AI Made Marketing Cheap. It Did Not Make Brands Interesting.

This Week's Marketing: July 13th Edition

Let us know if this sounds familiar: your Monday planning meeting now has more material than it could have produced in an entire month three years ago.

There are forty campaign concepts, twelve landing-page variations, a competitive teardown, several hundred keyword ideas, six versions of the product narrative, synthetic customer personas, social posts for four platforms, and a polished video storyboard. Most of it arrived overnight. None of it is obviously wrong.

The problem begins when someone asks which idea your company should actually use.

Several concepts are plausible. A few sound professional. Nearly all resemble something a competitor could have produced with the same prompt, the same model, and the same half-hour of attention. The team has solved the production problem and discovered a much less convenient bottleneck: someone still has to know what is good.

That is the more interesting marketing story of 2026.

AI has lowered the cost of producing a presentable campaign, article, advertisement, image, email sequence, research summary, and product story. It has widened the number of directions a team can explore and compressed the time required to turn one direction into a usable asset. Those are real advantages.

They are also becoming widely available.

When every competent team can generate more options, production volume becomes less distinguishing. The competitive question moves upward. Who recognizes the strongest idea? Who knows what to reject? Who can connect a creative choice to a real customer tension, a product truth, and a business strategy? Who can make the company sound like itself after the software has made everyone sound fluent?

AI made average marketing cheaper. It did not make average marketing more memorable.

The marketing conversation is moving past AI novelty

The shift was visible throughout the conversations surrounding Cannes Lions this year. McKinsey’s post-Cannes analysis described an industry moving away from broad AI optimism and toward the harder question of what actually produces business value. Its analysts also noted fatigue around the technology itself and concern that marketers had over-rotated toward AI while underweighting creativity, creators, and the work that moves the P&L.

That distinction matters. Marketers have spent several years asking what AI can generate. Most of the answers are now familiar. It can produce drafts, variations, summaries, concepts, classifications, research starting points, audience segments, workflow triggers, and personalized outputs at a speed no traditional team can match manually.

The useful question now is what deserves to survive that generation process.

Digitas North America CEO Amy Lanzi made a similar argument in a recent Cannes conversation with The Verge. She described AI as a meaningful operational tool that can accelerate workflows and improve how agencies use data, while rejecting the idea that it will independently rescue weak advertising or replace human creativity. Her broader point concerned systems: marketing still needs media, CRM, creative, creators, data, and business outcomes to work together.

AI can reduce the friction between an idea and its execution. It cannot guarantee that the idea carries any cultural, emotional, or commercial weight.

This is why teams should resist treating AI adoption as a marketing strategy by itself. A strategy describes which market the company wants to shape, which audience it wants to earn, which problem it understands, which promise it can keep, and which choices it will make differently from competitors. A tool can help explore those questions. The company remains responsible for answering them.

Efficiency is becoming an entry requirement

Marketing leaders still need efficiency. No serious team should romanticize slow research, manual formatting, duplicated work, endless revision cycles, or the peculiar corporate ritual of paying experienced people to copy information between tabs.

AI belongs wherever it can remove that drag.

It can summarize a transcript, compare several hundred customer comments, identify repeated objections, reorganize a content archive, generate alternative structures, prepare a first-pass brief, and adapt an approved idea across formats. A good AI content strategy workflow should give humans more time for interpretation rather than use automation as a reason to remove interpretation from the process.

The danger arrives when efficiency becomes the final objective.

A team that reduces the cost of producing forgettable material has improved a cost center. It has not necessarily improved marketing. The company may publish more frequently, test more variations, and create a larger inventory of assets without becoming more recognizable, more persuasive, or easier to choose.

Surface’s existing work on the quality-versus-volume trap applies beyond written content. Volume creates learning only when the work contains enough difference to teach the team something. Testing twenty near-identical variations mostly teaches the team which shade of sameness performed marginally better.

The practical inference is that efficiency will matter increasingly as infrastructure. Every competitive marketing organization will be expected to produce, adapt, analyze, and distribute material quickly. The differentiating layer will sit above that infrastructure in the decisions governing what gets produced.

The companies selling the AI future are buying brand advertising

The AI industry itself offers one of the clearest demonstrations.

Companies developing products that promise new forms of discovery, automation, personalization, and productivity are investing heavily in familiar brand-building channels. Sensor Tower data reported by Marketing Brew found that U.S. AI brands spent approximately $424 million on advertising during the first quarter of 2026, more than three times their spending from the same period a year earlier. Anthropic’s measured spending increased 1,184% year over year, while OpenAI’s increased 800%, with investment distributed across television, streaming, YouTube, LinkedIn, and other channels.

