When Data Makes Marketers Less Creative
There is something slightly strange about modern marketing. We have more information about consumers than marketers have ever had before, yet I am not convinced that we are necess…
I am not arguing that marketers should stop using data. That would be ridiculous. Data is one of the most useful things available to a modern marketing team. It can reveal patterns that intuition misses, expose assumptions that are wrong and prevent companies from wasting enormous amounts of money. The problem begins when measurement stops being a tool for understanding and becomes the thing that determines what we are willing to create.
There is a subtle difference between using data to inform creativity and using data to constrain it. The first makes creative work smarter. The second can make it safer, more predictable and increasingly similar to everything else.
Data Is Very Good at Explaining the Past
One of the fundamental problems with data is that it is usually describing something that has already happened.
A campaign generated a 2.4% click-through rate. One audience converted better than another. A particular image produced more engagement. Tuesday performed better than Thursday. A particular subject line increased open rates. These are useful observations, but they are observations about the past. They tell us what happened under particular circumstances.
They do not automatically tell us what people will respond to next.
This sounds obvious, but marketers routinely turn historical performance into creative rules. If a particular type of headline worked, we produce more headlines like it. If short videos performed better, we make shorter videos. If people clicked when we used a discount, we use more discounts. If a particular audience converted, we spend more money targeting people who look like them.
Eventually, the marketing team can become extremely good at repeating what has already worked while becoming less willing to try something that has never been tested.
That is where creativity starts to suffer.
The Dashboard Can Become the Creative Director
Imagine a marketing team reviewing its monthly performance. The dashboard shows that product-focused posts generated more clicks than brand storytelling. The team therefore decides to reduce storytelling and increase product posts. The next month, the numbers confirm the decision. Product posts continue to generate more immediate engagement.
The conclusion seems obvious: product posts work.
But perhaps the more important question is what the dashboard cannot tell us. Maybe the brand storytelling is helping people remember the company without clicking. Maybe it is making the product advertising more credible. Maybe it is contributing to long-term brand preference in ways that a short-term attribution model cannot see.
This is one reason the obsession with measurable outcomes can become dangerous. Marketing does not only create immediate actions. It also creates memory, familiarity, associations and expectations.
A person does not necessarily click an advertisement because they like a brand. They may simply remember the brand when they eventually need the product months later.
That is considerably harder to put into a weekly performance report.
A recent argument from advertising executive Sean Cummins made exactly this point, suggesting that marketers have become so focused on digital tools, data and constant optimisation that they can lose sight of marketing's fundamental job of building memory. His argument is that memorable brands are often built through consistent ideas repeated over time rather than constantly changing creative in pursuit of the latest performance metric.
That is an uncomfortable idea for marketers who have been trained to optimize everything.
The Problem With Testing Everything
A/B testing is useful. There is no reason to pretend otherwise.
If you are spending significant amounts of money on advertising, testing different creative, audiences, landing pages and calls to action can save money and improve performance. The problem is not testing. The problem is assuming that everything valuable can be tested in the same way.
Some ideas need time.
If you launch a completely new brand idea and judge it after seven days because another advertisement generated more clicks, you may be measuring the wrong thing. A genuinely distinctive campaign may initially feel unfamiliar. It may require repetition before people understand it. It may produce discussion rather than immediate conversion.
The easiest thing to measure is not necessarily the most important thing to measure.
This is particularly relevant when comparing short-term performance marketing with brand building. A discount advertisement can produce an immediate response because it gives people an obvious reason to act. A memorable brand campaign may not generate the same immediate behaviour, but it can influence future purchasing decisions.
Dove's Real Beauty campaign is an excellent example. Research conducted before the campaign found that only 2% of women surveyed considered themselves beautiful. Dove used that finding not simply to create another product advertisement, but to challenge the conventional way beauty advertising represented women. The result became a long-running brand platform rather than a short-term creative execution.
The interesting part is that the data did not tell Dove to make an advertisement featuring real women. Someone had to interpret the research and make a creative leap.
The data identified a problem. Creativity found the opportunity.
Dove Didn't Let the Data Write the Campaign
This is an important distinction.
