Producing another asset is easier. Knowing whether it deserves anybody’s attention still takes judgement. Four questions to ask before commissioning more.
AI has made content much easier to produce. Getting somebody to care about it remains a considerably more expensive business.
For a stretched marketing team, the saving is welcome. Work that once needed an afternoon can be ready for review before lunch. There is more room to experiment with ideas that previously never survived the discussion about resources.
But. And there is always a but. HubSpot’s 2026 research found that 83.5% of marketers said they were expected to produce more content because of AI. You still have to approve that work and defend what it costs to put it in front of people. A faster draft does not tell you whether the idea deserves the investment. It can, however, get it far enough through the process that stopping it becomes somebody’s problem.
The content bottleneck has moved
Production has often limited what a content plan could achieve. With cost and resource often being the driving factor over what the business needed.
AI loosens some of those constraints considerably, although original evidence and expert contributions still take time. Audience attention was scarce long before generative AI arrived. What changes is how easily we can create material that competes for it, and how quickly an organisation can mistake that capacity for a reason to publish.
RETHINK SOCIAL · ARTICLE COMPANION
AI made content cheaper. It did not make attention easier.
Explore what happens when the same monthly effort is divided between easy-to-produce generic content and fewer assets built around a distinctive point of view, useful evidence and clearer customer value.
Important: this is a thinking model, not a performance forecast. Every rate is an editable assumption so you can test your own content economics.
The bottleneck moves
Useful attention index over 12 months
Durable value
Trust + utility signals that remain active over time
Discovery potential
Illustrative retrieval/search opportunity index
Output trade-off
How many assets the test scenario produces each month
WHAT DESERVES TO EXIST?
Four questions before you commission the next asset
12-month detail
| Month | Current-mix assets | Test-scenario assets | Attention | Trust | Utility | Discovery |
|---|
Methodology, assumptions and cautions
This is a thinking model, not a forecast. It shows the trade-off described in the article: when AI reduces production friction, the limiting factors become attention, trust, utility, distinctiveness and discoverability.
Generic assets are assumed to be easier to create but decay faster and earn fewer useful signals. Distinctive assets take more effort but are modelled as more useful, more trusted and more likely to be retrieved, cited, saved or returned to.
The test scenario is simply the alternative allocation you choose with the slider. It is not a prescribed Rethink mix and it does not imply that 50% is optimal.
The discovery multiplier is an editable scenario assumption. It is not a claim that search or AI systems apply a fixed uplift to a content type.
No ROI claim: the simulator does not convert attention, trust or discovery into revenue. The link to business outcomes still needs evidence from your own journey.
The brief becomes more important. Without a clear purpose, capacity gets allocated to whoever asks for it. The content plan becomes a record of requests the team could accommodate.
The costs that follow the draft
An agency offering extra variations at little additional cost may be providing excellent value. Those variations could address different audience needs or allow a worthwhile test. But somebody still has to choose between them and arrange distribution. There is more performance to interpret, often without agreement on what the extra work was supposed to establish.
Cheap to make does not mean cheap to use.
There is also the effort of declining work. Once a draft exists, colleagues have something to react to and become attached to. A suggestion that could have been challenged in a brief turns into a discussion about wording. You can spend a surprising amount of senior time improving something that had a doubtful reason for being commissioned.
Competent enough to publish. Easy to forget.
A clear explanation can be almost entirely interchangeable with what the next business publishes. It may answer a basic question adequately while giving the audience little reason to remember where the answer came from.
If you are paying to distribute it, that distinction matters. Necessary information has a job to do, but a claim that the piece will differentiate the brand needs more substance. When the idea could belong to any competitor, the logo is doing a great deal of the work.
Specificity gives people a reason to pay attention. Showing whether a product fits into the space a customer actually has resolves an uncertainty that describing it as “convenient” leaves untouched. The value comes from the evidence. AI can assist with production without being the source of that value.
The same scrutiny applies to human-written work. A generic brief does not become a strong idea because a person spent longer answering it.
Useful does not have to mean instructional
Content can deserve attention because the observation is funny or the execution is recognisably yours. People do not open social apps hoping every brand will set them homework. Memorable entertainment can be doing exactly the job required. That matters because a narrow definition of usefulness could leave every brand publishing worthy explainers. Expertise is valuable when it contributes something the audience needs. Creative judgement matters just as much when the task is to make a brand worth remembering. Adding “make it more engaging” to a prompt settles neither question.
Google’s guidance for generative search also encourages original experience and useful, non-commodity material. It offers no guaranteed result, but it gives marketers another reason to question a polished summary of information already freely available.
Will the meaning survive beyond the feed?
Accessible content may appear in search or contribute to an AI-generated answer before somebody visits the website, sometimes without a visit at all. The audience may encounter an extract, separated from the introduction that made its purpose obvious.
A product claim whose limitation appears much further down the page is easier to misunderstand when extracted. Keeping the qualification close to the claim protects its meaning, although it cannot guarantee how an AI system represents it.
Core search systems still matter. Google says its generative features rely on its search ranking and quality systems, and warns against manufacturing query-variation pages primarily to manipulate search or AI responses. More versions of substantially the same answer do not resolve a lack of useful substance.
This makes the commissioning question more demanding: is there an idea here that remains valuable when someone encounters it without the surrounding campaign?
Save beats scroll
Some content has a useful life well beyond its first burst of engagement. In its July Search Console update, Google described tracking supported social and video content across its discovery services. One suggested use is identifying older videos regaining search traction.
That connects with our “Save beats scroll” thinking. A buyer may return to an explanation when a purchase becomes urgent. A colleague may receive it weeks after publication. Neither behaviour fits comfortably into a report that treats the first few days as the final verdict.
A save does not prove a sale. It gives you a reason to investigate whether the work remains useful, rather than assuming its job ended with the initial response.
Before another asset gets commissioned
Start with purpose and audience need, then choose the format. This has long been part of our planning approach. AI makes it easier to skip because producing a draft can feel quicker than resolving the brief.
These four questions belong in that conversation.
- Is there something specific here?
Identify what makes this worth receiving from your business. Evidence customers cannot find elsewhere gives you a strong starting point. So can a creative observation that is unmistakably yours. Some essential information will be ordinary; commission it with that job in mind. - Does it answer a meaningful audience or buyer question?
“Is this for someone like me?” can matter as much as a question about compatibility. Be clear about the audience’s reason to engage. The subject may be relevant while the angle misses what people actually want to know. - Can the idea travel clearly through other discovery environments?
Check whether the meaning survives without the surrounding campaign. Someone arriving from search needs enough context to understand the demonstration or claim. Not every asset needs to work everywhere. - What happens if we do not make it?
Name the gap it would leave. An unanswered objection may justify more work. An empty publishing slot deserves more scrutiny. Existing material might need better distribution, while a worthwhile experiment may warrant something new.
The answers should change where effort goes. A topic with strong evidence behind it may deserve substantial investment. Another needs a straightforward update. Treating both as identical units in a monthly deliverables list conceals the judgement that matters most.
Spend the saving deliberately
More production capacity gives teams room to pursue ideas that were previously too expensive to test. That is worth being ambitious about.
But every saved hour does not have to return as another asset. Sometimes its best use is getting closer to the customer or giving an idea more thought. Before accepting a higher output target, establish what the additional work is expected to contribute. The production saving is already yours. You do not have to spend it all on producing.
