, (b) control resources, and (c) control decision rights. Let’s unpack each one and see how we can navigate the politics around them.
Control of Information
Controlling the information is akin to controlling the narrative. It’s as simple as that. Stakeholders want to manage bad news, and good news, on their own time and terms. They also want to be able to corroborate the data and findings. Nothing is more frustrating than a non-stakeholder jumping to conclusions (because they mis-interpret the data) about the stakeholder’s performance. Given the infrastructure of modern organisations, controlling the information comes down to controlling the access to data. Stakeholders want access to their own data in the data warehouse / data lake, and they want their own supplementary stash of data that is not available to anyone else (typically extracted from process spreadsheets or directly from the operating systems).
How should an analytics leader navigate this politics? Run every piece of information with the stakeholder before it goes into general release. This builds trust; both ways. As the designated analytics leader, your objective is not to embarrass your stakeholder (unless it is your explicit intention to call bullshit on the stakeholder) but to support them to achieve their goals. But that doesn’t mean you should bury bad news, but rather, give analysis and recommendations to solve the bad news, in conjunction with presenting it. The stakeholder must directly experience that your analytics team is providing unique value that the stakeholder’s team cannot. If all the analytics team is doing is reporting, then it won’t do.
Control of Resources
The control of resources comes down to controlling the prioritisation of work / projects. I will be the first to admit that data analysts don’t always share the same priorities as their stakeholders; we go where the data leads us, and where the data says there is impact. The stakeholders may have divergent views on what they think is impactful … for them … and for the broader business. Controlling the (analytics) resources obviously allows the stakeholder to control the priorities of their deliverables.
The issue is that stakeholders, in general, are not data analytics experts, and don’t view deliverables as having the potential to create re-usable assets. And this is precisely where an analytics leader makes a strong, mutually beneficial case for the analytics resources not to be reporting into or beholden to the stakeholder. What the stakeholder loses in prioritisation control, they gain in speed and efficiency from the asset-oriented thinking.
Control of Decision Rights
With the ascent of data analytics as a recognised need and capability, many organisations have additionally imbued it with meaningful authority. For example, some organisations require their analytics leaders to sign off on outcome estimates, to sign off on pricing changes, etc. But stakeholders will fight tooth-and-nail before they relinquish their decision rights (i.e. approval authority) to someone else. Control on decision rights ultimately comes down to accountability and flexibility.
The savvy analytics leader can navigate this particularly nasty politics by stepping up and having skin in the game. Co-owning the stakeholder’s goals is a great way to start. Another way is to be held accountable for the outcomes of what you approve (or don’t approve).
Internal Wars
Another kind of politics that an analytics leader might find themselves embroiled in is the battle with their own kind. The internal wars. The data analytics function can sometimes be its own worst enemy. There is often tension between country, regional and global teams. This is about control as well. It’s about the control of thought leadership, the control of methodologies and processes, and the control of priorities.
Navigating the internal wars can often be more treacherous than the “stakeholder” wars. Promotions and bonuses are the weapons of the choice. They can be powerful weapons. Navigating the internal wars is tricky. It requires strategic thinking and planning. You need to gather your forces and allies. Regional and global teams typically have no operational capabilities, and so must rely on country teams for deliverables. Making yourself indispensable as a dependable delivery partner with your stakeholders creates a strong defensive bulwark that allows you to push back on your internal superiors.
Conclusion
Analytics leaders are not schooled in the “dark arts” of corporate politics. It is not in their personality and technical experience to engage in it. But as analytics leaders, they have an obligation to create an effective working environment for their teams. Playing the politics is inevitable. They can approach it like another form of pattern sensing; another problem-solving opportunity.
