The teams that still do it by hand

An uncomfortable interview series

We build a generation workspace, so it is tempting to believe that everyone who has not adopted one is simply behind. Over December and January we interviewed fourteen studios and in-house teams who have looked at tooling like ours and decided against it, on purpose, with reasons.

Almost none of the reasons were the ones we expected. Nobody said the output was bad. Two people said it was too good, which took some unpacking.

Reason one: the review cost moved, it did not disappear

The most common answer, by a wide margin. Generation makes producing a first draft nearly free, and every team we spoke to had discovered that the bottleneck was never the first draft.

“We went from three concepts a week to thirty. Our art director still has the same number of hours. So now she is a filter instead of a designer, and she hates it.”

That is a real cost and it is not obviously worth paying. A team producing four carefully considered options is not slower than a team producing forty and discarding thirty-six — it just looks slower on a dashboard.

Reason two: the house style is the product

Several of the studios sell a recognisable look. Their clients are buying the fact that a piece of work is unmistakably theirs. Generation tooling is very good at approximating a style and, at least so far, noticeably bad at holding one across a long project.

One illustration studio put it plainly: the tooling could match their style in about eighty per cent of frames, and the remaining twenty per cent cost more to fix than the whole set would have cost to draw. That maths is specific to them, but it is not unusual.

Reason three: provenance obligations

Three of the fourteen work under contracts that require them to document where every asset came from. Not as a philosophical position — as a clause. For them the question is not whether generated work is acceptable, it is whether they can produce an audit trail that survives a client’s legal review.

This is the reason most likely to change, and the one the industry talks about least. Tooling that treats provenance as a first-class output rather than a compliance afterthought will win these teams. Nothing else will.

Reason four: it is genuinely not their bottleneck

A small but consistent group looked at the tooling, agreed it worked, and concluded that image production was simply not what slowed them down. Their constraint was client approvals, or scheduling, or a single overloaded senior person. Speeding up the fast part of a slow pipeline changes nothing.

Discovery      ▓▓▓▓▓▓▓▓▓▓▓▓▓▓░░░░░░░░  14 days
Approval       ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓░░░░  18 days
Production     ▓▓▓░░░░░░░░░░░░░░░░░░░   3 days   <- the part tooling helps
Revisions      ▓▓▓▓▓▓▓▓▓░░░░░░░░░░░░░   9 days

We drew this diagram with one of them on a whiteboard. It ended the conversation faster than any demo could have.

What we took from it

Two things. First, the honest pitch for a workspace like ours is narrower than the industry pitch: it helps most when volume is the constraint and consistency requirements are moderate. Outside that band, the value is real but small.

Second, the teams working by hand are not a backlog of future customers. Some of them are running a better process than the one we would sell them. Saying so out loud costs us nothing and buys a great deal of credibility, which is why this post exists.

If you are one of them

Keep going. The pressure to adopt tooling because everyone else has is not a business reason. When your constraint moves — and it usually does eventually — the tooling will still be here, and it will be considerably better than it is today.