Personalization in marketing: 71% expect it, McKinsey finds

Personalization in marketing pays off when the company has reliable customer data, clear segmentation criteria and permission to use that data, and the same research that measures the return shows where personalization turns into a nuisance. The most quoted figure on the subject comes from McKinsey: 71% of consumers expect personalized interactions and 76% get frustrated when that does not happen (McKinsey, 2021). The sample was 1,013 adults in the United States, surveyed in September 2021, and the 71% adds up those who answered “somewhat agree”, “agree” and “strongly agree” on a six-point scale. The other famous figure from the same article, that faster-growing companies drive 40% more of their revenue from personalization, comes from a subset of 20 consumer goods companies that reported their own share of revenue from personalized marketing. On the other side, a Gartner survey of 1,464 B2B buyers and consumers found that personalization created a negative experience for 53% of them (Gartner, 2025).

In the CRM projects I lead, personalization usually arrives as a request from marketing, with the 71% on the opening slide, and the conversation changes direction once we open the database. The industry field is free text, the contact’s job title has not changed since the record was created and nobody can say which legal basis supports sending to half the list. I checked every number in this text against the original source.

What did McKinsey measure when it says 71% expect personalization?

The article, published in November 2021 by seven of the firm’s authors, presents the Next in Personalization 2021 report. The chart notes carry the method that almost never travels with the number: 1,013 American adults, with a sample weighted to match the US population aged 18 and over. Respondents rated the statement “I expect personalized communications and products/services tailored to my needs from the brands/businesses that I buy from”, and the frustration figure came from a separate statement about irrelevant recommendations.

In the same survey, 76% said personalized communications weigh on their decision to consider a brand, and 78% said they are more likely to buy again from, and to recommend, companies that personalize. The most useful data point for operations sits in a less quoted chart, where consumers rated the importance of each action on a first purchase. Making it easy to navigate in store and online scored 75%, relevant recommendations 67% and messaging tailored to their needs 66%, while addressing the customer by name scored 54%, below almost everything else. For these consumers, personalization means relevance, and a first name in the email subject line sits low on the list.

The aggregated scale also calls for care. Someone who chose “somewhat agree” counts toward the 71% with the same weight as someone who chose “strongly agree”, and the article does not show the split among the three answers. A moderate expectation of relevance calls for an email that addresses the right problem, and a strong demand calls for a data operation that few mid-sized Brazilian companies have.

Where does the 40% more revenue from personalization come from?

The article states the figure in two ways: “companies that grow faster drive 40 percent more of their revenue from personalization” than slower growers, and “companies that excel at personalization generate 40 percent more revenue from those activities” than average players. According to the chart note, McKinsey ran a benchmarking survey in February 2021 with 20 consumer goods companies that have no direct relationship with the end consumer, split the 20 into two halves by the previous year’s revenue growth and asked each what percentage of revenue came from personalized marketing actions or tactics.

The 40%, therefore, compares the share of revenue that ten companies attributed to personalization with the share that another ten attributed, with no external measurement of incremental revenue and no control group. The full survey covered 100 consumer companies and estimated revenue from personalization by company type: between 5% and 10% for those that do not sell directly to consumers, between 10% and 20% for brick and mortar retail and about 25% for digitally native brands. Growing companies can invest more in data and marketing and therefore attribute more revenue to personalization, and the survey design cannot separate cause from effect.

B2B personalization: what does the survey of buying decision makers show?

In 2026 McKinsey published its Global B2B Pulse Survey report, with nearly 4,000 B2B decision makers in 13 countries, Brazil among them (McKinsey, 2026). More than 90% of organizations say they personalize content across email, websites, social media and other touchpoints, and the companies McKinsey calls leaders are four times more likely to use one-to-one personalization, 20% versus 5%. A leader, there, is a company whose respondent reported market share gains above 10% in the year, and a laggard, one that reported losses above 5%.

The detailed chart tells a more modest story. In email campaigns, 15% of respondents from leaders said they use one-to-one personalization, against 7% of laggards and 8% of the total sample. Most B2B companies personalize by segment, and the report places the difference in degree and precision, with leaders using account context, buying history, behavioral signals and next-best-action recommendations.

The same report asked what would lead a decision maker to leave a supplier. Half, 50%, said they would “definitely” or “probably” leave a supplier that does not offer a sales experience tailored to their business, and 39% pointed to too many unwanted communications. The most cited reason was inconsistent information across teams, which I discuss with the rest of the journey data in the digital B2B buyer. The salesperson who quotes a price different from what CS mentioned, the prospecting email sent to someone who is already a customer and the fifth message of the week to the same director are personalization failures that come from teams without access to the same account history.

When does personalization bother the customer?

The Gartner (2025) survey heard from 1,464 B2B buyers and consumers in North America, the UK, Australia and New Zealand, in November and December 2024. The firm reports that personalization generated negative experiences for 53% of customers, and that this group was 3.2 times more likely to regret a purchase and 44% less likely to buy again. Customers who experienced personalization in a recent journey were 1.8 times more likely to pay a premium and, at the same time, 2 times more likely to feel overwhelmed by the volume of information and 2.8 times more likely to feel time pressure.

