Speculative Design for Product Managers: Testing Your Roadmap Against the Next Ten Years

Speculative design gives product managers a way to test a roadmap against a future nobody has shipped yet. Rather than asking what to build next quarter, it sets the product ten or more years out, makes that future specific enough to argue with (a named concept, a press release dated 2038) and then works back to the assumptions, bets and discovery questions that should change in the current planning cycle. It produces decisions, not forecasts.

In product organisations the practice has a reputation as a vision exercise — an offsite, a mood board, a deck shown once. Done properly it is closer to discovery: it treats the long-range future as evidence about which bets on the roadmap are fragile. For the general definition, start with what speculative design is.

Nor is it new to product companies. In 2021 three researchers at Facebook, Sanya Attari, Charley Scull and Mahboobeh Harandi, presented a paper at EPIC, the conference on ethnography in industry, called "Leveraging Speculative Design to Re-Imagine Product Roadmaps". They ran speculative experiments with business owners to explore futures of digital advertising, starting from a diagnosis any PM will recognise: businesses iterate on products quickly, often reactively, and struggle to get past short-term needs.

What product managers already have

Product managers arrive at this work better equipped than most, because much of the job is already writing about things that do not exist.

A PRD describes a product before it is built. A roadmap is a dated story about the future, with a Now, a Next and a Later. OKRs state an outcome and leave the route open. Discovery (interviews, assumption mapping, prototype tests) is a discipline for finding out cheaply that you were wrong. And many product teams already write a working-backwards PR/FAQ, which Colin Bryar and Bill Carr describe in Working Backwards (2021): a one-page press release and a short FAQ that define a product's customer experience before anything is built.

All of this transfers. PMs think in assumptions and bets, and they own the planning artefacts where the output of futures work has to land. What the toolkit lacks is distance, and the habit of following a consequence past the first one.

Why product managers can't see past the quarter

The product manager's native horizon is the quarter. Delivery runs in two-to-six-week increments, objectives are set quarterly, and anything beyond the current cycle sits in a column labelled Later, which everyone knows is a slide, not a commitment. This is not a failing. It is how software ships.

The cost is a vision vacuum. The product vision is a paragraph written three years ago, usually containing the word "seamless", and nobody can say what it rules out. One product lead who took our course put it more precisely than we have:

"My pitch was, 'as a team we are always focused on what to build in the next quarter but we don't have a clear vision of where to go next.' Taking the speculative design course provides tools to create that vision at multiple points in the future. As a result, I have been able to lead multiple workshops with my team which has led to more interest from the team and help us be more future oriented."

The fix is not to stretch the roadmap. Inside two years the future is the roadmap, and a futures conversation set there collapses into defending decisions already made. We set the horizon beyond ten years, where the platform has been replaced and nobody's current release is at stake. That distance makes it possible to say about 2038 what nobody can say about Q3: that the thing being built might be the wrong thing. We explain why in time horizons.

Where speculative design fits in product management

Each existing artefact gets a counterpart that runs alongside it, not instead of it.

The press release row needs a note, because PMs who know the PR/FAQ assume they already do this. A working-backwards press release is a commitment device, written to secure agreement to build. A speculative press release is written from a decade away about a product nobody has decided to build, and its job is to be argued with. So it needs the paragraphs a PR/FAQ never contains: who objected, what the launch made obsolete, which regulator took an interest.

A worked example: the journey that isn't a flight

Take a hypothetical product manager who owns search and results for a European travel booking app; call it Wayfare. Next year's roadmap is smarter fare alerts, a flexible-dates calendar and a new sort order, and every item assumes a trip is a flight chosen on price.

The brief. Source of change: limits on short-haul flying in Europe. Domain: how people search for and book journeys between European cities. Horizon: 2038. Audience: the product leadership team that signs off the search roadmap.

Signals. On 23 May 2023 France published the decree banning domestic flights where a direct train takes under two and a half hours, in practice three routes from Paris-Orly: to Bordeaux, Lyon and Nantes. That same week European Sleeper, a Dutch-Belgian start-up, began running a night train between Brussels and Berlin via Amsterdam, the first direct overnight link between the two capitals in more than a decade. On 18 October 2023 the EU adopted ReFuelEU Aviation, requiring fuel suppliers at EU airports to blend in a rising share of sustainable aviation fuel: 2% from 2025, 70% by 2050. None is large on its own. The trend: flying made scarcer and dearer where rail can compete.

Future wheel. Assume it is fully realised: in 2038, flying between European cities is restricted wherever rail can do the journey in a few hours, and the remaining flights carry the cost of mandated fuel. The first ring, more train travel and dearer flights, is the one everybody in travel already expects. In the second, journeys get longer and cross several operators, so the product stops being a ticket and becomes a connection. The question customers care about is no longer how cheap, but who sorts it out when the second train is late. Results have to show combinations no single operator sells. In the third ring, trips become fewer and longer, and when a strike strands thousands overnight, the booking app is the company people expect to find them a bed.

