Stop Using AI Just to Save Time. Use It to Find Growth

Most of the business conversation about artificial intelligence starts with productivity. AI can write an email faster, summarise a report, generate a presentation, create social posts, analyse a spreadsheet and turn a meeting transcript into a set of actions within seconds. Those capabilities are genuinely useful, and businesses would be foolish to ignore opportunities to remove repetitive work.

However, I increasingly think efficiency may be one of the least interesting things AI can offer marketing. Saving thirty minutes on a task is helpful, but discovering something about customers or the market that creates a new source of growth could be worth vastly more.

That distinction matters because there is a risk that businesses use one of the most significant technological changes of our time primarily to produce more of the same marketing they were already doing.

Doing the wrong thing faster is still doing the wrong thing

Imagine a business decides it needs more social media content. Someone introduces an AI system that can create an entire month's worth of captions in fifteen minutes. The organisation has unquestionably become more efficient because a task that previously consumed several hours can now be completed almost instantly.

The problem is that nobody has yet established whether social media content was the issue.

The real constraint might be poor customer retention, weak local awareness, an unclear proposition, inadequate distribution, a confusing website or low consideration among potential customers. The business may therefore have used AI to become dramatically more efficient at executing an activity that was never particularly important.

This is why the strategic application of AI interests me much more than the production application. The biggest opportunity is not simply asking how technology can help marketers produce existing outputs more quickly. It is asking how it can help us identify better things to do.

Marketing does not have a content shortage

There is already more marketing content in the world than any person could reasonably consume. Businesses produce enormous volumes of social posts, videos, emails, podcasts, blogs, graphics and advertising every day, and generative AI is going to increase that supply dramatically.

If almost every business can suddenly create competent copy, imagery and basic video at negligible cost, simply producing more content becomes a weak source of competitive advantage. The bar for production has fallen, which means judgement becomes more valuable.

The difficult questions remain difficult. Which customers should we prioritise? What should the brand stand for? Where is the growth opportunity? Why are people choosing competitors? Which parts of our proposition matter most? Where is demand changing? Which distinctive assets are actually being remembered?

AI can help marketers explore those questions, but it does not make them irrelevant. In many ways, it makes strategic thinking more important because the cost of producing activity has become so low.

When everyone can make more marketing, knowing what is worth making matters more.

Use AI to interrogate the business

Most organisations already possess enormous amounts of information about their customers. The problem is that it sits in different places and is rarely analysed together. There are website analytics, sales reports, customer reviews, survey responses, CRM records, complaints, emails, search queries and conversations happening every day between customers and frontline staff.

Historically, making sense of that information could be time-consuming. Qualitative research might require somebody to manually read hundreds of comments, categorise recurring issues and identify patterns. Large spreadsheets required a degree of analytical confidence that many smaller businesses simply did not have internally.

AI can significantly lower that barrier.

A business can analyse large numbers of customer reviews and ask which themes repeatedly appear in positive feedback. It can explore cancellation comments and look for patterns in why customers leave. It can compare enquiry data with sales outcomes and investigate whether certain types of lead are more likely to convert. It can organise open-text survey responses into themes and help marketers see patterns that would otherwise be difficult to detect.

That does not mean handing the strategy to the machine. AI can hallucinate, misinterpret data and produce confident answers that deserve to be challenged. Human judgement, good data and a proper understanding of the business remain essential.

What changes is the speed at which marketers can explore questions.

Ask growth questions instead of productivity questions

The way businesses use AI is partly shaped by the questions they ask it. If the instruction is “write ten Facebook posts”, the value created is primarily a time saving. That may be useful, but the upside is naturally limited.

A more interesting question might be to analyse hundreds of customer enquiries and identify the recurring reasons people decide not to buy. The AI could be asked to examine whether particular services generate different objections, whether certain customer groups behave differently or whether language repeatedly appears that suggests the proposition is being misunderstood.

