← Journal Process · · 6 min

Why the Concept phase deliberately ignores what is possible

Why our Concept phase overreaches on purpose, what it should produce, and how the Abstract phase turns an ambitious idea into a buildable, costed plan.

Milo Perović
Founder & lead engineer

  • Architecture
  • CAEC
  • Concept
Why the Concept phase deliberately ignores what is possible

Most hardware projects die of caution long before they die of ambition. The first meeting turns into a list of reasons something cannot be done, and the product that ships is a reshuffle of last year’s parts. That is why the Concept phase of our process deliberately ignores what is possible: for a short stretch, nobody is allowed to say “we can’t”.

This post is for founders and product managers about to start a connected-device project, and for engineers who are sceptical that an unconstrained concept is anything more than an expensive mood board. We think it is the cheapest risk reduction you can buy, provided the next phase is disciplined about pulling it back.

What the Concept phase is for

Our process has four phases: Concept, Abstract, Exact and Conclusion. Concept answers one question: what should this product be if nothing stood in the way? We work out the ideal architecture, the materials, the interactions and how it should look, and we produce conceptual 3D renders so everyone is reacting to the same picture rather than to a paragraph of requirements.

The output is not a specification. It is a reference point. Every later decision gets measured against it, so when we drop a feature in the Abstract phase we know exactly what we gave up and why. Without that reference, compromises accumulate silently and nobody can say where the product drifted.

Why starting from constraints produces worse products

Constraints are real, but the ones teams bring to a first meeting are often not the binding ones. They are assumptions inherited from a previous product, a supplier’s catalogue or a half-remembered conversation about certification. If you design inside those assumptions from day one, you never find out which of them were soft.

A few patterns we see repeatedly when a project starts constraint-first:

  • The chip picks the product. Someone already knows a particular MCU, so the radio, the power budget and the feature set are bent to fit it, rather than the other way round.
  • The enclosure is a box around the PCB. The board is laid out first and the industrial design is asked to cover it, which is how you end up with a product that looks like a project box.
  • Cost is fixed before value is understood. A target bill of materials is set before anyone knows which features users would pay for, so the expensive-but-essential part is cut and a cheap-but-irrelevant one survives.
  • The app is an afterthought. The companion software is scoped as “a dashboard” because the hardware absorbed all the early thinking.

None of these is fatal on its own. Together they produce a product that is buildable and unremarkable, and the team only realises it after tooling money is spent.

What we allow ourselves to ignore

Ignoring limits does not mean ignoring physics. We will not sketch a coin-cell device that streams video for a year. What we set aside, temporarily, are the constraints that are negotiable with enough money, time or cleverness:

  • Unit cost and bill-of-materials targets.
  • Manufacturing method, whether that ends up as injection moulding, CNC or something else.
  • Which component families, vendors and ecosystems the team already knows.
  • Certification effort and lead times.
  • Development schedule, within reason.

What we do keep are the hard edges: the laws of physics, safety, and the core job the product exists to do. A concept that is impossible is useless; a concept that is merely expensive is informative.

The questions we ask instead

Rather than “what can we afford?”, the Concept phase asks questions about the user and the use. Where does the device live, and who touches it? What would make someone choose it over the thing they already have? What should happen when it fails? What would it feel like in the hand, and what colour and finish would make sense in the place it sits? These answers shape architecture far more than a parts list does.

How the Abstract phase pulls the concept back

An unconstrained concept only pays off if something disciplined follows it. In our process that is the Abstract phase, where the concept is refined against manufacturing reality and the client’s actual needs. It is also where most of the conversation with the client happens, because this is where the trade-offs are made.

We work through the concept feature by feature and ask three things of each one:

  1. What does it cost to keep? In parts, in power, in enclosure complexity, in firmware effort and in certification exposure.
  2. What does it cost to lose? In user value, differentiation, or the ability to add it later without a hardware revision.
  3. Is there a cheaper way to get most of it? Often the answer is a different sensor, a software feature instead of a hardware one, or a simpler mechanism that delivers the same experience.

The result is an architecture and a costed plan. Some features survive intact, some are simplified, and some move to a later version with the hooks left in place: a spare connector footprint, a firmware abstraction, room in the enclosure. Because the concept exists, those deferrals are deliberate rather than accidental.

ConceptAbstract
Main questionWhat should it be?What can we build, for whom, at what cost?
ConstraintsPhysics and safety onlyManufacturing, budget, schedule, certification
Typical outputsIdeal architecture, materials, conceptual rendersRefined architecture, costed plan, agreed feature list
Client involvementReacting to ideasMaking trade-off decisions

Common objections to an unconstrained Concept phase

“It wastes time.” The Concept phase is short compared with the build and test work that follows, and it is the only part of the project where changing your mind costs almost nothing. A render is cheap to redo; a PCB respin or a mould modification is not.

“It sets expectations we cannot meet.” Only if the concept is presented as a promise. We present it as a reference point, and we say up front that the Abstract phase will cut it down. Clients who have seen the ideal version tend to make better cuts, because they understand what each feature was for.

“Our constraints really are fixed.” Sometimes they are. A regulated market, a mandated battery size or an existing retail price point can be genuinely non-negotiable, and we treat those as hard edges from the start. The point is to separate the truly fixed constraints from the habitual ones, not to pretend the fixed ones away.

“Engineers will hate it.” In our experience the opposite is true. Firmware, hardware and industrial design working on the same ideal picture surfaces cross-discipline ideas early, before anyone has committed to a layout or an enclosure split line. You can see how those disciplines fit together across our product engineering services.

What we would do before your Concept phase starts

You do not need our process to get the benefit. Before any new connected product, a few days of deliberate overreach will tell you more than a week of debating a parts list. A short checklist:

  • Write down every constraint you believe you have, then mark each one as physics, regulation, contract or habit. Only the first three survive the Concept phase.
  • Describe the product from the user’s side first: where it lives, who handles it, what it replaces and what happens when it breaks.
  • Get a visual early. Even rough renders align a team faster than any requirements document.
  • Keep the ideal version on file. When you cut a feature later, record what it was for and what it would take to bring back.
  • Plan the pull-back as a separate step with its own decisions and its own budget, rather than letting compromise happen piecemeal during the build.

If you want to know how long each phase typically takes and what it costs, our FAQ on cost, timelines and IP covers the ranges. The short version: the phase that ignores what is possible is usually the one that makes the rest of the project cheaper.

Working on something similar?Free scoping call →