From 15 Questions to 5: How We Perfected Onboarding Without Losing Insight
Part 2 of Building StepZero.eco: how we built a sustainability platform from discovery to launch.
The short version
Our first onboarding had 15 questions. Each one was justified. Each one was also quietly killing signup conversion, because nobody wants to fill in a form before they know whether your product is worth their time. We audited every question against one rule: does this directly power something the user sees within 60 seconds of finishing? Ten questions failed that test. The five that survived now drive a personalised results page that feels considered rather than generic, and the other ten live in the profile where users add them willingly once they are already invested. The full story is about how we made that cut, what "defaulting gracefully" actually means in practice, and why the moment after onboarding is the most important screen in your entire product.
You are designing onboarding for a new product and the list of questions keeps growing. What kind of business are you? How big is your team? Do you own or rent your space? What is your turnover? Do you have vehicles? Do you ship products?
Fifteen questions. Each one perfectly reasonable on its own. Each one backed by a real feature somewhere in the product. And each one quietly killing your signup conversion, because every field you add is another moment where someone decides that this is not worth their time and closes the tab.
When we built stepzero.eco, the first version of onboarding had exactly this problem. Here is how we fixed it.
The audit that changed everything
stepzero.eco helps small businesses figure out what sustainability actions to take, personalised to their specific situation. Behind the scenes, a matching engine sorts through a library of 230 curated actions to find the ones that actually fit. It considers your industry, how you operate, your team size, your premises, and several other factors.
The first pass at onboarding took the comprehensive route: ask all the questions upfront so we could deliver perfectly tailored results from day one. The logic felt watertight. Better data in means better recommendations out, and since the matching engine could use all of these inputs, why not collect them all while we had the user's attention?
The problem was equally obvious. Every additional onboarding step increases drop-off. We were asking people to answer 15 questions before they had any idea whether the product was worth their time. That is not onboarding. That is an interrogation, and no one signs up for a sustainability platform because they enjoy filling in forms.
Every onboarding question must earn its place by directly powering a feature the user will see within 60 seconds of completing signup.
So we sat down and audited every single question against one rule: every onboarding question must earn its place by directly powering a feature the user will see within 60 seconds of completing signup. Not "might be useful later." Not "helps us segment users for analytics." The question must connect to something the user experiences immediately. If you cannot draw a straight line from the question to something visible on the very first screen after signup, the question does not belong in onboarding.
We went through all 15 one by one. For each, we asked: if we remove this, what breaks in the first 60 seconds? If the answer was "nothing the user would notice," it got cut. Ten questions failed that test, which was uncomfortable but clarifying.
The five questions that survived
- What kind of business are you? This is the single most powerful input. It determines which of the 230 actions are even relevant. A cafe and a consultancy need completely different sustainability guidance. Without this, the results page would be generic and useless, and the whole premise of personalised recommendations falls apart.
- What matters most to you right now? This captures intent and priority. Is the user motivated by cost savings? Customer expectations? Doing the right thing? This shapes the tone and ranking of results, and it makes the experience feel personal from the first screen. Two businesses in the same industry with different motivations should see their recommendations presented differently.
- Where do you work, and how much control do you have over the space? Office, retail, warehouse, home, mobile — combined with whether you own or rent. These two data points together eliminate dozens of irrelevant actions. There is no point suggesting solar panels to someone renting a desk in a co-working space, and filtering those out early is what makes the results feel considered rather than generic.
- Which sustainability areas interest you most? Eight areas (energy, waste, water, travel, people, community, suppliers, carbon reporting), and users pick the ones they care about. This drives the structure of their personalised plan and makes the results feel curated rather than overwhelming.
- Your contact details. Name, email, postcode. Postcode feeds into local support recommendations and location-relevant suggestions.
Five questions. Under two minutes. Each one directly powers something the user sees on their results page within a minute of finishing.
What happened to the other ten
The remaining questions did not disappear. They still improve the personalisation. But we moved them to the business profile, where users can add detail at their own pace after signup, once they are already invested in the product and have seen enough value to want the experience to get even more specific.
Turnover, employee count, vehicle usage, shipping, outdoor space: they all live in the profile section now. When users add this information, their action plan automatically refreshes and gets more precise. But the initial experience works well without it.
The matching engine still runs on first load with sensible defaults for the missing information. A business that has not told us their turnover gets actions relevant to all sizes. Not perfect, but good enough to deliver a strong first impression that feels relevant and useful. And when they add details later, the plan sharpens in ways they can see and appreciate, which actually makes the profile completion feel rewarding rather than obligatory.
The payoff
The screen right after the fifth question shows personalised sustainability actions tailored to everything the user just told us. Industry, priorities, premises, focus areas. All reflected immediately.
This is the moment that justifies every question we asked. The user sees their answers turned into something useful, fast. No "thanks for signing up, now go explore." No empty dashboard waiting to be populated. Real, specific recommendations built around their situation, presented in a way that makes it obvious their answers mattered.
The difference between these two experiences is profound. One version of onboarding feels like a compliance form that exists for the company's benefit. The other feels like a short conversation that leads somewhere useful for the person having it. Same product, same data needs, completely different relationship with the user from the very first interaction.
What this taught us about respecting the user's time
The instinct to collect everything upfront is understandable, because it feels like you are being thorough. But thoroughness at the wrong moment is just friction wearing a lab coat. The real discipline is figuring out which questions earn their place in the first two minutes and which ones can wait until the user has a reason to care about giving you more.
Every onboarding field that does not visibly power something immediate is a silent tax on goodwill. And the frustrating thing is that you rarely see the damage directly, because the people it affects most are the ones who leave before they ever become users. You never hear from them. You just see a conversion number that feels lower than it should be.
If you audit your own onboarding against the 60-second rule and find that half the questions fail the test, that is not a problem. That is clarity. Move those questions somewhere they can still be useful, design your system to work gracefully with incomplete data, and let the first experience be strong enough that people want to tell you more.
The payoff is not just higher conversion. It is a fundamentally different relationship with your users from the very first interaction, one where they feel like the product respects their time rather than extracting their patience.
This is part 2 of the Building StepZero.eco series. Next: how we phased a complex product into shippable releases without cutting corners or losing the thread. If you are wrestling with similar onboarding decisions, we would genuinely enjoy the conversation. Say hello.