A Huge Value Add for Data Scientists: Product Mindset
“I don’t want to be a product manager - the product org doesn’t offer me growth that I want.” My answer to this is “Great! That’s okay.” Since I started writing this newsletter, I’ve had numerous data scientists push back on the idea that data science needs to turn into data product.
However, I’ve never said data science (or data scientists) need entirely change. Nor do I think making everyone into a data product manager is the right approach. We’ve already seen the downsides of trying to cram everyone into a single bucket with the advent of the data science title causing too many to forget the critical role that data analysts play for business.
Data scientists who can think about and build data products are incredibly powerful (and well paid). No one needs to go through a formal title change. In fact, the idea of data product will be far more useful if we think about how to help others adopt aspects of it while not changing their current role. Imagine if you had data scientists, data analysts and larger organizations looking at data as a product to create sustainable value. That would be transformational.
Let’s be practical for a second - what does it really look like for data scientists to have a product mindset? After all, writing an article about this isn’t useful if it doesn’t enable people to think differently or change their behavior. Over the past 4 weeks, I’ve talked to many data scientists I know and asked them this exact question “where does a product mindset help you, or hinder you?” Here are 5 main takeaways.
Value for the business through value for the customer. A product mindset is all about creating value for the customer so that the customer does things beneficial for the business. In a couple of these conversations, we talked specifically about the value of platforms. An experimentation platform or ML platform itself might not have a direct ROI (in some cases they do), but it allows people to make higher quality decisions and innovate with product faster than they otherwise might have. Many of the products data scientists are responsible for don’t easily translate into revenue. The connection to revenue is through what they enable customers to do. If you create value for the customer (particularly internal ones), the business wins.
Building relationships with those building the product. I heard repeatedly that a product mindset for data scientists enabled something harder to quantify: better relationships with their product partners. Some of the most senior data scientists said that the key element in their success was the close partnership and trust they developed with product. Because they took the time to understand the tools and insights useful to their product partner, their product partner often was willing to dig more deeply into the challenges and constraints of data science. This mutually beneficial outcome came up enough times that I believe there’s something critical here.
Making tough tradeoff decisions. In a given day, data scientists have to make many decisions, in particular, how to solve a problem, how to build the right model, how to share it with other audiences. But one key issue I heard about frequently was “what problem to work on”. Most of us are familiar with quarterly priorities and how what is “important” often feels in flux. Data scientists who can think clearly about the product can also anticipate what data products will “move the needle” for product innovation. I think about this a lot in the context of product experimentation - we can do 50 different things. But what data product change will move the needle for the business? This is the hardest question.
Career trajectory. I’m not going to make a causal statement here, but the correlation feels fairly clear - those data scientists who count themselves as product obsessed tend to also have faster track careers and outcomes. Again, there is a lot of bias here, but imagine that you are sitting in a promotion or salary discussion and there are three people in front of you. Imagine that one of them has built relationships, understands the product and “gets” what data products move the needle for those in product, marketing and sales. I guarantee that person gets a serious look at promotion.
Keeping pace with the industry. I think every year I read about a new “gold standard” in AI and ML. Right around the time conferences publish their papers, I realize that I’m pretty much out of date with the most cutting edge techniques. If you’re purely focused on the techniques in data science, there’s zero doubt you’ll be overwhelmed. There’s too much to keep up with. But if you take the product focused mindset and ask “what is new, and are these new approaches going to be broadly useful for the business if we build them into our workflows or data products?”, then you’ve created a very useful filter for business value.
As data science continues to evolve, its value is realized more and more through sustainable data products. What makes those products sustainable? A connection to business and product value. If you are a data scientist, work with them, or want to become one, your product mindset (alongside your other skill sets) will be transformational in the coming years.
Thanks so much for reading and engaging. Have a wonderful weekend!