86 AI Personalization Statistics for 2026 (Every Number Sourced)
The most quoted statistic in personalization does not publish a sample size.

The most quoted statistic in personalization does not publish a sample size.
“71% of consumers expect companies to deliver personalized interactions” appears in vendor decks, agency pitches and conference keynotes every week. It comes from a McKinsey article published in November 2021, and I went to that article to pull the methodology. The sentence reads, in full: “Our research shows that 71 percent of consumers expect companies to deliver personalized interactions.” There is no sample size on the page. No fielding window. No country. No link to an underlying survey. The number that anchors an entire industry’s business case is published as a bare assertion.
That is not an argument against the figure. McKinsey has no reason to invent it. It is an argument for knowing what you are actually holding when you put a number on a slide, because a good share of the personalization statistics in circulation are older, narrower, or more heavily caveated than the pages quoting them let on.
So I took 86 AI personalization statistics back to the organization that produced each one. Salesforce, HubSpot and the Marketing AI Institute for what marketers are doing. McKinsey, Adobe and Litmus for what consumers say they want. Pew Research Center and Cisco for the privacy backdrop. BCG for the revenue forecasts, clearly labelled as forecasts. Klaviyo’s audited SEC filings for the only numbers here that a regulator has looked at.
Several claims did not survive that process. Some were misattributed to the wrong report edition, some to the wrong survey inside the right report, and three failed for a different reason each, from having no traceable source to resting on four customer interviews. Each exclusion and each correction is listed in the Sources and Methodology section at the bottom, with the reason.
If you only quote four numbers from this page, quote these. Each is restated in context further down with its full source note.

