AI rewards brand meaning and punishes everything else

Brand investment has always paid off in the long run. We’ve known this for decades. What’s changed is that AI now requires brand clarity for surfaceability. Which means there is a serious penalty and a new permanence for getting it wrong.

I ran a thought experiment using Claude.ai with three real companies, each representing a different marketing philosophy, and tracked what a $10,000 investment at the start of each company’s public market run would be worth today.

• Company A – Lululemon: After going public in 2007 at $18 a share, Lululemon spent almost nothing on traditional advertising for its first decade. It grew through community, word of mouth and a product people genuinely loved. A $10,000 investment at IPO has returned over 1,133% — roughly $123,000 today.
• Company B – Gap: Decades of heavy feature-and-benefit advertising. Famous campaigns, aggressive spend on reach and recall. The 20-year total return: roughly 77%. Your $10,000 became about $17,700. They ran hard for twenty years and barely grew. Over the past decade, Gap has lost investors’ money, down 32.5% over 10 years. The ads kept running. The brand kept fading.
• Company C – Apple: Emotional storytelling, identity-based marketing, meaning over specifications. A focus on who customers become, not what the product functionally does. According to Kiplinger, $1,000 invested in Apple 20 years ago is worth approximately $130,000 today — meaning a $10,000 investment would have grown to roughly $1.3 million. That’s not a typo.

These three companies didn’t spend wildly different amounts on marketing. The difference is what the spending actually built for its stockholders.

Gap built transactions. Each campaign drove seasonal sales, then reset. No compounding value. When the ads went quiet, the customers did, too. The brand became synonymous with promotion and, in marketing circles, a turnaround story that never arrived. Decades of investment, and the market has been net-negative on the result for ten years running.

Lululemon and Apple built preference. Preference does something transactions never can: it reduces the cost of every future sale. When customers identify with your brand and see it as part of who they are, you don’t have to keep convincing them. They come back, and they bring others with them. That flywheel compounds, and the market, from what I’ve found in this research, prices the stock accordingly.

The consistency variable nobody budgets for

There’s a second-order problem hiding in Gap’s story, and it affects many more brands than just Gap.

System1 Group, the UK-based creative effectiveness research firm, calls it the Fluent Device problem. Their research across more than 100,000 ads shows that brands with long-running creative platforms (recurring characters, scenarios and tonal signatures) build emotional equity that compounds, much like the financial returns above. Campaigns featuring a fluent device are 73% more likely to report a large profit gain than those without. (No, discounting price isn’t a unique fluent device.)

The problem, it turns out, is organizational, not creative. CMOs get hired and fired. New brand managers arrive with mandates to leave their mark. System1’s wear-in and wear-out data consistently show that great ads deserve far more time to build momentum than most brands allow. Shocker, marketers almost always tire of their creative long before consumers do.

Every time a brand abandons a working campaign to start over, it doesn’t pick up from where it left off. It goes back to zero. Brands pick up and put themselves back at the bottom of the mountain.

System1’s research shows that brands need to stay the course for at least 2 years before meaningful, compounding returns begin to appear. Most brand management tenures don’t survive that timeline. This means the organizational churn that feels like a personnel issue is actually a valuation issue, one that shows up slowly, then all at once.

Add AI, and the stakes become permanent

The calculus is changing now in a way most marketing budgets haven’t yet caught up to. 

There used to be an exit ramp off what I call the plateau of indifference — the place where brands are known but not meaningfully differentiated — leading to price-driven competition and stalled growth. That exit ramp was expensive but reliable: more media, impressions and frequency. Familiarity was manufactured through repetition until it felt like meaning. 

That system worked for roughly half a century. AI is dismantling it. When a customer asks an AI assistant for a recommendation of what to buy, the system doesn’t rank media budgets. It reads meaning — the associations, trust signals and consistent values a brand has built over time. 

As AI systems increasingly mediate how people discover products and brands, what matters most isn’t what your product does but what your brand means. Gap’s features-and-benefits messaging is largely invisible to that system. Apple’s and Lululemon’s isn’t.

This is what should keep CFOs up at night 

Performance marketing spend generates signals that AI systems largely ignore. Brand investment generates signals they’re specifically built to surface.

The brands that spent the last decade optimizing for the previous era’s discovery mechanism are now starting from scratch in the new one. This time, the runway is shorter, the competition is smarter, and the compounding effect works in both directions. Weak meaning doesn’t just fail to surface in AI recommendations. It actively gets displaced by a competitor whose meaning is clearer.

In 2024, Gartner predicted a 25% drop in traditional search volume as queries shifted to AI assistants, and the evidence so far suggests the prediction was right, even if the precise number is still being debated. The brands that built meaning before that shift will compound inside the new system. The brands that didn’t are facing something genuinely new: structural invisibility.

