Insights · Deciding · September 2, 2026 · 5 min

One in five of our workshops ends with "do not build this". That is the product.

Why a studio that sells software spends two days a month telling people to buy a platform instead.

Why a studio that sells software spends two days a month telling people to buy a platform instead.

Product Discovery is priced from your scope, from $2,500 to $12,000, and about a fifth of the time it ends with a recommendation not to build. Clients sometimes ask why we would run a process that talks a paying customer out of a six-figure project. The answer is that the discovery is the product, not the build.

A "no" that pays for itself

A build that should not have happened costs everyone: the client pays for software a platform already provided, we spend a year maintaining something that should not exist, and the reference is lukewarm. A Product Discovery that says no is a fraction of the investment in the build, and produces a backlog, an estimate and a written recommendation the client can use with any vendor, including a SaaS one.

Where Product Discovery ends up
Build with us
57%
Buy or extend a platform
21%
Put it on UnflowAI™
14%
Wait
8%
Example figures. Unflow workshop log, rolling twelve months.

The four outcomes

They are always the same four, and we publish the reasoning for each rather than presenting a conclusion.

  • Buy a platform. Your process matches how most companies do it, and configuration closes the gap.
  • Extend the one you have. The platform is right; the 30% it misses is an integration, not a rebuild.
  • Put it on UnflowAI™. The rules change often, few systems are involved, and being wrong is cheap.
  • Build. Several systems, customers in the loop, rules that move, and real cost when it is wrong.
Acme Customer PortalShipmentsDocumentsInvoices
All · 128In transit · 41Delayed · 3Delivered
ReferenceRouteETAStatus
SHP-20481 · 12 palletsDallas → MemphisToday, 16:40In transit
SHP-20477 · 3 palletsLaredo → DallasTomorrowDelayed 1 day
SHP-20470 · 8 palletsHouston → AtlantaDelivered 09:12Delivered
SHP-20466 · 1 palletDallas → El PasoDelivered 08:30Delivered
SHP-20462 · 6 palletsMemphis → DallasToday, 19:10In transit
SHP-20459 · 2 palletsSan Antonio → LaredoTomorrow, 10:00In transit
SHP-20455 · 14 palletsDallas → AtlantaDelivered yesterdayDelivered
SHP-20451 · 4 palletsHouston → DallasDelivered yesterdayDelivered
Synced with TMS 4 minutes ago
When the answer is build
Web applications

Customer portals, operations platforms and back-office systems, built around how your team wants to work. Most take 6 to 30 weeks.

See the service →

Why we publish the ratio

A number that costs us sales is the only kind worth publishing. Every studio says it is honest about scope; the ones that mean it can tell you how often they walk away, and from what.

The workshop that told us not to build is the reason we came back two years later for the one we did.

A client who bought a platform first
Not sure what to do next?

Should you build it at all?

Answer a few questions about how your team works, and we’ll tell you which way we’d go, with the reasoning and a playbook sent to your inbox.