Data Scientist to Data Product Manager: How to Find a Superstar

Ever since I started writing about data product and data product management, I’ve had great discussions about exactly how to find a data product manager. For some, they already have the title. For others, they might already be technical product managers in a data adjacent space. But in many cases, the best data product managers are sitting right in front of us as data scientists, but we just don’t know what to look for in evaluating potential. Yet the key challenge remains that while companies want to invest in data product, it is often not clear exactly where they can find suitable candidates. While data product managers can come in many forms, today I want to focus on the data scientist to data product manager pipeline.

Put yourself in a hiring manager’s shoes. Your organization has made a commitment to managing data science products. Its clear you can’t ask someone to do this part-time. You have support to go hire. Then the first meeting with the recruiter comes and they ask you “tell me about who we should be looking for and what skills they have.” It is not unusual to have a deer in the headlights moment. There’s no obvious profile that’s perfect. Existing titles are only so helpful. Here are some ways that I’ve approached this question as a hiring manager, far before I ever get to that conversation with the recruiter.

1. Know who and what you have internally. One of the biggest mistakes we make is assuming that the perfect candidate exists outside the physical (or virtual) walls of our company just because we don’t have the role in its current form. However, my belief is that if you have strong data scientists who are motivated to work with product and focus on strategy, you likely have some superstars waiting in the wings, you just need to work with them to identify the describe the opportunity. If you see data scientists that naturally focus on building healthy products, that care about user input and seek to make life better for internal users and who want to improve the organization’s ability to broadly use data, you’ve already got a good list of potential candidates. Start here, discuss with this group, sell the opportunity. Do this before you ever get into looking externally.

2. For internal candidates, understand what they have (and what they’d need to develop). The upside to potential candidates that are current data scientists is that you probably know their strengths and areas of opportunity. You don’t need to interview to get signal, you have their work product. This also means you probably realize no one is “perfect” for a data product role, simply because they haven’t done all of it before. That’s okay! Investing in your people internally means that you can lean on what they are good at, and provide time and structure to come up to speed on areas in which they need growth. Don’t let one concern block you from taking a chance.

3. Be very clear on what data products need ownership. This applies to both internal and external candidates. Data products don’t exist in a vacuum. Data scientists have different skillsets and areas of comfort, and it is important to ensure you match that domain expertise with your key areas of need. This is not only helpful for evaluating internal candidates, but for helping narrow the potential matches of external data scientists. If you advertise for a data product manager without specific filters or requirements, you’ll have hundreds to thousands of applications. Your recruiting team (or just you) will be faced with an overwhelming sense of dread in navigating the external market. My belief is that success in a data product role is not a commodity - domain knowledge related to the product is necessary.

4. Evaluate product sense and customer mindset before technical skills. For a complex role like data product, it is important that the initial conversations and interviews are high information and only a few candidates pass through them. Otherwise the process will quickly become overwhelming. Remember, you aren’t interviewing for a data science role. The rarest skillset is someone with a data science background who also excels with product sense and working with customers. That means initial conversations should go heavy on product sense and customer engagement. Hold a high bar here. This might mean you bring in your product counterparts to help draft the interview or even help you navigate evaluation criteria. That’s okay - don’t try to do this alone!

5. Even with the right skills and product sense, development will be important. Whether you hire someone internally or externally, you should expect that they won’t come in ready to crush it on day one. Superstars in this space have a learning curve in at least one or two areas. What makes them stand out is their ability to recognize those areas of opportunity and addressing them quickly. I think if you wait to hire someone until you know they will succeed without a doubt, you’ll likely never hire someone for the role. It is always a bit of a risk, but using some of the filters, things to look for and qualities above, I think you can find that potentially transformative data product manager.

---