1. 78% of marketers say they need more personalized content than they are able to produce (Salesforce, State of Marketing, 10th edition, n=4,450 marketing decision makers, fielded 8 October to 17 November 2025).
2. 98% of marketers hit barriers to personalization, and data issues are the most common culprit (Salesforce, 10th edition).
3. 42% of customers trust businesses to use AI ethically, down from 58% in 2023 (Salesforce, State of the AI-Connected Customer, 7th edition, n=15,015 consumers plus 1,570 business buyers across 18 countries, fielded 26 July to 20 August 2024).
4. Only 26% of consumers describe their digital experience with a brand they already have a relationship with as “excellent” (Adobe and Econsultancy, 2024 Digital Trends Report, consumer survey n=6,800, fielded 1 to 13 February 2024).
Adoption is no longer the interesting question. Three separate 2025 and 2026 surveys put AI use among marketers between 49% and 75%, depending on how loosely each defines “use.” What separates them is the gap between having AI and having the data to point it at, which is where the Salesforce numbers get uncomfortable.
5. 75% of marketers are already dabbling in some form of AI, whether predictive, generative or agentic (Salesforce, 10th edition). Salesforce’s own word is “dabbling,” which is doing a lot of work in every deck that quotes this as adoption.
6. Only 13% of marketers are currently using agentic AI (Salesforce, 10th edition).
7. 75% of marketers are turning to AI specifically to close the personalized-content gap (Salesforce, 10th edition).
8. 49% of marketers already use AI to tailor content (HubSpot, 2026 State of Marketing Report, more than 1,500 global marketers, published January 2026). HubSpot calls this the most-adopted trend of the year.
9. 19.20% of marketers are already using AI agents to automate marketing initiatives end to end (HubSpot 2026).
10. 26% of marketing and business leaders place themselves in the “Integration” phase, embedding AI into workflows and processes (Marketing AI Institute and SmarterX, 2025 State of Marketing AI Report, 1,882 respondents, fielded February to April 2025).
11. 61% of marketers believe marketing is experiencing its biggest disruption in 20 years because of AI (HubSpot 2026).
12. 75% of marketing organizations were either experimenting with or had fully implemented AI two years earlier (Salesforce, State of Marketing, 9th edition, n=4,850, fielded 5 February to 12 March 2024).
13. 91% of marketers say personalization improves engagement (HubSpot 2026).
14. 93% saw a great impact on marketing-driven leads or purchases from personalized experiences (HubSpot 2026).
15. 32.82% of marketers say AI tools are saving their teams 10 to 14 hours per week (HubSpot 2026).
16. 75% of marketers who have AI are satisfied with their ability to connect touchpoints, against 60% of those without it (Salesforce, 10th edition). A 15-point gap is real, but it is self-reported satisfaction, not measured performance.
17. High-performing marketers are 2.2 times more likely than underperformers to have optimized for AI search (Salesforce, 10th edition).
18. 84% of marketers confess to running generic campaigns (Salesforce, 10th edition). This is the number I would put next to any personalization vendor’s pitch deck.
19. 69% of marketers still struggle to respond to customers promptly (Salesforce, 10th edition).
20. 46% report lacking the customer preference data needed to provide content relevant to customer needs (Salesforce, 10th edition).
21. Only 31% of marketers are fully satisfied with their ability to unify customer data sources (Salesforce, 9th edition).
If you are working through which tools actually close that gap, the AI marketing use cases breakdown covers where the output holds up and where it does not.
This is where the canon lives, and where it is oldest. Four of the most-cited consumer expectation figures in personalization come from one McKinsey article published in November 2021, which publishes no methodology. The freshest large-sample consumer work here is Adobe’s, and Adobe runs two separate surveys inside one report, which is the single most common misattribution I found.
22. 71% of consumers expect companies to deliver personalized interactions (McKinsey, The value of getting personalization right or wrong is multiplying, November 2021). The article states no sample size, no fielding date and no country.
23. 76% of consumers get frustrated when personalization does not happen (McKinsey, November 2021).
24. Companies that grow faster drive 40% more of their revenue from personalization than their slower-growing counterparts (McKinsey, November 2021). This is a share-of-revenue measure. It is routinely misquoted as “grow 40% faster,” which the source does not say.
25. Shifting US industries to top-quartile performance in personalization would generate over $1 trillion in value (McKinsey, November 2021). A modelled opportunity, not a measured one.
26. Personalization can reduce customer acquisition costs by as much as 50%, lift revenues by 5% to 15%, and increase marketing ROI by 10% to 30% (McKinsey, What is personalization?, May 2023). The explainer states these three figures with no inline citation and no named study behind them.
27. Over 75% of consumers are turned off by content that does not feel relevant (McKinsey, Discussing the future of AI-powered personalization, July 2025). This is an edited interview co-published with Jasper, a generative AI vendor, and it publishes no new dataset.
28. 80% of consumers consider consistent experiences across different online channels “important” or “critical” (Adobe 2024 Digital Trends, consumer survey n=6,800, fielded 1 to 13 February 2024, 13 countries).
29. 70% assign similar ratings to personalized product recommendations (Adobe 2024).
30. 91% of consumers say responsible data use is either “important” at 28% or “critically important” at 63% (Adobe 2024).
31. Only 26% describe their digital experience with a brand they already have a relationship with as “excellent” (Adobe 2024). The same report separately surveyed 8,600 executives between 1 January and 19 February 2024. The consumer figures above are not from that executive sample, and pages that attribute them to n=8,600 have merged two different surveys.
32. 80% of consumers are more likely to make a purchase when brands offer personalized experiences (Epsilon, The Power of Me, n=1,000 US consumers aged 18 to 64, published January 2018). This is the oldest number on the page and it predates generative AI entirely. It is included because it remains widely cited, and it should be dated every time it is used.

33. Only 25% of baby boomers find personalization “extremely” or “very important” (Litmus, 2024 State of Email in Lifecycle Marketing, n=1,000 US consumers, published November 2024).
34. Nearly 60% of millennials and Gen Z find it “extremely” or “very important” (Litmus, November 2024).
35. 51% of Gen Z are “very” or “somewhat comfortable” with companies using their data (Litmus, November 2024).
36. 49% of millennials say the same (Litmus, November 2024).
37. Only 20% of baby boomers are comfortable with it (Litmus, November 2024). The boomer-to-Gen-Z spread is 31 points, which is wider than any vendor-versus-vendor gap in this entire dataset.
Pew Research Center is the only non-vendor primary in this corpus, which makes it the most useful. Everything Salesforce, Cisco and Adobe report about privacy sentiment corroborates Pew directionally, but Pew is the one that does not sell a customer data platform.

