Why We Don't Present Three Design Concepts

Why Don't We Present Three Design Concepts?

Because picking from a menu is not a design decision, and the research most often used to justify it actually argues for something different. We explore several directions internally, merge the best of them, and bring one recommendation with the reasoning attached.

Almost every prospective client asks how many concepts they will see. It is a reasonable question, because three concepts has been the industry norm for decades and is what most studios still offer.

We think it is the wrong offer, and the interesting part is that the evidence for exploring multiple directions is strong. It just does not support showing them all to the client.

What Does the Research Actually Say About Multiple Directions?

That exploring several is clearly better than committing to your first idea. The Nielsen Norman Group's work on parallel and iterative design, in the version updated by Therese Fessenden in December 2024, reports figures from a 1996 study that are hard to argue with.

Simply picking the best of four parallel versions produced "measured usability 56% higher than the average." A merged design that took the best ideas from across the four did better still, at "70% higher." And that merged design plus a single round of iteration reached "measured usability was 152% higher than the average of the original designs."

The same article reports a 1993 Nielsen Norman Group study finding that iterative design "measured usability improvements by 38% per iteration," and a case study where a target metric improved 233% across six iterations, "equating to 22% per iteration."

Read those three parallel design numbers in order, because the ordering is the argument. Choosing the best of several is good. Merging them is better. Merging and then iterating is dramatically better.

So Isn't That an Argument for Showing Three Concepts?

It is an argument for making three, which is not the same thing. The research describes a design technique. Showing all of them to a client is a sales ritual, and it makes the weakest of the three outcomes the one you get.

When a client picks one of three, you land on "pick the best" and stop. Nobody merges, because merging after a selection means telling the client that the thing they chose was not quite right. And the iteration that produces the largest gain gets spent on refining a selection rather than improving a synthesis.

So we do the parallel work. Two or three genuinely different directions, explored far enough to know what each one is good at. Then we do the part the research says matters most: we merge, iterate, and bring one thing.

Is This Just the Paradox of Choice?

No, and we want to be careful here, because that is the reason studios usually give and it does not hold up well.

The choice overload hypothesis, as the literature defines it, "states that an increase in the number of options to choose from may lead to adverse consequences such as a decrease in the motivation to choose or the satisfaction with the finally chosen option." It is an appealing idea and it has been repeated constantly in design writing.

But the meta-analytic review by Benjamin Scheibehenne, Rainer Greifeneder and Peter M. Todd, published in the Journal of Consumer Research in 2010, examined "63 conditions from 50 published and unpublished experiments" with 5,036 participants and "found a mean effect size of virtually zero but considerable variance between studies." Their conclusion was that some moderators proposed in individual studies "may still explain when and why choice overload reliably occurs," which is a long way from a general law.

So three options probably will not paralyse your client. That is not our objection. Our objection is about what the choice does to the work.

What Actually Goes Wrong When a Client Chooses?

The decision gets made on the wrong criteria, by the wrong process, and the reasoning disappears.

A design direction is a set of connected decisions about hierarchy, tone, density, colour, type and motion, each made for a reason. Presented side by side, three of those collapse into an aesthetic preference. The question in the room stops being "which of these serves the buyer we are trying to reach" and becomes "which do you like."

Worse, it becomes which does the group like, and groups do not have taste. They have compromise. What emerges is usually a request for the layout from option two, the colours from option one and the typography from option three, which is not a design. It is a collage of three incompatible sets of reasoning.

We have watched that conversation many times, and it is the origin of a large share of the redesigns that fail to move anything, for reasons we set out in why website redesigns fail.

Doesn't One Concept Make the Client Powerless?

Only if you present it badly. This is the fair objection and it is worth answering properly.

A single direction presented as a finished artefact is a take it or leave it demand, and clients are right to resist that. A single direction presented as a recommendation is something else entirely: here is what we made, here is why, here is what we tried and rejected and why, and here is what we are least sure about.

That last part does most of the work. We bring the rejected directions to the conversation as reasoning rather than as options. The client sees the thinking and can challenge any part of it. What they cannot do is pick a winner without engaging with the argument, and that constraint is the point.

What Does the Presentation Actually Look Like?

Context first, then the work, then the open questions. We start with what we learned about the audience and the commercial goal, so the design is being judged against something rather than against taste.

Then one direction, shown on real content at real breakpoints. Not a hero image in isolation. A page someone could use, because a direction is only real once it has to survive a long paragraph, an awkward product name, and a phone.

Then the explicit list of things we want the client's judgement on. Usually three or four questions, and they are never "do you like it." They are questions like whether the tone is right for the buyer, whether a claim overstates the product, or whether a section is missing something a sales conversation always covers.

That structure is a design review with a purpose, which we have written about separately in running a design review that produces decisions.

How Do You Reduce the Risk of Getting It Wrong?

By front loading the agreement, not by hedging with options. Three concepts are, when you look at them honestly, a way of deferring the hard conversation about direction until after the visual work is done.

We have the conversation first instead. Who is the buyer and what do they already believe. What does this site need to do that the current one does not. Which competitors will they see beside you. What are you not willing to look like. Those answers narrow the space enormously, and they are cheap to get wrong on a call and expensive to get wrong in Figma.

We also show work earlier and rougher. Grey boxes and real words before anything is styled, which both removes the aesthetic distraction and surfaces structural disagreement while it is still free to fix. That is the reasoning behind designing in grayscale first.

Does This Ever Go Wrong?

Yes, in one specific way: when we have misread the brief and the single direction is not close enough. That is a real risk and it is on us rather than on the method.

When it happens, the recovery is faster than it would be with three concepts, because the reasoning is on the table. A client can say "the tone is too corporate for our market" and we know precisely which decisions to revisit. With three options, the same feedback produces a request to see three more.

The honest summary is that this method puts more pressure on the discovery phase and less on the presentation. We think that is the correct place for the pressure. It is also why we will not quote a design phase before a scoping conversation.

Would We Recommend This to Anyone Hiring a Studio?

We would recommend asking a different question. Rather than how many concepts you will see, ask how the studio will explain its decisions, and what they will do when you disagree with one.

A studio that can answer that clearly is doing the parallel exploration and the merging whether or not they show you the byproducts. A studio whose answer is "we will show you three and you pick" has told you that the decision is being handed to you, which sounds like control and is actually the absence of expertise you are paying for.

None of this means the three concept model is dishonest. Plenty of good work has come out of it. We just think the evidence points somewhere else: explore widely, merge ruthlessly, iterate once more than feels necessary, and bring one thing you can defend.

If you are about to brief a website project and want to talk about how the process should work before anyone opens a design tool, we are always happy to have that conversation at phoenix.studio.