There is no contradiction once we understand the category problem.

AI companies can explain that their models are faster, more capable, better at reasoning, more integrated, safer, cheaper, or more useful for a particular workflow. Their competitors can make many of the same claims. Product capabilities change quickly. Benchmarks become obsolete. Features travel between platforms. Buyers receive an expanding number of technically credible options.

The companies therefore need something feature comparison alone cannot provide: memory.

They need buyers to recognize a name, understand its character, associate it with a type of work, and feel some confidence before examining the latest model card. They need category language, visual systems, product rituals, founder narratives, customer stories, public demonstrations, repeated media exposure, and a coherent explanation for why their version of the future deserves attention.

The same pressure applies to less glamorous B2B categories. When competing companies offer similar workflow automation, analytics, integrations, dashboards, or AI assistants, the technically accurate description begins to sound like category boilerplate.

Brand becomes the system that gives those capabilities shape.

Taste is a business capability

Taste sounds suspiciously soft inside an operating plan because marketers often confuse it with personal aesthetic preference.

A creative director liking one font more than another is preference. Taste becomes strategically useful when a team can evaluate several possible expressions against the company’s customer, product, category, ambition, and moment.

Taste means recognizing that an idea is technically polished and emotionally vacant.

It means understanding when a clever campaign distracts from the product truth, when a popular visual trend makes the company less distinctive, when a provocative opinion lacks enough substance to survive scrutiny, and when a simple sentence carries more authority than a page of generated sophistication.

Taste also means knowing what belongs together.

A product interface, event, executive voice, customer story, advertisement, social post, and sales deck do not need to look identical. They should feel as though the same organization made them. That coherence helps people build a mental model of the company.

A strong content brief can establish the audience, intent, evidence requirements, internal links, structure, and strategic point of view. It cannot completely automate the editorial decision that tells a team whether the final piece has become worth reading. Standards create the floor. Taste determines how far the work moves beyond it.

The teams most likely to benefit from AI will therefore be the teams with the strongest ability to select. They can generate widely, judge rigorously, and commit clearly. Teams with weak judgment may simply create more plausible distractions.

Interesting brands give people something to carry

A brand becomes interesting when people can take part of it with them.

Sometimes that part is a sentence. Sometimes it is a visual device, a product interaction, a report, a term, a customer ritual, a founder’s way of explaining the problem, or a strangely specific point of view. It gives customers something they can remember, repeat, show, quote, recommend, or use to explain their own identity.

Recent consumer branding provides literal examples. Vogue Business has described a wave of food, beverage, protein, water, and wellness products borrowing the visual logic of fashion and beauty. Packaging, creator placement, product design, and social recognizability help ordinary consumable products function as attainable status objects. People encounter them in routines, photographs, influencer feeds, and public spaces rather than only on a supermarket shelf.

B2B companies rarely have a beautiful bottle to place in a celebrity’s hand. They still create social objects.

A benchmark report can become a social object. So can a distinctive dashboard, an assessment, a certification, a useful framework, a customer score, a category term, a conference ritual, a printed field guide, a product-generated image, or an annual index that gives an industry language for something it previously felt but could not name.

The product itself can create moments worth sharing. A useful automated insight, a surprisingly elegant report, a clear before-and-after comparison, or a result customers feel proud to show can travel further than a corporate announcement asking people to believe the company is innovative.

Interestingness does not require making everything strange. It requires giving the audience something more specific than functional adequacy.

Creators reveal what brands cannot automate

Creators matter to this conversation because their advantage is rarely pure production capacity. Many creators have small teams, modest resources, and little of the infrastructure available to a large marketing organization.

Their leverage comes from audience understanding, recognizable judgment, repeated presence, and accumulated trust.

Recent creator-economy conversations at Cannes have focused increasingly on creators as business builders rather than temporary campaign inventory. Leaders from UTA’s creator division described creators launching products, developing intellectual property, building companies, and seeking more durable ownership beyond brand sponsorships. They also warned that over-commercialization can weaken the audience relationship that made the creator valuable in the first place.

That tension is instructive for brands.

An audience trusts a creator because the creator has demonstrated a pattern of attention. People understand what the person notices, what they ignore, what they recommend, how they explain things, and where their boundaries sit. The individual’s judgment becomes part of the product.