38. 81% of Americans say they feel very or somewhat concerned about how companies use the data they collect about them (Pew Research Center, Views of Data Privacy Risks, Personal Data and Digital Privacy Laws, n=5,101 US adults, fielded 15 to 21 May 2023).
39. 71% say the same about government use of data, up from 64% in 2019 (Pew, 2023).
40. 73% feel they have little or no control over what companies do with their data (Pew, 2023).
41. 67% say they understand little to nothing about what companies are doing with their personal data, up from 59% (Pew, 2023).
42. 72% of Americans say there should be more regulation of what companies can do with personal data, and just 7% say there should be less (Pew, 2023). The remaining 21% either want no change or are unsure.
43. 56% frequently click “agree” to privacy policies without actually reading them (Pew, 2023). Consent, in practice, is not evidence of comfort.
44. 42% of customers trust businesses to use AI ethically, down from 58% in 2023 (Salesforce, State of the AI-Connected Customer, 7th edition, n=15,015 consumers plus 1,570 business buyers across 18 countries, fielded 26 July to 20 August 2024). A 16-point fall in twelve months, measured by a company that sells AI.
45. 70% of Americans who have heard about AI have little to no trust in companies to make responsible decisions about how they use it in their products (Pew, 2023). Pew, with no product to sell, lands in the same place.
46. 23% of consumers describe themselves as regular users of generative AI, almost double the 12% of a year earlier (Cisco, 2024 Consumer Privacy Survey, more than 2,600 adults across 12 countries, fieldwork conducted June 2024). Use is climbing while trust is falling, which is the tension every personalization roadmap has to survive.
47. More than 75% of consumers say they will not purchase from an organization they do not trust with their data (Cisco, 2024).
48. 53% of respondents said they were aware of their country’s privacy laws, the first time a majority has said so since the survey began (Cisco, 2024).
49. In 2019, only 36% were aware (Cisco, 2024). A 17-point rise in awareness over five years.
50. 84% of generative AI users were concerned their data could be shared (Cisco, 2024).
51. They entered it anyway: 37% put health information into generative AI tools, 36% work information and 29% financial information (Cisco, 2024). Stated concern and actual behaviour diverge by a wide margin, which is worth remembering before treating any survey’s privacy answers as a forecast of what people will do.
This is the weakest evidence base on the page and the one quoted with the most confidence. Population-level measured ROI for AI personalization barely exists. What exists is consultancy forecasting, self-reported marketer satisfaction, and vendor case studies. All three are useful. None of them is a benchmark.
52. BCG projects that over the next five years, $2 trillion in revenue will shift to companies that understand how to create personalized experiences (BCG, October 2024). Forward-looking language is not optional here. BCG sells personalization transformation consulting.
53. Personalization leaders grow revenue roughly 10 percentage points faster annually than laggards (BCG, October 2024). The leader and laggard tiers are defined by BCG’s own Personalization Index of 200 brands across seven sectors.
54. Top retailers on that index could achieve an estimated $570 billion in incremental growth by using first-party data (BCG, Retail Spotlight, November 2024). Also a projection.
55. Brands integrating advanced digital technology and proprietary data saw revenue increases of 6% to 10% (BCG, Profiting from Personalization, May 2017).
56. That was two to three times faster than companies that did not (BCG, 2017).
57. BCG forecast an $800 billion revenue shift over five years to the 15% of companies that get personalization right, in retail, health care and financial services alone (BCG, 2017). This is a 2017 forecast about a window that has now closed, published before generative AI existed commercially.
58. 70% of organizations report their personalization metrics have somewhat or significantly improved over the past three years (Adobe, 2026 AI and Digital Trends Report, Adobe with Oxford Economics, 3,000 executives and practitioners plus 4,000 customers, fielded October to November 2025).
59. 64% report improved lead generation (Adobe 2026).
60. 59% report improved retention (Adobe 2026).
61. 57% of organizations say their digital customer experience is on par with or behind peers, and only about 36% consider themselves ahead (Adobe 2026). Self-rated maturity where most respondents put themselves at average or worse is unusual, and more believable for it.
Every report in this corpus, from three competing vendors, lands on the same conclusion: the models are ready and the plumbing is not. That is the most reliable finding in the whole dataset, precisely because Salesforce, Adobe and Twilio Segment all sell different fixes for it.

