Companies are using AI in marketing for specific jobs: adapting creative to local audiences, generating product imagery, personalizing ads, and speeding campaign workflows. AI Weekly’s September 28, 2026 roundup counts 24 named deployments, but that is the roundup’s classification—not a representative census. The cases described below come from Google customer stories, Capgemini Research Institute’s 2025 report, and Axios’s reporting on an IBM pilot; their results are company- or publisher-reported, not comparable independent benchmarks.
How widespread is AI use in marketing?
Capgemini Research Institute’s 2025 survey found that 72% of surveyed organizations used generative AI in marketing either extensively or to a limited extent, compared with 37% in its 2023 comparison. The latest wave was based on fieldwork in June–July 2025 and a sample of 1,500 organizations; it should not be read as a rate for every company worldwide. The report also says 77% used generative AI for content creation in 2025, up from 58% in its 2023 comparison.
As an Amazon Associate I earn from qualifying purchases.
Those figures describe reported use, not how much work AI performs or whether the use delivers a business return. The cases below illustrate the range of tasks—and the very different measures companies and publishers use to describe outcomes.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
What are companies using AI to do?
| Organization or campaign | Marketing task | Data or material used | Evidence status in the cited account |
|---|---|---|---|
| PODS with agency Tombras | Neighborhood-specific outdoor ad headlines | Live data and local neighborhood context | Google customer story, 2024 |
| Cadbury | Localized video advertising | Generative AI-created ad variations promoting local stores in India | Capgemini case summary, 2025 |
| IBM | Campaign image generation and variation | Adobe Firefly-generated images and variations | Early pilot, as reported by Axios in 2024 |
| PUMA India | Localized product photography | Product photography customized with Google Imagen | Google customer story, 2024 |
| Radisson Hotel Group | Personalized advertising at scale | Vertex AI and Gemini models with datasets in BigQuery | Google customer story, 2024 |
| Ulta Beauty | Personalized content creation | Collaboration with Adobe; the report also cites Microsoft Copilot | Capgemini case summary and executive comment, 2025 |
| Kraft Heinz | Product-content design and personalization | TasteMaker, a custom retrieval-augmented generation engine | Capgemini case summary, 2025 |
| Standard Chartered | Campaign concept development and design execution | ChatGPT for concepts and Adobe Firefly for design | Capgemini case summary, 2025 |
| Formula E | Condensing race commentary for audio audiences | Existing two-hour race commentary and Google Cloud generative AI | Google customer story; no quantified outcome stated in the cited passage |
| Globo | Personalizing streaming content | Google Cloud AI | Google customer story; no quantified outcome stated in the cited passage |
| Cook Medical | Internal information access, content creation, and decision support | A generative AI chatbot using company data | Executive account quoted by Capgemini, 2025 |
| Airtel Business | Customer intelligence, recommendations, and personalized messaging | AI-supported customer intelligence | Executive account quoted by Capgemini, 2025 |
Which deployments report measurable outcomes?
PODS: adapting creative to local context
Google’s September 2024 customer story describes PODS and agency Tombras using live data to adapt truck advertisements by New York City neighborhood. Google reports that the campaign covered all 299 neighborhoods in 29 hours and generated more than 6,000 headlines. These figures describe that campaign’s output, not a general AI production rate.
#1 Best Overall
Cadbury: localizing video ads in India
Capgemini Research Institute’s 2025 report says Cadbury used generative AI to make localized video ads featuring a Bollywood star and promoting local stores across India. Its case summary attributes reach of 140 million or more, more than 2,500 unique ads, and a 32% engagement spike to the campaign.
IBM: an early Firefly pilot
Axios reported on March 6, 2024, that IBM used Adobe Firefly in an early marketing pilot to create 200 images and more than 1,000 variations. Axios said the campaign achieved engagement 26 times higher than the benchmark for those efforts. This is a reported result from an early pilot and its specific benchmark, not proof that AI campaigns generally produce that lift.