Dove could have taken its research and produced an advertisement saying, "Only 2% of women think they are beautiful. Dove believes every woman is beautiful." That would have been factually consistent with the research and almost completely unremarkable.
Instead, the research became the starting point for a much bigger creative idea.
The campaign challenged the conventions of the category. It featured women who did not look like traditional advertising models. Later work such as Evolution showed how an ordinary woman could be transformed through makeup, lighting and Photoshop into an idealized advertising image. The creative did not simply communicate the research. It dramatized the underlying problem.
That is what good marketing does with data. It does not merely repeat the finding. It asks what the finding means and then turns that meaning into something people can experience.
Marketing Week's analysis of the campaign notes that Dove's research into women's perceptions of beauty helped lead to the Real Beauty idea, while the brand combined longer-term emotional brand work with more product-specific activity. When Dove subsequently reduced its brand investment, sales declined, illustrating the danger of treating brand-building activity as disposable simply because its effects are harder to measure in the short term.
The data mattered enormously.
It just wasn't the creative idea.
The Most Interesting Ideas Often Look Bad in a Spreadsheet
One reason marketers become overly dependent on data is that data makes ideas easier to defend internally.
A creative director says, "I think this is interesting." That is a difficult statement to put into a board presentation.
A performance marketer says, "This audience produced a 4.8% conversion rate compared with 3.2% for the control group." Suddenly everyone feels much more comfortable.
Numbers create authority.
The problem is that creativity often begins where certainty ends.
An unusual campaign can be difficult to justify before it exists. If nobody has done it before, there is no historical benchmark. If it is genuinely different from what competitors are doing, there may be no comparable data. If it is designed to create cultural attention rather than immediate clicks, the usual performance metrics may be inadequate.
This creates a strange incentive. The safest creative idea is often the easiest to defend because it resembles something that has already worked.
And if every marketer makes decisions that way, every brand gradually starts producing variations of the same thing.
Coca-Cola's Share a Coke Was Simple, Not Safe
Consider Coca-Cola's Share a Coke campaign.
The basic idea was almost absurdly simple: replace the Coca-Cola logo on bottles and cans with people's names and encourage consumers to share them. From a distance, it could look like an obvious execution. But the campaign required an enormous amount of operational and creative work, including thousands of pieces of original artwork and extensive trademark checks to determine which names and words could appear on packaging.
What made the idea powerful was not a sophisticated targeting algorithm. It was the creative transformation of a mass-produced product into something that felt personal.
A data-driven marketer could certainly have helped determine which names were most common, which markets were most responsive and where the campaign performed best. But data alone would not have produced the central creative thought.
Someone had to imagine changing one of the world's most recognizable brands in a way that temporarily made the bottle feel like it belonged to an individual.
That is an important role for creativity. It connects information to an idea.
The Danger of Optimizing the Personality Out of a Brand
Another problem appears when companies optimize creative execution one small improvement at a time.
Change the headline. Improve the call to action. Move the button. Test another image. Shorten the copy. Change the opening three seconds. Add a stronger offer. Try a different thumbnail.
Each individual decision can make sense.
After hundreds of decisions, however, the work can become strangely lifeless.
The brand gradually gets stripped of anything that feels unusual because unusual things are difficult to optimize. Creative teams learn which colors produce clicks, which formats perform best, which words attract attention and which images generate engagement. Eventually they can produce technically excellent advertising that looks remarkably similar to everyone else's advertising.
This is particularly obvious on social media. Scroll through enough corporate LinkedIn posts and you start seeing the same structures repeatedly: a photograph of an executive, a short story, a lesson, three takeaways and an optimistic conclusion. None of these posts are necessarily terrible. The problem is that very few are memorable.
Data can tell you which version generated more engagement.
It cannot tell you whether anyone will remember the company three months later.
Data Can Also Give Marketers False Confidence
There is another issue that does not get discussed enough: numbers can be wrong without being inaccurate.
A dashboard can accurately report that Campaign A generated more clicks than Campaign B. That does not necessarily mean Campaign A was better marketing.
