Argue from the document, not from taste

Why “I think it looks better” always loses

When a designer defends a decision on taste, they have moved the discussion onto ground where the client’s opinion is exactly as valid as theirs, and the client is paying. That argument cannot be won and should not be.

The brief exists so that you never have to have it. A decision defended from an agreed document is a decision defended from something the client already said yes to.

“We agreed the one thing this has to do is get people to the open day. The date is the only element that survives at arm’s length. Enlarging the logo means the date stops being the only one, and then nothing is.”

That is not a matter of opinion. It is an argument against a shared standard, and the client can engage with it: by disagreeing about the goal, which is a conversation worth having, or by accepting it.

The shape of a good defence

Four sentences, in this order:

1. The goal, restated from the brief. “The one thing is the open day.”
2. The mechanism. “Ranking these three levels is what makes one thing survive a glance.”
3. The evidence, if you have any. “Five people I showed it to all said ‘open day, Saturday’.”
4. The cost of the alternative, stated plainly and without drama. “We can enlarge the logo. The likely effect is that people remember the brand and not the date.”

Then stop talking. The commonest failure at this point is continuing to argue after you have made the argument, which reads as anxiety and invites the client to reassure you by finding a compromise neither of you wants.

Show the rejected option once

Showing the alternative you considered and set aside is useful: it demonstrates the decision was made rather than defaulted to, and it lets the client see the trade-off rather than take your word for it.

Do not show three options where two are deliberately weak in order to steer the choice. Clients recognise it more often than designers think, and once they have recognised it every subsequent presentation is read as a manipulation. The same judgement applies here as to your audience: a technique that works only while hidden is one to be suspicious of.

Concede fast on everything that is not load-bearing

Go through the feedback and mark each item: does the message depend on this?

Most items will not. The typeface, the exact blue, the corner radius, the order of two items of equal weight: concede these immediately and pleasantly. It costs the work almost nothing and buys you the standing to hold the one or two things that matter.

A designer who defends every decision is not principled; they are unable to distinguish, which is a competence problem rather than an integrity one, and clients read it correctly.

When they overrule you

They are allowed to. It is their business, their money and their risk, and they may know things you do not.

What to do:

1. State the risk once, clearly, in writing. “Noted: adding the fourth message means the open-day date is unlikely to be what people remember. Happy to proceed.”
2. Then do it properly. A grudging execution is worse than either option and it is visible. If you are doing it, do it well.
3. Keep both versions. Sometimes the result comes back and the conversation reopens, and having the other version ready is worth more than having been right.

What not to do: sabotage, sulk, execute it badly to prove a point, or bring it up afterwards. All of these are more damaging to you than the design decision ever was.

The limitation

Arguing from the brief only works if there is a brief, it was written down, and it was agreed. Without that, every discussion is taste against taste, and you will lose most of them for the structural reason above.

There is also a category this cannot reach. Sometimes the real objection is not about the work at all: somebody needs to be seen to contribute, somebody’s superior has an opinion, somebody was not consulted earlier and this is where it surfaces. No argument about hierarchy addresses any of that.

The move is to recognise it rather than to keep making design arguments at a problem that is not a design problem. Asking directly (“is there someone else who needs to be comfortable with this?”) is usually met with relief, and it converts an unwinnable debate into a scheduling issue.