Data Partnerships and Agreements

Returning to some concepts Lan Nguyen and I discussed a few weeks back, today, I want to dive into data partnerships.

Why should you care? In a world where Large Language Models (LLMs) dominate, all things data will become important for all PMs, no matter their role. I would argue it was the case before November 2022, but it is become more evident as more and more products rely on data partnerships.

If you are being asked about your experiences with data partnerships/agreements, there are 4 things to keep in mind.

Before I dive into the list, I was to illustrate some examples where this might matter:

• Amazon working with a supplier

• Google asking a Retailer to commit to supporting a rollout

• Uber working with Retailers for deliveries

The 4 Concepts

1. Baseline on Risks

   • Shared understanding is key. Predictions will be wrong to start.

2. Agree to worst-case scenario planning

   1. If you have 10 people in a room, you probably have a minimum of 5 different ideas of what a worse-case scenario really looks like. Talk about it as you plan.

3. Communicate Constantly

   1. At Amazon, we called it Hits, Misses, and Learnings. Share them early and often.

4. Scaling: Set agreement with more firm plans

   1. As you start to solidify your learnings, agree to more acceptable limits and shared risks.

Let’s dive in and assume the case of Uber working with say Macy’s for clothing deliveries. As soon as we start talking about delivering clothing, we have both inventory control issues as well as predictive purchasing agreements that need to be considered. If Uber invests in a big advertising push or overpredicts and Macy’s is left holding unusable inventory, the word will get out and no one will be willing to join in as an additional partner.

I hope you get the picture. Sometimes we in the tech world get soo focused on selling technology we forget about the risks in the physical goods space. Next time you interview, if you get a data question, keep these concepts in mind.