futurest
ApproachAI ProductsResearchInnovationTeamPodcast
APPROACH

The right question first. Then the best answer.

That isn't a slogan, it's a sequence. Most AI projects don't fail on the technology. They fail because nobody tested the question before the building started. Our approach puts that test first — and makes it fast enough that it holds nobody up.

Why the question matters more now, not less

AI has made building dramatically cheaper. A prototype that would have taken a quarter five years ago now takes two weeks. Plenty of companies draw the wrong conclusion from this: just start, it barely costs anything.

The flaw in that logic: cheap to build does not mean cheap to build the wrong thing. The cost of an AI product nobody uses isn't in the development. It's in the organisation's burned attention, in the trust lost among staff, and in the time that wasn't spent on the project that would have worked.

Which is exactly why method hasn't gone out of fashion. It has only got faster.

THE THREE TESTS
Desirability — does it solve a real problem?

The question: What problems does the user actually have, and what job are they trying to get done — functionally, emotionally, socially?

How we answer it: Design research and design strategy. User interviews, observation, journey and touchpoint mapping. AI helps us analyse qualitative material at a scale that used to be impossible without losing the nuance — but it doesn't replace talking to real people.

How you know this test was skipped: The product ships, technically flawless. Nobody uses it.
Feasibility — can we build it?

The question: Is the solution technically and organisationally viable — with the capabilities this company actually has?

How we answer it: Foresighting, trend and technology analysis, and an honest read on strengths and gaps. We use our own deep research tooling for innovation management. And because we run an AI product ourselves, we know first-hand what works with today's models and what stays demo magic.

How you know this test was skipped: The pilot works. Scaling doesn't.
Viability — does it hold up?

The question: Does the business model fit the usage pattern — and does it lead to sustainable profitability?

How we answer it: Business model design, scenarios, defensible assumptions instead of wishful numbers. With AI projects there's a question that didn't exist before: what does it cost to run, per user per month, once the thing is genuinely in use?

How you know this test was skipped: The product is popular and loses money with every new user.

How a project runs with us

1.
Understand

Trend analysis, user research, problem definition. It ends with the question we're going to work on — written down, specific, testable.

2.
Design

Innovation sprints and workshops that turn the question into concepts that hold. AI tooling across the whole innovation funnel, so we can test more options, not fewer.

3.
Build

Rapid prototyping through to a product that runs. We build with you, not before you — from one team, without the friction that appears when advice and delivery sit in different companies.

4.
Embed

A product only counts as a success when the organisation can carry it. Enablement, learning journeys, handover.

What makes us different

There are consultancies that write excellent strategies and leave delivery to someone else. There are development studios that build excellently — whatever they're handed. The gap between the two is where most innovation efforts die. futurest has both sides: the method from nearly twenty years of digital product development and a decade of AI projects with large enterprises and mid-sized companies — and an AI product of our own in the market, AIDGEN, which we built and operate ourselves. We don't recommend anything we haven't done ourselves.

Get in touch