As we reviewed last week, building an impactful data, analytics, and AI capability requires a range of new roles and competencies. While the scope and responsibilites of roles such as data quality analyst or data architect are usually easy for the business to grasp, data product management can be a little bit trickier. Is this role not essentially a project management one? Although there are a number of similarities, there are important differences, as we will see.
I will caveat this by saying that the following represents my perspective and the way that I view data product management vs. traditional project management. My aim with data product management is to achieve maximum return on investment, customer satisfaction, and obviate the need of supporting roles such as business analysts and project managers. While this will not suit every organisation, these points will hopefully engender a good discussion.
Firstly, a project manager (PjM) is tasked with completing a set of activities with defined constraints, usually scope, cost, and time. This results in an output of a certain quality. A data product manager (DPM) instead is tasked with developing a data product (e.g. an algorithm, or data set) that delivers value to the customer over an extended period of time. If it costs more, takes longer, or if the scope changes – this is fine if the customer is happy at the end.
Secondly, the DPM is a leadership role, effectively acting as an entrepreneur-in-residence. While part of a larger team, they have the autonomy to identify opportunities in the business, define a plan to realise these opportunities, and to mobilise resources both inside and outside the team to deliver on their plans. While this does not mean that there is no prioritisation that takes place, it does mean that a DPM has far more autonomy than a traditional PjM.
Thirdly, while a PjM tends to engage solely with project stakeholders, a DPM is expected to have much broader relationships with the business. This enables them to act as trusted advisers to senior leaders, guiding executives through the complexity of the data and technology landscape. In this way the DPM plays a role not unlike a business analyst or a business partner, but one who is genuinely empowered to mobilise resources to capitalise on the opportunity.
Fourthly, given the DPM's relationships both inside the data and analytics team and outside it, they are the source of truth for priotisation. Put bluntly, as tempting as it might be to tell a DPM what to work on, any data leader worth their salt knows that as the voice of the customer, it is your job to listen to what the DPM has to say. They should be the compass that guarantees the customer-centricity and long-term success of your team and its data products.
Fifthly, not having a PjM promotes accountability. I once worked on a simple project where multiple team members told me they could not get started until the PjM was recruited. Why not I asked? Oh, because "we need the PjM to draw up a project plan (read: Gantt chart) first." Mind you, this was for a two-week project! Eliminating the PjM role means everyone understands they are responsible for managing their work – the DPM simply orchestrates them.
Finally, introducing the DPM role helps shift the mindset of the team from projects to products. Instead of thinking of WHAT needs to delivered, think of HOW you can generate the most impact in the long term. Success in the PjM role is sadly often defined as delivering a project according to specifications, even if six months later it gets abandoned. Product management is about building intellectual property that creates sustainable value for years to come.
There are a lot of other things that can be said on this topic, but I will keep my powder dry for a future debate, perhaps a panel at a conference? That said I hope that for those of you that do not yet have DPM roles in your team, you seriously consider introducing them as a way of increasing productivity and customer impact. While it is not easy to build a high-performing product management culture, the results will likely positively surprise you.