Perhaps Campaign A used a more aggressive headline that attracted curiosity but lower-quality leads. Perhaps it offered a discount that increased immediate conversions but trained customers to wait for promotions. Perhaps the campaign generated engagement because it was controversial but created negative associations with the brand.
The numbers can be perfectly correct while the conclusion is wrong.
This is why marketers need interpretation, not just measurement.
Data is evidence. It is not a verdict.
Creativity Needs Room to Be Wrong
One of the less obvious costs of excessive measurement is that it can make failure psychologically harder to accept.
If every campaign needs to prove its likely return before approval, genuinely experimental ideas struggle to survive. The organization becomes interested primarily in ideas that can be justified in advance.
But nobody can know in advance which creative idea will become culturally important.
Some of the best marketing campaigns initially look strange. They may be difficult to benchmark. They may require an audience to understand something unfamiliar. They may create reactions that are impossible to predict from previous campaign performance.
This does not mean marketers should abandon discipline and spend money on every creative idea someone has after lunch. It means companies need to distinguish between reckless experimentation and calculated creative risk.
A portfolio approach makes more sense. Some activity can be optimized aggressively because the objective is immediate performance. Some investment should protect and build the brand. Some should be reserved for experiments that may fail but could produce something disproportionately valuable.
If every part of marketing has to behave like a performance campaign, creativity eventually becomes a casualty of the approval process.
The Best Use of Data May Be Asking Better Questions
I actually think marketers sometimes underestimate how creative data can be.
The problem is not data itself. The problem is using data only to answer narrow questions.
Instead of asking, "Which advertisement got the highest click-through rate?" a marketer could ask, "Why did customers respond to this idea?"
Instead of asking, "Which audience converted most cheaply?" they could ask, "What does this tell us about the problem customers are trying to solve?"
Instead of asking, "Which content format generated the most engagement?" they could ask, "What does our audience find genuinely interesting enough to spend time with?"
Those are much more valuable questions because they turn data into insight.
And insight can lead to creativity.
The best marketers I have worked with are rarely people who choose between data and intuition. They use both. They look at what happened, try to understand why it happened and then use that understanding to make a decision that may go beyond the evidence.
That final step is where creativity enters.
The Marketer Still Has to Make the Leap
There is a temptation to imagine that if we collect enough data, eventually the correct marketing decision will become obvious.
I don't think it works that way.
More information does not remove judgment. In many cases, it makes judgment more important because there are more possible interpretations.
Two marketers can look at exactly the same customer research and reach completely different conclusions. One might see an opportunity to create a premium brand. Another might see an opportunity to compete on convenience. Both can point to evidence supporting their position.
At some point, someone has to make a choice.
That is why I do not want marketing to become a discipline where the safest answer automatically wins because it has the strongest spreadsheet behind it.
Marketing is not only the science of predicting behavior. It is also the art of creating something people did not know they wanted to see.
Data Should Inform Creativity, Not Replace It
I have no desire to go back to the days when marketers could spend enormous budgets based almost entirely on instinct and then declare a campaign successful because someone in the room liked it.
We know too much now to work that way.
We should use data. We should test. We should measure. We should understand customers and pay attention to what the evidence is telling us. We should absolutely stop repeating ideas that consistently fail.
But we also need to recognize the limits of measurement.
The campaign that gets the most clicks today is not necessarily the campaign that builds the strongest brand tomorrow. The audience with the lowest acquisition cost is not necessarily the audience with the highest lifetime value. The post with the most engagement is not necessarily the idea that people will remember. And the creative that wins an A/B test is not necessarily the creative that deserves to define the brand.
There is a point where optimization becomes a form of imitation. We keep improving what we already know how to measure because it is safer than creating something we do not yet know how to measure.
That is where marketers need to be careful.
Data should make us more curious, not less creative. It should help us understand people better, not reduce people to rows in a spreadsheet. It should challenge our assumptions, not replace our judgment.
The best use of data is not to tell a creative team exactly what to make.
It is to give them a better understanding of why something might matter, and then give them enough freedom to figure out what to do with that knowledge.
Because the moment we let the dashboard become the creative director, we may discover something rather uncomfortable: we have optimized our marketing so efficiently that it no longer has much to say.