Similarly, instead of simply asking AI to summarise customer reviews, a marketer might ask which parts of the experience appear most associated with recommendation, which repeatedly create dissatisfaction and which could potentially become distinctive strengths in future marketing.

The difference is not really about prompt engineering. It is about the type of problem we are choosing to solve.

One application saves the marketer twenty minutes of writing. The other could reveal an issue or opportunity worth thousands of pounds.

First-party data becomes even more valuable

This also changes the conversation about customer data. Businesses spend enormous amounts of time building followings on platforms they do not control, and those platforms decide how much of the audience sees any particular message.

First-party data gives an organisation a different kind of relationship. With appropriate permissions and governance, it allows businesses to understand behaviour over time rather than treating every customer interaction as an isolated event.

AI potentially makes that information easier to use.

A gym might examine joining dates, attendance patterns and cancellations to understand where retention weakens. An event organiser could compare registration, attendance, geography and repeat participation. A retailer could look at products frequently bought together, while a professional service business might analyse the characteristics of enquiries that become valuable long-term customers.

None of those analyses require AI to make the final commercial decision. What the technology provides is an ability to explore data more quickly and ask more questions of it.

That is a considerably more interesting use of AI than producing another month's worth of generic social posts.

The value shifts towards judgement and creativity

There is an irony in the rise of generative AI. The easier it becomes to produce competent marketing material, the more valuable genuinely distinctive human thinking may become.

AI can generate fifty headlines in seconds, but somebody still needs to understand which idea is appropriate for the brand. It can suggest campaign concepts, but it cannot fully reproduce the knowledge somebody develops by spending years around a particular community, customer base or culture.

Marketing has always involved a mixture of evidence and creativity. Evidence can tell us a great deal about how brands grow, how advertising works and how customers behave, but there still needs to be somebody willing to notice an opportunity and develop an interesting response.

AI should increase the amount of information available to that person and reduce the amount of low-value work surrounding them. It should not make every business sound and behave the same.

There is a genuine risk that businesses use identical AI systems, similar prompts and the same marketing conventions to produce an endless sea of competent but forgettable communication. Distinctiveness may therefore become even more important in an AI-heavy marketing environment.

Efficiency eventually reaches a ceiling

There is also a simple economic reason why growth may ultimately be the more valuable application.

Efficiency has a natural limit. A four-hour task can become a two-hour task, then a twenty-minute task and perhaps eventually a task that takes almost no time at all. Once that happens, there is very little additional value available from making it faster.

Growth works differently. Finding a new customer segment, improving the proposition, identifying a retention problem or discovering an underdeveloped source of demand can create value that continues long after the initial analysis.

That does not mean productivity improvements should be ignored. Giving a marketing team ten hours back every week is clearly useful. The question is what happens to those ten hours.

If the additional capacity is simply used to produce even more marketing content, the organisation may end up with a larger volume of activity without creating much additional value. If the time is invested in understanding customers, exploring opportunities, developing ideas and making better strategic choices, the impact could be much greater.

AI should make marketers more strategic

There has been considerable discussion about whether AI will replace marketers. I suspect the more interesting divide will be between marketers whose value is primarily based on producing outputs and those whose value comes from understanding markets and solving business problems.

If somebody's role consists almost entirely of creating basic social captions, routine emails and generic content, AI will inevitably change the economics of that work. If a marketer can diagnose a business problem, understand customer behaviour, identify an opportunity, develop a strategy, create a distinctive idea and connect marketing activity to commercial outcomes, AI becomes something very different.

It becomes leverage.

That is why the question I am increasingly interested in is not how AI can help businesses produce more marketing. It is how AI can help marketers understand something about the customer, market or business that they could not previously see as easily.

Saving thirty minutes is useful, and there will be thousands of worthwhile efficiency applications for AI. However, the biggest prize may not be completing the same marketing work faster.

It may be finding the next thing worth doing.