62. 89% of organizations have the cloud-based technology to support generative AI (Adobe 2026).
63. Only 51% have cloud-based technology for agentic AI (Adobe 2026). A 38-point drop-off between the AI everyone has deployed and the AI everyone is planning for.
64. Only 47% of organizations are using generative or agentic AI for journey design or omnichannel activation (Adobe 2026).
65. 16% have agentic AI embedded organization-wide for customer support (Adobe 2026).
66. 13% have it embedded for brand discovery and search (Adobe 2026).
67. 72% of companies are using a customer data platform for personalization (Twilio Segment, State of Personalization Report 2024, n=521 business leaders across 12 countries, fielded 8 April to 5 May 2024).
68. 48% are using a data warehouse (Twilio Segment, 2024).
69. 86% of business leaders expect a significant shift from reactive to predictive personalization across their industry (Twilio Segment, 2024).
70. 89% of leaders predict ethical AI use will be a competitive business advantage (Twilio Segment, 2024).
71. 54% of brands plan to implement a data platform with robust privacy controls (Twilio Segment, 2024). The 35-point gap between believing ethical AI wins and actually building for it is the most honest number Twilio Segment published.
72. By 2025, 59% of decision makers expected their teams to be using AI daily, and 91% at least weekly (Twilio Segment, 2024).
73. 92% of businesses were already using AI-driven personalization to drive business growth (Twilio, State of Personalization Report, fourth annual edition, 3,001 consumers plus 500 business leaders, published 2 May 2023). This figure is widely attributed to the 2024 report. It is not in the 2024 report. It comes from the 2023 edition, and “using” includes pilots.
Channel data is where the abstraction stops. A marketer can tell you whether email is personalized far more reliably than whether “the customer journey” is.

74. On average, marketers are able to fully personalize across five channels; high performers manage six and underperformers only three (Salesforce, 9th edition, n=4,850, fielded February to March 2024).
75. Mobile messaging is the most personalized channel at 57% (Salesforce, 9th edition).
76. Email follows at 54% (Salesforce, 9th edition).
77. Social media sits at 52% (Salesforce, 9th edition), with audio at 43%, organic search at 42% and TV or OTT at 41% trailing behind.
78. 45% of teams currently use AI in email marketing, and a further 21% want to start (Litmus, 2024 State of Email Innovations Report, nearly 1,000 marketers, fielded 29 January to 29 February 2024).
79. 51% of marketers require over two weeks to produce a single email (Litmus, November 2024). That production floor, not model quality, is what most personalization programmes are actually limited by.
80. 72% of marketers do not know their email ROI (Litmus, 2024 State of Email Innovations). Personalization ROI claims in email should be read against that.
81. 67% of consumers report receiving too many emails, rising to 81% among baby boomers (Litmus, November 2024).
More channel-level benchmarks sit in the email marketing statistics roundup and the content marketing statistics page.
Every “state of” report above is vendor-sponsored self-report. Klaviyo went public in September 2023, so its customer counts and revenue are filed with the SEC and audited. For platform-level adoption, this is the only evidence here that carries a legal penalty for being wrong.

