Data Product and Team Building at Motiva AI

In 2016, Motiva AI spun out of the venture studio started by David and Chris, The Data Guild, and many of our core team members moved over to Motiva from the Guild (David, Chris, Matt, Paul, and Jami). The Data Guild’s mission-driven philosophy of tackling global challenges is very much baked into Motiva’s DNA, along with a particular product development process and approach to building world-class data product teams.

Data products rely on some mix of underlying data assets and UX, and can include machine learning components or not. They differ from other software products in some respects. They are generally more sensitive to underlying data sourcing and quality needs, may require specialized processing to make those assets viable, require careful attention to when and how human users interact with the product, and often require a deep understanding of the larger landscape of modeling approaches in order to meet the product’s design goals. 

These requirements shape our approach to data product development. Motiva’s approach generally follows the Guild’s, a process honed under partnerships with and taught to product teams at large multinationals. The basic approach is:

  1. Understand the human needs, constraints, and behavioral context
  2. Define the data landscape that supports possible product solutions. Do we have the right data, how good is it, are there opportunities to create data moats, etc. 
  3. Rapidly iterate with key persona proxies to validate pain/need, priority, market potential, and technical viability(data, modeling approaches) – prior to dev commit. Often involves testing ideas out with current customers, in services engagements, etc.
  4. Fullish-stack MVP including data solution (aggregation, normalizing, cleaning), just enough product engineering and design to be functional. Go/no-go with feedback from customers and stakeholders.
  5. Ship v.01 quietly and iterate quickly on early versions. 
  6. Scale as appropriate.

Building data products involves at a minimum data engineering, product, strategy, and data science team members – a bit like an A-Team: small, diverse, and very fast. As the candidate product gets validated and matures, the team grows to include other skills – design, core engineering, additional data engineering, etc. Designing the right kind of data product team is key to ensuring this process succeeds.  We hire highly talented people like everyone else, but we put a premium on the following characteristics.

  • T-shaped. We value people who are both broadly curious about the world and deeply fluent in a particular domain (hence “T”). The ability to find and share patterns and similarities outside of one’s own expert area is critical to effective product building, regardless of your role. This fits very few people in general.
  • Deep empathy. We value those who are able to deeply understand others – whether those are other team members or customers or partners. We become better collaborators and problem-solvers by cultivating empathy. 
  • Connectors. We seek people who can develop deep relationships with others. They can be introverts or extroverts, but the ability to build and maintain human relationships is key. This ability is often related to the above.
  • Driven to serve the world. Our team is motivated by solving global problems. The more profound these are, the more we get excited. 
  • Humility. There are lots of smart, talented people in the world, and there’s usually someone smarter than you. If you can’t accept that, you’re going to have a tough time listening to others, developing healthy relationships, or understanding others. 
  • Resilience. How well do you handle adversity? We all struggle at different times; how do you bounce back and how do you help others stay strong?
  • Technical excellence. Goes without saying, but we look for exceptionally strong skills/experience (or evidence you can develop these) in your field of expertise.  
  • Creativity. We look for people who live between worlds, can see things others don’t, and bring new ideas to life. This quality can be expressed in many different ways, but ultimately tells us whether and how you deliver. 

Building successful data products is much more likely when team members share these qualities. In addition, Motiva has been fully remote from Day 1, and we’ve developed a culture and set of practices that enable everyone to become the best they can be.