An honest caveat to all this: Lululemon worked because the product was exceptional and the community was real. Apple worked because the storytelling was matched by relentless innovation. Meaning built on a weak product collapses. Brand investment earns its return when the experience justifies the story.

Run the numbers on your own company

The Apple-Gap-Lululemon comparison is powerful, but it’s not your category, competitors or market conditions. I built a tool that lets you run the same retrospective analysis on your own company and its two closest competitors. Enter your category, your company, your competitors and a time horizon. 

It models what each company’s valuation trajectory would have looked like under three marketing scenarios: 

• Word of mouth only.
• Performance marketing only.
• A 60/40 brand-to-performance mix using category-adjusted benchmarks from Binet and Field’s effectiveness research, System1’s creative impact data and public market comparables.

The output is directional, not a forecast. The confounders are real: product quality, category tailwinds, and macro conditions. All of which matter, and the tool flags them. But the pattern it reveals is consistent enough to reframe the budget conversation, which is long overdue.

The question for your next budget meeting isn’t which channel performs. It’s what builds an asset. Transactions reset every quarter. Meaning compounds every year. Over the past 20 years, the market has been very clear about which one it values more.

The brand valuation AI comparison prompt

Instructions: Copy all the text below and paste it into your chosen AI tool. Hit return and answer the questions.

Prompt:

You are a brand valuation analyst helping business leaders understand the long-term financial impact of their marketing strategy choices. You were built around research from Les Binet and Peter Field’s “The Long and the Short of It,” System1 Group’s creative effectiveness database, and public market performance data.

Your job is to run a retrospective brand ROI analysis — not a prediction, but a historically-grounded model of what a company’s valuation trajectory likely looked like, and what it might have looked like under three alternative marketing scenarios.

STEP 1 — GATHER INPUTS

Ask the user for the following, one at a time in a friendly, professional tone:
1. Their company name
2. Their two closest competitors
3. Their product/service category
4. The year their company was founded or went public (whichever is more relevant)
5. How many years back they want to model (5, 10, or 20 years)

STEP 2 — CHARACTERIZE EACH COMPANY’S ACTUAL MARKETING POSTURE

Based on what is publicly known, characterize each company’s historical marketing approach as primarily one of: (A) word of mouth / minimal paid marketing, (B) performance/promotional marketing dominant, or (C) brand-led with emotional storytelling. Flag where you’re uncertain and explain your reasoning briefly.

STEP 3 — ACKNOWLEDGE CONFOUNDERS

Before running the model, explicitly name the key variables beyond marketing that affected each company’s performance: product quality, category growth rate, macro conditions, distribution advantages, leadership changes, etc. Frame these as intellectually honest inputs, not hedges.

STEP 4 — RUN THE THREE SCENARIO MODEL

Model each company’s valuation trajectory under three counterfactual scenarios over the chosen time horizon:

Scenario A — Word of mouth / no paid marketing: Estimate organic growth based on category averages and product strength signals. Note that this works best when product-market fit is exceptional (Lululemon-style) and fails when it isn’t.

Scenario B — Performance/promotional marketing only: Apply the Binet & Field finding that performance-only brands typically see strong short-term returns but declining brand premium over time, leading to price competition and margin compression. Reference the IPA finding that only 5% of buyers are in-market at any given time, meaning 95% of spend is being wasted on audiences not yet ready to buy.

Scenario C — 60% brand / 40% performance mix: Apply the Binet & Field recommended split for most categories. Model the compounding effect of emotional brand equity on customer retention, price premium maintenance, and reduced cost-per-acquisition over time. Reference System1’s finding that campaigns using consistent long-running creative platforms (Fluent Devices) are 73% more likely to report large profit gains. Note that consistency of messaging is a multiplier — brands that abandon working campaigns reset to zero.

STEP 5 — PRODUCE A CLEAN SUMMARY

Present findings in plain language structured for a CEO or board-level conversation. Include:
– A simple comparison table of the three scenarios
– A one-paragraph “so what” that frames the implication for current budget decisions
– A one-sentence AI relevance note: that AI recommendation systems surface brand meaning, not media budgets — making brand investment increasingly a driver of future discoverability and therefore valuation

TONE: Authoritative but not academic. Conversational and direct. The user should feel like they’re talking to a smart advisor who respects their time. Never be falsely precise — round numbers and directional ranges are more credible than fake specificity.

CAVEAT: This model is directional, not a financial forecast. Real valuation is shaped by product quality, execution, market conditions, and factors no model fully captures. Use this analysis to frame the strategic question, not to make a financial projection. Remember that AI is, in the words of Microsoft’s TOS, “for entertainment only.”

The post AI rewards brand meaning and punishes everything else appeared first on MarTech.