The concept. Wayfare Through. Tagline: "One booking. Every connection. Our problem if it breaks." A single booking for a multi-operator journey of a day or two (Copenhagen to Rome, say) covering the trains, a hotel night in the halfway city and a guarantee: if any leg fails, Wayfare rebooks and pays. Sketch: a ticket screen with a two-day timeline, a moon over the night in Munich, and a status line reading "All connections protected".

The PM then writes the launch press release, dated April 2038. Its headline: Wayfare retires its "cheapest flight" sort. Its third paragraph covers regional airports complaining that the change steers travellers away from them; its fourth, the rail operators who refused to sell through Wayfare until liability for missed connections was settled.

Implications. Starting with human experience: a visit to your mother takes two days, with a night in a city you never meant to see. Socially, people short of time rather than money (shift workers, carers, single parents) lose easy access to distant family. Culturally, the stopover city becomes a destination. Ethically, Wayfare decides which connections it will guarantee, and so which cities get the overnight trade. Economically, margin moves from commission to the cost of the guarantee. How it goes wrong: a rail strike on a Friday in August, and every protected connection breaks at once.

Backcasting. What had to be true first: more countries adopting the French rule, with longer thresholds; fuel mandates pushing fares on busy routes above the cost of train plus hotel; rail operators agreeing to sell through third parties with shared liability for missed connections; employers writing rail-first rules into travel policies. And one rung in Wayfare's own hands — rail timetables and multi-operator combinations in its search index, with a way to judge how reliable each connection is. More on the method in backcasting.

What changes this quarter. Not a feature. The PRD for fare alerts gains a written assumption, "the trip is a flight, chosen on price", and a named signal that would trigger a review: a second country adopting a French-style rule. Discovery adds three sessions in which customers react to the two-day Through booking, not to ask whether they would buy it but to hear what they would trade time for. And rail timetable ingestion moves from Later to Next, because it is useful now where rail already competes and a precondition in nearly every version of 2038.

Where it goes wrong for product managers

Shipping the future. The concept gets a ticket and an MVP. At that point it is a feature with a long name, and its implications are lost. A speculative concept is a probe into the roadmap, not an item on it.

Running discovery on 2038. Customers cannot evaluate a ten-year future in a survey, and asking "would you use this?" returns their present preferences in costume. Show them an artefact and listen to what they object to: that is evidence about values, not a prioritisation score. Our guide for UX designers covers testing reactions rather than usability.

Horizon creep. The session opens in 2038 and by lunchtime is arguing about next year's release, because the quarter is every PM's gravity. Fix the horizon in writing, and park anything with a current release date attached.

The vision deck. The PM writes one future, falls for it, and uses it to align the team. That is a pitch. The value comes from several futures that disagree, held long enough to see which roadmap items survive all of them.

A first exercise you can run this week

1. 1.Open your current roadmap. For the three largest items, write down one assumption each makes about the world rather than about the user, such as "customers choose on price" or "we own the customer relationship".
2. 2.Pick the assumption that would hurt most if it broke. Find two real signals from the last three years, named and dated events, that suggest it could.
3. 3.With your product trio, spend forty-five minutes on a future wheel, on a whiteboard or in Miro. Write a year at the top, ten or more years from now, and go three rings.
4. 4.From the outer ring, write one concept: a name, a tagline, two sentences of description, a sketch on a sticky note.
5. 5.Write its launch press release, dated that year, on one page. Include who objected and what it made obsolete.
6. 6.Work back: list three changes that had to happen first. For each, ask whether anything on your roadmap depends on it not happening.
7. 7.Take one result into your next planning session: an assumption written into a PRD, a discovery question, or a Later item that moves.

Where to learn speculative design as a product manager

Speculative Design Basics is our self-paced online course: 42 lessons, no prior experience assumed, taught by J. Paul Neeley. For product managers the Process chapter matters first, because it covers time horizons and how speculative design relates to existing design processes. Tools & Methods covers signals, future wheels, future prototyping and backcasting, and the Creating Value chapter deals directly with user research and product and service development. The full methodology is in our practitioner's workflow; for the portfolio side, see the guide for innovation teams, and for every other role, speculative design for every role.

Key takeaways

• •Speculative design lets product managers test the roadmap against a future ten or more years out, then work back to this quarter.
• •Inside two years the future is just the roadmap, so the work is set beyond ten.
• •A speculative press release is not a PR/FAQ: it is written to be argued with, and says who objected and what it made obsolete.
• •Customers cannot evaluate a ten-year future in a survey; treat their reactions to an artefact as evidence about values, not demand.
• •The concept is a probe into the roadmap, never an item on it.