The Unknowns in Product Development by Spotify Director of Product, Flavia Neves

Working with uncertainty is being in a situation that requires a decision, for which you have significantly less information than you need to make the decision obvious to all participants. No product decision guarantees any results for sure, because market users have too many moving computers for anyone to express the logic of all the dynamics that are taking place. There is no single plan that guarantees good results, but there are some strategies that can increase the chances of creating great technology products.

1. Innovation comes with a ton of uncertainty.

Move quickly, learn fast, build to explore.

Get a rough iteration in front of users ASAP to start collecting quality inputs.

You need to know what your problem is and what goal you want to achieve in order to solve it. This is where the quality of the research matters. Start by using the data to formulate a problem, define your goals, create a hypothesis to test. Don't waste time on very complex analysis, planning all the details and developing a perfect plan created just for study.

Thinking should be directed to the creation of a correct hypothesis, to the creation of a goal. Make it crystal clear to everyone around you.

Accept the uncertainty and the long odds.

Moving fast and creating a flow of high quality materials will help you pave the way forward. But this does not make the path clear, but simply helps you navigate and move forward. Therefore, from the first day, accept uncertainty and great chances.

It is very easy to be tempted to organize, plan, classify in order to feel secure and in control. But this is unlikely to change the odds and give you less time to repeat. Most of the time, the path is unclear and the path to a successful product is through a bunch of failed tests and iterations that will help you move forward. Repeat based on your knowledge and take one step at a time.

Be the cheerleader.

It`s our job as PMs to keep everyone motivated and with the eyes on the future.

It is difficult to make decisions in uncertainty. Therefore, it is important to keep the team motivated during this time.

2. Data is great, just don`t go overboard!

It is important to find the right balance between the two extremes: making decisions based on intuition and fear that for some reason there is no data to work with.

The mindset

When you look at data, the goal should be to prove your assumptions wrong.

Data shopping isn`t cool!

Don`t stretch the numbers to tell a story and resist to the temptation of jumping to conclusions.

The solution to the problem of uncertainty is to make more discoveries, to dig deeper, to pave the way for knowledge that will lead to strategic decisions. Don't look for patterns where there aren't any, and don't jump to conclusions just because it makes our lives easier. It's important to have an idea of when you need to repeat an experiment or when to keep experimenting in order to make decisions that will ultimately stand the test of time.

Take calculated risks.

It won`t make you popular but learn to say “we don`t have enough data to answer that“.

Try answering questions you don't feel comfortable answering: "I'll sort this out" or "I'll get back to you with an answer or a plan for how I'm going to get that answer." Formulate it as a hypothesis and how would you test it. You have to develop product intuition, use your intuition throughout your career.

3. Dance to the tune and never forget the end goal.

Product management is a cross between paradigm-breaking changes and the optimization or enhancement of existing offerings. What we do and in which direction we invest, more or less, depends on the ultimate goal.

The north star is your best friend

Having a north star you can fall back on will help you make the toughest decisions.

If you have a polar star or a good strategy to guide you, you will probably make a better decision or more good decisions than bad ones. Regardless of the situation or what you want to achieve, you have several different options.

It worked once...double down!

Use previous iterations to guide your future optimizations.

If you've found an underused area where there's plenty of room to explore, and you already have some knowledge that can help you make better decisions, redouble your efforts. This will help you prioritize. It directs your attention only to small, insignificant improvements. Your goal is to optimize. Develop under-optimized areas.

Is it time for a paradigm shift?

Sometimes the signs point towards a paradigm shift so make sure you don`t overlook them.

Signs that a paradigm shift is needed:

  • Your positive impact becomes less frequent or small increments decrease over time.

  • Most of the experiments that you conduct lead to significant negative consequences.

  • It becomes very difficult to make discoveries and come up with hypotheses that don't require a lot of user experience rewriting.

Nobody likes a frankenstein product.

Small iterations are preferable but the new proposed state needs to be functional.

Flavia Neves, Spotify Director of Product