Not all stakeholders stand on equal ground. Discover which stakeholders should take top precedence.
As a management consultant, my clients are constantly asking for my help in managing their stakeholders. It seems like the bigger the company, the more people come out of the woodwork with requirements on how you should build a solution. This can be quite frustrating, because oftentimes two stakeholders have diametrically opposite requirements.
How can you make everyone happy at the same time? You can't, which is why you must protect your sanity by prioritizing stakeholders. On a data science effort, it's important to have clear agreements with your different stakeholders because not all them are created equal.
The "golden rule"
My best advice for stakeholder management is to follow the "golden rule": he who has the gold makes the rules. Your primary stakeholder must be your project sponsor — that is, the one who's paying for everything. This is the person (and by extension the department or function) who has the overriding say on what happens on the project.
I was on a critical compliance effort with the federal arm of a large technology firm where an audit turned into an investigation. Even though the federal division had the most context and impact, it had no money. The VP of Sales and Marketing funded the effort, because 1) he would derive residual benefit from the effort and, more importantly, 2) he had money. This is a dysfunctional stakeholder arrangement that you should stay away from.
The primary benefactor of your analytic solution should also be the one paying for it. If Marketing wants better insights into a few key markets, it makes sense for the VP of Marketing to sponsor an effort where your data scientists develop a predictive analytics engine for them. In this case, make it clear to all other stakeholders that Marketing has the biggest stake in the effort so they get the most votes.
Stay within the lines
The second-tier stakeholders are the ones who have a legitimate regulatory stake in your project. Back in the early 2000s, I was involved in a number of Sarbanes-Oxley (SOX) compliance efforts. And although it seemed to be a somewhat bloated and over-hyped market for consultants, SOX compliance is no joke for top management. If not handled properly, the CEO and/or CFO could go to jail. With this kind of visibility, it's no mystery why many of us were helping large companies obtain and maintain compliance.
When there are legitimate regulatory concerns, the stakeholders impacted by these concerns should have their appropriate priority level. I say legitimate, because this principle is often abused. Many times you'll find stakeholders playing the regulatory card against corporate standards or even guidelines and best practices. I'm not saying that corporate standards shouldn't be followed; I am saying that people holding this stake shouldn't share the same status on your project as those representing a real government concern, never mind the person paying for everything.
The also-rans are the rest of the stakeholders, and they take the lowest priority. In a sports race, there's the winner and the first-runner up, and the rest are called also-rans — the same applies with stakeholders. If a stakeholder isn't paying for the project and doesn't have a legitimate regulatory concern, their requirements must take a backseat. Most likely, you'll have a lot of also-rans on the same project, including IT, Finance, HR, and R&D. You may be concerned about how to prioritize the also-rans, but in practice their voice is typically irrelevant once the primary stakeholder voices their direction.
You cannot let the also-rans push you off the ledge. Unfortunately, it's usually the stakeholders with no money that cry the loudest — don't endorse this on your projects. Make it very clear who owns the requirements (the sponsor) and whose requirements we must comply with (those with a legitimate regulatory concern). All others can submit requirements for consideration; however, don't feel obligated to make them happy. That's a recipe for insanity.
Stakeholder management is a key element in any successful project, and data science is no exception. The key to stakeholder management is realizing that not all stakeholders stand on equal ground. Give the most attention to the stakeholder paying for your project and thoughtful consideration to those who have a regulatory concern. Everyone else has nothing more than a wish list that you'll document and accommodate — resources permitting. Stop the madness now!