Rank #2
PUMA India: product photography
Google’s 2024 account says PUMA used Imagen to customize product photography for its website. PUMA India reported a 10% increase in click-through rate. The account describes localization and time savings as aims; it does not quantify a time-saving result.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →Radisson: personalized advertising
Google reports that Radisson Hotel Group worked with Accenture and Google Cloud, using Vertex AI and Gemini models alongside datasets in BigQuery to personalize ads at scale. In Google’s 2024 customer story, the group’s ad-team productivity is reported as 50% higher and revenue from AI-powered campaigns as more than 20% higher. These are attributed customer-story outcomes; the account does not establish that the figures are comparable with other companies’ metrics.
Ulta Beauty: content productivity
Capgemini’s 2025 report describes Ulta Beauty producing personalized content at scale through collaboration with Adobe. It quotes CTIO Mike Maresca saying generative AI was improving productivity by around 30%, owing to tools such as Microsoft Copilot. That is an executive-reported estimate, not a separately specified productivity test.
Kraft Heinz: faster product-content design
Capgemini’s 2025 report describes TasteMaker, Kraft Heinz’s custom retrieval-augmented generation engine for scaling content creation and personalization. The report says the product-content design timeline fell from weeks to hours—an eightfold reduction in the case summary.
Rank #4
Standard Chartered: more campaigns and assets
Capgemini’s 2025 report says Standard Chartered used ChatGPT to develop marketing concepts and Adobe Firefly to execute designs. From 2023 to 2024, the report records 150% growth in total campaigns, 133% growth in total assets, and a 21% reduction in average working days per campaign. Reducing campaign time from 21 days to five was the bank’s stated target, not a reported achieved result.
What do the other examples show?
Formula E: turning existing material into a new format
Google describes Formula E using Google Cloud generative AI to condense two-hour race commentary into a two-minute podcast in any language. The example illustrates repackaging existing content for different formats and audiences; Google’s cited passage gives no quantified business outcome.
Globo: tailoring streaming content
Google’s 2024 account describes Globo using Google Cloud AI to personalize streaming content. The cited passage does not specify a numerical result, so the example supports the use case—not a claim about its commercial impact.
Cook Medical: using company information
Capgemini’s 2025 report quotes Global Marketing Director Terrence Wiggins describing a generative AI chatbot that uses real-world company data. He says the company uses it for information access, content creation, predictive analytics, and decision-making. The report does not attach a quantified marketing result to this account.
Airtel Business: tailoring recommendations and messages
Capgemini’s 2025 report quotes Airtel Business’s Kaustubh Chandra saying AI has helped improve customer intelligence for product recommendations and messaging, enabling campaigns and pitches tailored to customer personas. The cited excerpt does not state Chandra’s role or provide a numerical outcome.
How should these results be compared?
They should not be ranked by percentage alone. A click-through change, campaign engagement result, staff-productivity estimate, revenue figure, and reduction in design time measure different things. Even two percentages with the same label may use different baselines, definitions, audiences, or measurement windows.
- Identify the task. Generating many creative variants is different from personalizing an ad, improving product imagery, or shortening a design workflow.
- Check the maturity. IBM’s example is explicitly described as an early pilot. The other accounts describe customer use or campaign cases, but do not all provide a consistent production-status label.
- Look for the inputs and integration. Examples draw on live neighborhood data, product imagery, company information, or datasets connected to marketing systems. Those inputs help explain what a system could do in that particular workflow.
- Keep attribution visible. Google, Capgemini, and Axios are reporting customer or case-study claims. The figures are useful evidence that specific organizations reported outcomes, not independent forecasts for another team.
Taken together, these deployments show AI being used as part of marketing and content workflows: to create or adapt assets, connect content to data, and transform material for another audience or format. They do not show that AI has replaced marketing teams, nor do selected customer stories establish what a typical organization should expect.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
