ADPList Product-led Growth: Build Better: Guide To Optimizing Product-Led Growth

Some of the mistakes that product teams make that harm PLG:

- Underestimate strategy (important to understand your work, how your role is solving an issue, consistency in strategy, key to align with teams about goals, rushing to build solutions is a no-no).

This can be avoided by Understand (analysis and theory).Identify(putting in the roadmap and prioritzing).Execute (Jumping to solutions) <-- Missing one of these will be a disaster.

- Optimizing in silos. Every team should be thinking about growth.

- Pushing product without PMF.  If product does not have retention, you lose them. It is harder to get them back. Avoid this by pivoting to the right technique.

- If you want to fail don't revisit prior assumptions. Do not worry about previous wins and losses since they may not be applicable now anymore since times change and tech evolves. Avoid this by holding out (A hold-out group is a form of cross-validation that extracts, or “holds out,” one set of users from testing) 

- Blaming users for not using the product. Building product is just half the battle. Perhaps the user did not find it? Acquisition problem. It is all about distribution. Have humility, think about marketing, entry points, funnels. 

- Optimising the wrong thing. Vanity metrics -> you are improving a metric but no underlying benefit. Obsessing only on metrics eg clicks and not thinking from user's goals

- No analytics. Without it, difficult to justify the impact of your work. More advanced you are, more analytics are available. Team should have access to metrics. 

- Picking the wrong time to iterate. Sometimes a small change may not make that impact versus a big change. Users might benefit from that impact. Right-sizing effort.

- Not talking to customers. Happens so often. If there is old research, use that. What are the ongoing surveys that are going on? Does the product team know when they are losing/winning from the sales team? Ideally, they should be aware of that. 

- Test badly (If you are testing and you haven't learnt anything then that qualifies as a bad test, testing without enough power - consider many users, testing without enough time - stopping when the test succeeds, bad audience selection, have counter-metrics, testing as an afterthought is a no-no). Avoid this by critical thinking and data science