Gartner places the problem at the moments when the buyer switches tasks, such as moving from searching to selecting a product, when a personalized offer looks irrelevant to someone dealing with questions more complex than the offer solves. The firm contrasts passive personalization, which infers from behavior what to show, with active personalization, in which customers answer questions and share their own context, and says that those who received this approach were 2.3 times more likely to confidently complete critical purchase decisions.

What is the personalization paradox?

Academic research calls the personalization paradox the tension between the value customers see in a tailored offer and their discomfort with the data collection that makes it possible. Three peer-reviewed studies show when each side weighs more.

Awad and Krishnan (2006), in MIS Quarterly, surveyed more than 400 online consumers and found a dilemma: the customers who most value transparency about how information is used are the least willing to be profiled for personalization. The authors recommend designing the offer for those who accept to take part and admitting that the most privacy-sensitive minority will stay out, even with extra transparency features.

Aguirre, Mahr, Grewal, de Ruyter and Wetzels (2015), in the Journal of Retailing, combined a field study on a social network, data from an advertising campaign and three experiments. Click-through rates dropped sharply when customers realized their data had been collected without consent. With overt collection, more personalized ads generated higher click intentions, and trust signals offset part of the negative effect of covert collection, which the authors link to a sense of vulnerability.

Tucker (2014), in the Journal of Marketing Research, analyzed a randomized field experiment on a social network, with ads using personalized text and generic text. Midway through the test, the network gave users more control over their personal information, without changing what advertisers could use, and users became nearly twice as likely to click on personalized ads. Ads that were only targeted, without personalized text, kept the same effectiveness.

Customers respond better to personalization when they know the company has the data, understand what it is for and feel in control of its use, and they react badly when they discover the collection afterward. The studies deal with ads and online retail, and I carry the conclusion over to B2B as a hypothesis, still untested in that context.

Summary table of personalization research

Source	Year	Sample and method	Main finding
McKinsey, Next in Personalization	2021	1,013 US adults, September 2021, six-point agreement scale	71% expect personalization and 76% get frustrated without it (sum of the three positive answers)
McKinsey, Next in Personalization	2021	20 consumer goods companies, February 2021, self-reported personalization revenue	The faster-growing half attributes 40% more revenue to personalized marketing
McKinsey, Global B2B Pulse	2026	Nearly 4,000 B2B decision makers in 13 countries (charts show n = 3,664)	Over 90% personalize; one-to-one at 20% of leaders and 5% of laggards; 50% would switch suppliers without a tailored experience
Gartner	2025	1,464 B2B buyers and consumers, North America, UK, Australia and New Zealand, late 2024	53% had a negative experience, 3.2 times more likely to regret and 44% less likely to repurchase
Awad and Krishnan, MIS Quarterly	2006	Questionnaire with more than 400 online consumers	Those who value transparency are less willing to be profiled
Aguirre and coauthors, Journal of Retailing	2015	Social network field study, campaign data and three experiments	Collection without consent cuts clicks; overt collection makes personalization pay off
Tucker, Journal of Marketing Research	2014	Randomized field experiment with ads on a social network	With more privacy control, personalized ads got nearly twice the clicks

What does personalization require from the CRM in a B2B company?

In a B2B operation, I translate the findings into layers that need to exist in the CRM before any personalization project. The link between each finding and the signal you check in your database is my interpretation.

Layer	Supporting finding	What needs to exist in the CRM	Signal that it is missing
Account data	Inconsistent information is the main reason to switch (McKinsey, 2026)	A single history with deals, contracts and tickets tied to the right company	Duplicate accounts and proposals CS cannot find
Contact data	Relevance weighs more than the customer’s name (McKinsey, 2021)	Job title and buying committee role with an update date	Empty job titles or titles recorded over a year ago
Segmentation	Leaders personalize by account context (McKinsey, 2026)	Industry, size and customer stage in picklists, tied to the ICP	Industry in free text, with dozens of spellings
Journey moment	Passive personalization backfires at task switches (Gartner, 2025)	Deal stage and last interaction visible to marketing	Sales offers sent to someone in technical evaluation
Consent	Overt collection makes personalization pay off (Aguirre and coauthors, 2015; Tucker, 2014)	Legal basis, purpose and data source per contact, with working unsubscribe	Nobody can say where the list came from

The segmentation criteria come from an ICP written with real customer data, the subject of ICP, the ideal customer profile. In the data layers, personalization inherits the database’s errors, and a duplicate account gets two different “personalized” messages on the same morning, a problem I discuss in CRM data quality. The AI agent that writes the nurture emails, covered in AI in B2B marketing, personalizes with whatever it finds in the record.

