I’ll let you in on a secret. There’s an easy way you can assess how good a product trio is at discovery. Just ask them, “what did you learn and decide last week?” If the product trio is vague — “we spent the week crafting strategy”, “we learnt [insert something obvious here] and tweaked this design” — you might want to check for shells on their back (cause they be moving like snails).

Empowered product t should be learning every week. And with those learnings, they need to make good product decisions. It’s the most transformative principle of Teresa Torres’s book: Continuous Discovery Habits. As she defines it:

Continuous discovery is: weekly touch points with customers, by the team building the product, where they conduct small research activities, in pursuit of a desired product outcome.

When I first read this statement, I didn’t understand its power. I saw the value in testing assumptions with users regularly but missed the bigger picture. But I eventually found that bigger picture in those words: “in pursuit of a desired product outcome”.

We do research ultimately to make good product decisions.

We’re constantly looking to make decisions on two (often linked) questions:

—1. What customer problem do we focus on solving

—2. What solution is most likely to solve that problem*

*in a way that is desirable, feasible, and viable

When you elevate good decision-making to be the most crucial goal of any week, it clarifies every activity. Okay, you crafted a beautiful opportunity map based on insights from user testing, but what did you decide to based on that visualisation? Where are you prioritising your focus now? And what research will you run to evolve your understanding of the most valuable opportunities and update it again?

At Gousto, each product trio aims to learn and make a valuable product decision at least every week. And our ambition is to narrow these cycle times even further to days. Here’s how we think about moving fast.

Do almost everything to 70% fidelity

Too often, product teams I’ve worked with have over-sweated the detail when first investigating a problem space. We’ve spent hours arguing over how to phrase our job stories or conducting additional 6–7 user interviews to build out the edges of an opportunity tree. These pursuits often feel good, like you’re doing “proper customer-centric discovery” and becoming an expert in the field. But these activities come with a tremendous cost — time.

You do need to plan research and think about the broader opportunity space. But be mindful of over-planning and over-relying on general research instead of asking targeted questions, prototyping and live testing.

The first hour of planning/opportunity mapping after 6–12 user interviews and data analysis will often create 80% of the total “decision-making value” for where to focus.

“Most decisions should probably be made with somewhere around 70 percent of the information you wish you had, if you wait for 90 percent, in most cases, you’re probably being slow.” — Jeff Bezos

Instead of waiting for the perfect information set, set out your key assumptions that a decision rests on and validate them as you go. Don’t be academic. Think critically about data, but move at speed.

When we first began investigating how to improve our menu experience for signups, we knew very little about the problem space. We set ourselves a goal to get to validated and refined solutions that solved the most significant signup problems in less than a month.

In the first week, we ran unmoderated user testing with our website and competitors, investigated key behaviours on the signup menu, and collated insights into an opportunity tree.

We then debated what to do next. Some argued that we should do more user testing to flesh out the tree further and build confidence in our current opportunities. But we wanted to move fast. So we synthesised evidence for each opportunity and force-prioritised them.

This prioritisation led to a robust discussion on the evidence behind the ‘traffic lights’, after which the team agreed to focus on just two problem statements to validate and understand better in our next round of interviews. We also brought prototypes to those interviews to begin testing assumptions for how to solve those problems. By ignoring the other opportunities early, we achieved a much deeper understanding of an area that was most likely to create value (sounds obvious, but it’s often so hard for teams to do!).

Make research goals super clear before starting

When moving at speed, it’s easy not to step back and question what you’re doing. You can conduct what feels like interesting research, only to realise too late that what you were trying to answer has ceased to be relevant. To avoid this pitfall, we need to move intentionally at every step in our discovery.

Deciding what you want to learn is one of the highest leverage activities you can do. Spending a few high-quality hours a week documenting and debating what assumptions are most important to test pays off tenfold as you spend the rest of the week focussed only on research that matters.

It doesn’t take much effort to point your flashlight in a different direction, but that may be the difference between spotting a vast ‘monster’ of a risk from tanking the success of a new product or feature.

There’s no one best way to create a research plan, but at Gousto, we think about and discuss three things.

— 1. What we’re trying to learn: research goals and assumptions to test

— 2. How will those learnings inform decisions

— 3. Why are those decisions important

To navigate and answer questions 1–3, you’ll likely need to have thought about your most important assumptions and defined a product strategy. I plan to share thoughts on product strategy soon. But if you want to dive into some good material early, read this excellent summary of Rumelt’s Good Strategy, Bad Strategy, which is as applicable to product strategy as it is to broader business strategy.

A great recent example of super targeted research was the investigation two of our all-star designers Kelly & Elsa conducted into our ‘customisation’ experience on our menu (the capability for users to swap proteins for different ones, like meat for veggie). We had strong reason to believe the XP was suboptimal, meaning fewer people were using the feature. If we could get more people to understand and use our customisation feature, we’d increase users’ experience of our broad range of recipes and improve retention.

Our product team documented the why, key outcomes, and research goals. They then articulated a set of key assumptions that we’d prove or disprove via user testing. Each assumption, if confirmed, would lead to actionable solution design off the back of it. You can see one below, including the findings and next steps.

With clear research goals and assumptions, they focused on answering only the most essential questions. And within a week, we had robust evidence on the main customer pain points and could move forward at pace to create solutions for them.

Move fast, but don’t forget to step back from time to time

A week-to-week cadence of doing focussed research and making big decisions can feel uncomfortably fast. And without a doubt, you’ll make (avoidable) mistakes. But these mistakes are the cost of moving quickly, and the benefits are far, far more significant than spending too much time agonising over your next move.

But, there will be times when you need to step back to see the proverbial wood from the trees. Often, I’ve found that when big new pieces of data come in, it’s the right time to review the entire product (and research) strategy. Have a few core assumptions that inform broad direction changed? Should we be thinking about an entirely different approach? Raise your head above the water, take a deep breath, and when you know which way that island oasis of excellent results is likely to be, plunge back in.

Think critically, but don’t boil the ocean as you sail through it. Hold your product teams accountable for making at least one great decision a week that ladder up to creating powerful product outcomes for your customers.

“It is remarkable how much long-term advantage people like us have gotten by trying to be consistently not stupid, instead of trying to be very intelligent.” — Charlie Munger, long-time partner of Warren Buffet