82. Klaviyo served over 193,000 customers as of 31 December 2025, up from 167,000 a year earlier (Klaviyo, Form 10-K FY2025).
83. Revenue grew 31.6% year over year, from $937.5 million in 2024 to $1,234.0 million in 2025 (Klaviyo 10-K). Growth slowed from the 34.3% it posted in 2024, when revenue rose from $698.1 million in 2023.
84. 3,912 customers generated over $50,000 of annual recurring revenue as of 31 December 2025, up from 2,850, representing 37% growth (Klaviyo 10-K). The upmarket cohort is growing faster than the customer base as a whole.
85. Net losses narrowed from $46.1 million in 2024 to $31.8 million in 2025 (Klaviyo 10-K).
86. Gross profit grew 28.7%, from $716.2 million in 2024 to $921.5 million in 2025 (Klaviyo 10-K). Gross profit is growing slightly slower than revenue, which is the cost of serving larger customers.
Platform growth proves that businesses are buying personalization infrastructure. It does not prove the personalization works, and Klaviyo’s own attributed-value metric is its own methodology rather than an audited figure. The Klaviyo review covers what the product does and does not do for a content team.
The figure is real and correctly quoted, but it is not what most people think it is. It comes from a McKinsey article published in November 2021 that states “Our research shows that 71 percent of consumers expect companies to deliver personalized interactions” without publishing a sample size, a fielding date or a country. McKinsey has continued to reproduce the number in 2023 and 2025 pieces without publishing a new consumer panel. Cite it as a 2021 McKinsey figure with unstated methodology, not as a current measurement.
AI personalization is the use of machine learning to tailor content, product recommendations and customer interactions to individual behaviour and purchase history, in real time, across every touchpoint. In practice it covers three different things that get spoken about as one: predictive models that segment customers, recommendation engines that surface products, and generative AI that writes the personalized content itself. That definitional spread is why the adoption percentages on this page range from 13% to 92%. Each survey draws the boundary in a different place, so check what a report counted before putting its number next to another one.
For privacy and AI trust, Pew Research Center’s “How Americans View Data Privacy” is the strongest source in this corpus: n=5,101 US adults, address-based recruitment, fielded 15 to 21 May 2023, and Pew sells nothing. For marketer adoption, Salesforce’s 10th State of Marketing report is the freshest large sample at n=4,450, fielded October to November 2025. For platform-level scale, Klaviyo’s SEC filings are audited and cannot be spun.
Use them, but name the sponsor inline. Salesforce, Adobe, HubSpot, Twilio Segment, Litmus, Klaviyo, BCG and McKinsey all sell into the personalization stack, and each has a structural incentive to find that personalization works, data is the bottleneck and AI is the answer. Adobe recruits part of its executive panel from its own customer lists. The Marketing AI Institute states plainly that its respondents come from its AI-education audience. Where three competing vendors independently find the same directional result, as they do on the data-layer bottleneck, that agreement is worth more than any single absolute percentage.
Adobe found 80% of consumers consider consistent cross-channel experiences important or critical, while only 26% describe their experience with a brand they already deal with as excellent. Salesforce found 84% of marketers admit to running generic campaigns and 98% hit barriers to personalization. Both sides of that gap are self-reported, from different populations, but every report in this corpus finds it.
Per Salesforce’s ninth State of Marketing report, mobile messaging leads at 57%, then email at 54% and social at 52%. Audio at 43%, organic search at 42% and TV or OTT at 41% lag. Marketers fully personalize across five channels on average, high performers across six and underperformers across three.
Yes, with forecast framing. BCG’s October 2024 press release states that over the next five years, $2 trillion in revenue will shift to companies that understand how to create personalized experiences. That is a projection built on BCG’s own Personalization Index of 200 brands across seven sectors, published alongside the book Personalized by Mark Abraham and David Edelman. Write “BCG projects,” never “personalization generated.” BCG sells personalization consulting, and that belongs in the sentence.
The honest answer is that nobody has measured it at population level. What exists is self-report: Salesforce found 75% of marketers with AI are satisfied with their ability to connect touchpoints against 60% without it, and HubSpot found 32.82% of marketers say AI saves their teams 10 to 14 hours a week. Litmus supplies the useful baseline, which is that 51% of marketers need more than two weeks to produce a single email. Time saved is the best-evidenced benefit. Revenue lift is not.
Every figure above was traced to the organization that produced it and checked against the primary publication, not against an article quoting it. Where a report is gated, the vendor’s own press release or newsroom page was used and is linked. Vendor sponsorship is disclosed inline rather than in a footnote. Sample sizes and fielding windows are stated as the primary source states them, and where a source does not publish them, that absence is stated rather than filled in.
Three claims in wide circulation were excluded. Each failed a different test:
Six corrections were made to figures that are commonly published with the wrong attribution:
One figure was dropped for lack of a primary source rather than corrected. Klaviyo’s “18.2% of customers using SMS” does not appear in the FY2024 Form 10-K, which was read in full for this article.
Figures from before 2024 are retained only where no successor study exists, and each is date-stamped in the sentence that carries it. That applies to Pew (May 2023), McKinsey (November 2021 and May 2023), BCG (May 2017) and Epsilon (January 2018).
Last verified: 9 September 2026.

Hey, I’m Chintan, a creator and the founder of Elite Content Marketer. I make a living writing from cafes, traveling to mountains, and hopping across cities. Join me on this site to learn how you can make a living as a sustainable creator.
View Profile →
109 webinar statistics for 2026 from 7 million registrations: attendance runs 40% to 57%, and average watch time is 26 minutes of a 68-minute session.

80+ prompt engineering statistics for 2026, each traced to its primary source: 69% of LLM input tokens are system prompts, not the user’s question.

80 AI workflow automation statistics for 2026, each traced to its source: 88% of organizations use AI, 37% can find any EBIT impact, 6% are high performers.