What Brazil’s LGPD says about personalization data

I present the law for information only, without replacing legal advice. Brazil’s General Data Protection Law (Law 13,709/2018) lists in article 7 the grounds that authorize processing personal data, including the data subject’s consent and the controller’s legitimate interest. For consent, article 8 requires specific purposes, voids generic authorizations and guarantees revocation at any time through a free and easy procedure. For legitimate interest, article 10 limits processing to strictly necessary data and requires transparency measures, and article 18 gives data subjects rights such as access and correction of outdated data.

In practice, the CRM needs to store the contact’s source, the purpose of collection and the legal basis used, and unsubscribe has to work across every channel. Academic research and the law point to the same design, with overt collection and a declared purpose, and the LGPD points related to AI use are in AI and the LGPD.

What these data do not say

The 71% describes US consumers in 2021, in the middle of a pandemic-driven shift in habits, and adds up weak and strong agreement. It does not describe B2B buyers or the Brazilian market, and the 2026 B2B report, which includes Brazil, does not break results down by country in its public material.

The 40% has the weakest base on the list: twenty companies of a single type reported their own share of revenue attributed to personalized marketing, and each one attributes in its own way. The number shows an association between growth and attributed revenue, and nothing in it measures how much personalization caused.

The 2026 B2B report has inconsistencies readers should know about. The text mentions nearly 4,000 decision makers, the charts report n = 3,664, and the personalization chart cites fieldwork in April 2024 while the others cite December 2025. The split between leaders and laggards depends on the market share reported by the respondent.

Gartner’s press release does not describe the comparison base for its multiples or the definition of “negative experience”, and the full report is restricted to clients. The three academic studies work with ads and consumers between 2006 and 2015, before generative AI, and none of the sources measures the effect of personalization on sales cycle, deal size or renewal in B2B contracts. For that question, the most reliable answer is still a test on your own database, with a comparison group.

What leadership does with these numbers

The research justifies treating personalization as an operating capability, with data, criteria and permission, and advises against buying it as a tool feature to meet the expectation on the slide. The roadmap I use with leadership teams has seven steps.

1. Pick one moment and one segment. Start with one point in the journey, such as nurturing accounts in evaluation or renewal communication, and one ICP segment.
2. Audit the fields the message will read. Measure completion and the last update date of each field, and leave out of the first test any field less than half filled.
3. Unify the account history. Marketing, sales and CS need to see deals, contracts and tickets for the same company in a single record.
4. Record source, purpose and legal basis. Define with the legal team the basis for each use and test unsubscribe across every channel.
5. Prefer personalization the customer can see. Ask for context in forms, assessments and conversations and use the answer in the next communication.
6. Measure against a comparison group. Keep part of the database on the standard communication and compare response, pipeline generated and unsubscribes.
7. Add up frequency across teams. Count the messages the same decision maker receives per week from marketing, sales and CS, because too many communications is among the reasons to switch suppliers.

What this means for HR and L&D

Personalization depends on behaviors that happen far from marketing: the salesperson who updates the contact’s job title after the meeting, the CS rep who logs the ticket on the right account and the analyst who checks the legal basis before importing a list. To assess the effect of training, track before and after the completion of the fields personalization uses, the duplicate account rate and the percentage of contacts with a recorded source.

The expectation the research measures turns into results when the company keeps in the CRM the data, segmentation and permission behind each message, and that is the work of the in-company CRM adoption training. Classes are split by role, for leadership, sales, marketing and CS, and take place in the CRM the company already has, with the operation’s real records. The team leaves with the CRM best practices guide and the portfolio and database management manual, written for your operation, with the record, update and consent rules that personalization requires.

Frequently asked questions

Is it true that 71% of customers expect personalization?

The figure comes from a McKinsey survey of 1,013 US consumers, conducted in September 2021, and adds up the answers “somewhat agree”, “agree” and “strongly agree” on a six-point scale. It holds for that audience and that moment, and it does not describe B2B buyers or Brazilian consumers.

Does personalization increase revenue?

The research shows association: at McKinsey, consumer goods companies that grew faster attributed 40% more revenue to personalized marketing, in a subset of 20 companies with self-reported data. None of the sources measures the causal effect on B2B sales, and the reliable way to find out is to test on your own database with a comparison group.

Can too much personalization push customers away?

It can. In Gartner’s survey of 1,464 B2B buyers and consumers, 53% had a negative experience with personalization, with a higher chance of regret and a lower chance of repurchase. The academic studies show that the reaction gets worse when customers discover data collection they did not authorize.

What personalization data does the CRM need in B2B?

A unified account history, the contact’s job title and buying committee role with an update date, segmentation in picklist fields tied to the ICP and, for each contact, the source, purpose and legal basis for processing. Without these items, the personalized message repeats the database’s errors.

Does the LGPD forbid personalized communications?

The law requires processing to have a legal basis, such as consent or legitimate interest, a specific purpose and transparency, and it gives data subjects rights such as access, correction and revocation of consent. Within these rules personalization is possible, and the legal team should validate the basis for each use before launch.

How do you prepare the team to personalize with CRM data?

Train each role on what to record and when, focusing on the fields personalization uses and on the source and consent rules for contacts. Measure the completion of those fields and the duplicate rate before and after the training, in the company’s own CRM.