Brands often remove that pattern during the approval process. Legal teams soften the claim. Product teams add seven qualifications. Executives replace the interesting sentence with a safe one. The social team attaches a trending format. AI smooths the remaining language. The final asset is accurate, approved, optimized, and almost entirely free of evidence that anyone cared.

The growing interest in the Storyteller role in marketing reflects an attempt to recover that judgment inside companies. The role becomes valuable when the storyteller can work with product truth, customer experience, executive ideas, and business strategy. Hiring a good writer and giving that person nothing worth saying merely creates a more elegant bottleneck.

AI should move human work toward judgment

The strongest AI marketing operating model does not divide work into “AI content” and “human content.” It assigns responsibility according to the type of intelligence required.

AI is useful for breadth. It can inspect more sources, produce more alternatives, classify more records, and adapt an approved idea across more formats.

Humans remain responsible for consequence. They decide which evidence is credible, which customer tension matters, which claim the product can defend, which idea fits the cultural moment, which risk is worth taking, and which expression should become part of the brand.

A practical workflow looks like this:

1. Use AI to widen the field. Gather customer language, competitor patterns, cultural references, channel behavior, and several possible creative territories.

2. Use people to interpret the field. Identify the tension that matters, the product truth that can resolve it, and the perspective the company can credibly own.

3. Generate several expressions. Explore formats, headlines, narratives, images, demonstrations, and distribution opportunities without assuming the first polished answer is the right one.

4. Select against explicit standards. Evaluate distinctiveness, customer relevance, product accuracy, memorability, emotional force, channel fit, and business utility.

5. Adapt rather than duplicate. Let an article, video, executive post, sales asset, event, and product experience express the same idea according to the strengths of each format.

6. Measure what the idea changed. Track recognition, engagement, branded demand, sales-language adoption, creator or customer reuse, conversion behavior, and the quality of the leads that followed.

That process requires more judgment than asking a model for one campaign. It also makes the automation useful.

How to measure whether a brand is becoming more interesting

Interestingness should not become another vague brand metric that survives entirely because nobody can disprove it.

Marketing teams can look for observable effects:

• Do more buyers search for the company by name?

• Can prospects explain the company accurately before the first sales call?

• Do customers repeat the company’s language without being prompted?

• Do creators, partners, analysts, and communities reference the brand’s frameworks or artifacts?

• Do distinctive campaigns outperform generic campaigns after controlling for distribution?

• Are direct traffic, return visits, newsletter engagement, and branded demand increasing?

• Does the company appear in consideration sets beyond the channels it paid to enter?

• Do sales teams report stronger recognition, clearer expectations, or less category explanation?

• Are the leads attracted by the brand’s ideas more qualified for the product?

None of these proves that one creative decision caused a revenue outcome. Together, they can show whether the company is becoming easier to remember, explain, and choose.

The content marketing operating model for 2026 already separates visibility from trust and conversion. Brand interest belongs in the tissue connecting those layers. Visibility gives someone the chance to encounter the company. Interestingness gives the encounter a chance to remain.

What marketers should do now

Begin by auditing the material your team produced during the last month.

Remove the company name and logo. Ask whether a knowledgeable person could identify who made each asset. Review the claims competitors could make without changing a word. Look for product descriptions that explain capabilities without communicating why anyone should care. Find the places where automation accelerated work without adding customer understanding.

Next, identify one thing the company can create that competitors cannot reproduce through prompting alone. That may be customer access, proprietary data, an operational methodology, a founder’s experience, a product-generated artifact, a community, a visual system, or a distinctive way of framing the category.

Then protect human attention around that thing.

Do not spend the most experienced person’s time formatting drafts. Give that person responsibility for interpretation, selection, and standards. Let software widen the possibility space. Let accountable people decide where the company belongs inside it.

Finally, measure the automation program by what it frees the team to do. Faster production is useful. More interesting customer research, sharper product stories, stronger creative systems, better experiments, and more memorable brand experiences are considerably more useful.

The average is getting better and less valuable

AI will continue improving the average quality of marketing output.

The average campaign will look cleaner. The average article will be more structurally complete. The average advertisement will have more variations. The average marketing team will possess research, design, writing, analytics, and production capabilities that would have required several specialists a few years ago.

That is good news for the craft and uncomfortable news for anyone whose advantage depended on being merely competent.

The remaining opportunity is human and operational at the same time. Teams can use AI to remove unnecessary labor while becoming more attentive to the customer, more demanding about the work, more coherent in their choices, and more willing to express a point of view that belongs to the company.

Marketing became cheaper to make. Meaning did not.