The revision tax: why AI hasn't given marketing teams their time back
There’s something I’ve noticed more and more in conversations with marketing teams recently.
They’re producing more than they ever have. AI has made it quicker to get the first draft of an email, a social post, a campaign idea or a landing page onto the screen.
But it doesn’t necessarily feel like they’ve got that time back.
Instead, a lot of it is being spent at the other end: rewriting, refining, correcting and trying to make the work sound less generic.
Which is why a study Optimizely published landed the way it did. Covering just over 2,000 marketing leaders across 7 markets, it found that more than three quarters of marketers now spend at least 3 hours a week editing, fact-checking or correcting AI-generated output.
They call it a revision tax, which is a good name for it. Whatever time the tools gave back is quietly being handed over again at the other end.
The obvious reading is that the tools aren't good enough yet.
I don't think that's the whole story.
Those 3 hours are often the sound of a decision being made in the wrong place.
What's actually happening inside those 3 hours
When someone reviews AI-generated copy and finds it generic, the instinct is to blame the output.
In my experience, the problem often sits further upstream.
The brief was vague, so the work came back vague, and the person reviewing it is now supplying all the judgement that was missing from the brief in the first place.
They're doing strategy in the edit.
One paragraph at a time, without calling it that, and without anyone else in the business necessarily seeing it happen.
You see it when an email goes through three or four rewrites because it still doesn't quite land. The wording changes, the subject line changes, the CTA moves around. But somewhere underneath all that editing is a much more fundamental question that hasn't been answered clearly enough: who exactly is this for, and what do we want them to think or do?
Until that decision is made, no amount of polishing really fixes the problem.
That's an expensive way to make decisions, and a slow one.
It's also largely invisible on a P&L, which is why it can continue for months without anyone naming it. Nobody files a line item for “rewrote it because we hadn't decided who it was for”.
The businesses I see getting real value from these tools aren't necessarily the ones with the cleverest prompts.
They're the ones who already know who they're talking to, what they want that audience to think, and what they've decided not to do.
Give a tool that level of clarity and it's genuinely useful.
Give it a fuzzy brief and it will produce fuzzy work considerably faster than you could yourself.
The constraint moved
For most of my career, making the work was one of the hard parts.
I spent 15 years at Microsoft leading Xbox campaigns across more than 40 markets, and a serious chunk of that job was production logistics. Getting assets made, localised, approved and out of the door on time.
Strategy mattered, but capacity was a very real constraint.
For a lot of everyday marketing, that constraint has now changed dramatically.
You can produce 10 versions of a campaign in an afternoon.
So the question is no longer simply whether you can make it.
It's whether you should.
Which version? For whom? What job is it doing? Is it worth putting budget behind?
Those questions were always harder than production. They were just easier to hide when making the work itself took so much time.
Now there's much less standing in front of them.
Volume was never the win
I worked with a creative agency on a global retail brand, and we created 2,459 assets in the first 12 months.
It's a number I use because it's concrete, but the volume isn't actually the interesting part.
The interesting part was deciding which 2,459.
Every one of those assets existed because someone had made a call about who it was for, where it needed to appear and what job it was supposed to do.
The same logic scales down.
If you're a growing brand putting out 4 emails and 12 social posts a month, whether that activity works has far less to do with how quickly you can make it than with the thinking sitting behind it.
Faster production doesn't solve unclear priorities.
It just helps you execute them faster.
Three questions the tools won't answer for you
Most of my strategy work comes back to three questions.
Who are your customers, really?
Not the demographic summary, but what they're actually trying to do, what matters to them and what's currently getting in the way.
What makes you genuinely different?
Not the line you'd like to put on your website, but something meaningful enough that a customer might actually repeat it to someone else.
How do you reach them?
Given where their attention already goes, what they're receptive to and what role your marketing needs to play.
Answer those properly and the tools become a real advantage because you're pointing them at something specific.
Skip them and you've found a quicker route to the same generic output everyone else is producing.
Optimizely's UK research from earlier this month found 75% of consumers say the marketing they receive is irrelevant to them, and 69% say it's generic or not tailored.
More volume isn't going to close that gap.
What this means for how you resource marketing
If production is your constraint, you hire for capacity.
Executives. Agencies. Freelancers. More hands.
If judgement is your constraint, the answer looks different.
It's often a smaller amount of a more senior person rather than a larger amount of a junior one.
Someone who can hold the whole picture, make the calls about what not to fund, align the people involved and set a brief clear enough that everything downstream gets easier and faster.
That's a big part of what I do as a fractional marketing lead.
Those engagements don't need to be full time because the value isn't in adding another pair of hands to production.
It's in having someone owning the marketing agenda, setting priorities and helping the business make better decisions about audiences, budgets, channels, campaigns and growth.
AI hasn't removed the need for strategy.
If anything, it's made the cost of not having one much easier to see.
Put another way: the revision tax gets paid by whoever is holding the vague brief.
Better to fix the brief.
If this sounds familiar
If you're producing more marketing than ever but can't confidently say what all of it is doing for the business, that's worth a conversation rather than another tool.
If you need ongoing senior marketing leadership to set the priorities, align the team and make those decisions, take a look at my Fractional Marketing Leadership service.
Or, if you need to get the strategic foundations clear first, my Marketing Strategy Sprint is designed to do exactly that.
