Retailers Are Spending More on AI Than Other Consumer Companies and Getting Less Back

As featured in TotalRetail, written by Carrie Shea

New research across 300 consumer industry executives reveals a striking paradox: retail companies outspend nonretail peers on commercial AI, yet expect significantly lower returns.

I’ve spent 25 years working with consumer companies on revenue growth. I’ve sat in boardrooms where executives present beautifully designed artificial intelligence road maps. I’ve watched the same executives walk out of those meetings and continue doing exactly what they were doing before. The decks are impressive. The results, especially in retail, are not.

We recently surveyed more than 300 consumer industry executives, managers, and specialists on how their companies are applying AI to commercial functions, with a sample split nearly evenly between retail and nonretail industries. What we found was striking. Not because of what companies are doing, but because of what they openly admit they’re not doing.

The term that came up most often in our executive interviews was AI theater. Companies are piloting AI where results are easy to demonstrate to boards, not where they drive the most revenue and EBITDA growth. It’s not malicious. It’s human nature. But for retail in particular, the cost measured in lost revenue growth and EBITDA is enormous, and the gap with nonretail peers is widening.

Retail Spends More and Expects Less Back

Here’s what stopped me when we cut the data by industry: retail companies spend more on commercial technology and AI than their nonretail counterparts. The average retail respondent reported spending approximately $19 million on commercial tech and AI tools last fiscal year. Nonretail companies spent roughly $16 million. However, that spending advantage is not translating into returns. Retail respondents in our study expect an average of approximately $18 million in EBITDA improvement from AI implementation over the next 12 months to 24 months. Nonretail respondents expect nearly $28 million, a 55 percent higher expected return, despite lower investment.

That gap doesn’t happen by accident. It reflects choices about where AI investment money is going.

The Paradox at the Heart of AI Investment

The broader finding from our research that consistently surprises people: the consumer companies generating the highest returns from commercial AI are spending less than those generating the lowest returns.

High-impact companies spend an average of $15 million on commercial technology and AI. Companies not investing in high-impact use cases spend $20 million. On incremental AI investment beyond what’s already embedded in existing tech stacks, the gap is even more pronounced: $2.5 million vs. $10 million. Four times more spending for materially worse results.

This is not a technology problem. It is a prioritization problem.

Consumer companies, retail companies especially, are concentrating AI investment in marketing content development, ad creative optimization, and customer segmentation. These are not bad investments. But our research shows they deliver materially lower returns than the use cases companies are largely ignoring.

Where the Revenue Growth Actually is

The highest-impact commercial AI use cases in our study are not mysterious or experimental. They’re well understood. Companies are simply not investing in them at scale.

AI-powered commercial planning and analytics is delivering 40 percent revenue growth improvement for companies that use it. Fewer than 8 percent of consumer companies are applying AI here. AI-driven innovation is delivering 23 percent revenue growth improvement. Revenue growth management AI is delivering 16 percent improvement. In both cases, adoption is minimal.

The single highest-impact finding in our entire study: AI-powered customer onboarding delivers a 41 percent improvement in customer lifetime value. This is the most powerful commercial AI application we found. Ninety-one percent of consumer companies are not pursuing it.

What Makes Retail Distinct

When we looked specifically at retail vs. nonretail behavior, a few patterns stood out.

Retail companies are more likely than nonretail peers to buy AI tools and configure them rather than build solutions in-house. That’s not inherently wrong, but it does shape what gets implemented. Off-the-shelf tools make it easier to deploy quickly in familiar areas like content, segmentation, and ad optimization, and harder to pursue the more complex, high-return use cases that require process redesign and integrated data.

Retail companies also cited vendor lock-in as a disproportionate concern compared to nonretail peers. That anxiety makes sense given the complexity of retail technology stacks, but it may also be reinforcing a tendency to cluster around safer, more interchangeable tools rather than committing to the deeper implementations that drive outsized returns.

Nonretail companies, meanwhile, reported doubts about whether their executives truly understand AI, which presents its own problem. But that skepticism sometimes correlates with more rigorous business case development and more willingness to question where the investment is actually going.

One area where retail is ahead: adoption of AI agents. Retail companies are more likely than nonretail peers to be using AI agents for critical commercial tasks. That experimentation is real and worth building on. The opportunity is to direct that same appetite for agentic AI toward the use cases that move revenue, not just the ones that automate existing workflows.

Why Companies Keep Making the Wrong Choices

Our executive interviews surfaced honest and somewhat uncomfortable explanations for why companies gravitate toward lower-impact AI use cases.

Speed-to-value consistently beats total value. Use cases that show results in weeks win budget battles over use cases that require six months to 24 months to fully realize their impact, even when the long-term returns are dramatically higher.

Vendor demos are doing too much work. Several executives acknowledged that in the absence of strong internal data and rigorous business cases, teams are being swayed by polished presentations. This is not a criticism of vendors. It reflects how most companies are approaching AI investment decisions.

The highest-impact use cases are genuinely harder to implement. Revenue growth management, commercial planning, and innovation AI require cross-functional integration, end-to-end process redesign, and more complex datasets. They’re harder to get off the ground. That’s precisely why they remain underinvested, and why the companies that do invest here are building a competitive advantage that’s difficult to replicate quickly.

What the Winners Are Doing Differently

The consumer companies generating the highest returns from commercial AI share four characteristics:

• They are experimenting with AI agents, not just AI tools.
• They believe AI will expand their workforce rather than reduce it, and they expect that shift to happen sooner than their competitors anticipate.
• They are significantly more confident in their ability to build rigorous AI business cases internally.
• They prefer to buy proven solutions rather than build proprietary AI. The most sophisticated companies are not necessarily building their own.

4 Actions to Close the Gap

For retail companies ready to stop doing AI theater and start generating real revenue growth, our research points to four actions.

1. Benchmark your commercial AI spending against actual revenue growth impact.

Challenge every AI investment with the question: What is the revenue growth business case, and how does it compare to the highest-impact use cases available to us? Most companies cannot answer this question rigorously. That’s where the problem starts.

2. Move from point solution pilots to end-to-end process redesign.

Fragmented AI experimentation does not create durable competitive advantage. Consumer companies that map their end-to-end commercial processes and reimagine AI applications across the full value chain will generate returns that point solution pilots cannot match.

3. Standardize and accelerate adoption.

Develop repeatable AI toolkits: audits, benchmarks, prioritized use case libraries, vetted vendor frameworks, and standardized workflows. The winners in our research are not reinventing the wheel with every implementation; they’re building replicable systems.

4. Prioritize the use cases that drive revenue growth and margin expansion, not the ones that are easiest to demonstrate.

This requires internal discipline and leadership alignment. It also requires the willingness to make harder implementation choices. That is where sustainable competitive advantage lives.

The Window is Open, But Not Indefinitely

Retail companies are sitting on data infrastructure, established commercial functions, and AI tools that are increasingly accessible. The returns from getting this right are measurable in weeks.

The companies that close the AI investment gap now will widen their performance advantage in ways that will be difficult to close later. Those that continue spending more than their peers while targeting lower-impact use cases will find themselves competing from a structural disadvantage they created themselves.

The research is clear. The choices are available. The only question is whether retail companies are ready to stop performing AI theater and start using it.

To access the full research report or to schedule time to review the findings in detail, please contact us.