One-to-one marketing treats each customer as a market of one. The offer, the message and the timing follow what the company knows about that person. The idea is more than 30 years old, and it is easier to run now than it was then. The evidence says it pays in some places and adds little in others.
What is one-to-one marketing?
One-to-one marketing means tailoring marketing to each customer from data on that customer. Don Peppers and Martha Rogers launched the idea in their 1993 book, The One to One Future (Don Peppers). Two years later, with Joseph Pine, they made the case in Harvard Business Review. Customers, they wrote, want “exactly what they want” (HBR, 1995). They wrote that customers “do not want more choices.” Technology, they said, now lets companies give people what they want, when, where and how they want it.
Their argument rested on two shifts. Databases could hold data on each customer’s needs and preferences. Flexible production could customize goods at a low cost. Most managers, they said, still saw the world through mass marketing and mass production.
Their answer was a learning relationship. Each contact teaches the firm about the customer. The firm uses that to serve the customer better next time. The article tells readers to start one with each customer.
How is one-to-one marketing different from mass and segment marketing?
The difference is the unit of treatment. Mass marketing sends one message to everyone. Segment marketing sends a message to each group. One-to-one marketing sends a message to a person.
| Approach | Unit | What changes by unit | Data it needs |
|---|---|---|---|
| Mass marketing | Everyone | Nothing | Market size |
| Segment marketing | A group | Message and offer by group | Group traits |
| One-to-one marketing | A person | Offer, message, channel and timing | Each person’s history and preferences |
In practice the lines blur. Most programs run several layers at once. A welcome series is mass. A tier benefit is segment. A next-best offer is one-to-one.
How does one-to-one marketing work?
It works as a loop. The firm collects data on each customer. It picks the best offer for that person. It sends it, reads the result and learns. A click or a purchase becomes the data that tunes the next offer.
The data comes from two places. One is behavior the program records, such as purchases and visits. The other is what members tell you, which is zero-party data. Starbucks asks members to opt in before it sends special offers by email (Starbucks terms).
Start small, with five steps:
- Pick one decision. The next offer, the send time or the channel.
- Check the data. You need identified purchase history for the people you target.
- Test against segments. Beat the segment offer before you scale.
- Hold out a group. Measure the lift it shows.
- Honor objections. Wire the opt-out into every send.
Does one-to-one marketing pay off?
It can pay, and the gain differs by place. Zhang and Wedel studied customized promotions at three levels: mass market, segment and individual. They tested about 300 conditions in online and offline stores (Zhang and Wedel, 2009).
Better-tuned promotions raised profit for every type. The extra gain from one-person offers was small overall. It was smaller still in offline stores. In online stores, one-person offers paid where shoppers respond to deals. In stores, few shoppers redeemed the offers, which held results back.
Zhang and Wedel also found that loyalty offers earn more online than offline. Competitive offers do the reverse. For stores, they advise making mass offers the main program. Add tailored offers in suitable categories.
Targeting needs care too. Eva Ascarza ran field experiments on retention. Customers at the highest risk of leaving were not always the best targets. Her advice is to target by response to the offer, whatever the risk (Ascarza, 2018). The author calls the gain significant.
Measure lift. Redemptions alone mislead, because an offer can go to a customer who would have bought anyway. Compare the offer group with a random holdout. The gap is the lift.
Then price the lift. Divide the cost of the offers by the extra buyers they won. Compare that with the margin an extra buyer brings. Run the same test for each layer. Compare mass, segment and one-person offers on matched groups. That shows where the extra build cost pays.
Werner Reinartz and V. Kumar add a caution on loyalty and profit. Much of the common wisdom about retention, they wrote, is bunk (HBR, 2002). Treating each customer individually costs money, so the customer’s value has to carry the cost.
What privacy rules apply to one-to-one marketing?
The method runs on personal data, so privacy law applies. Under the GDPR, a person can object to direct marketing at any time. That covers related profiling. After an objection, the data can no longer be used for that purpose (Article 21).
Article 22 limits decisions made only by software. People can avoid those that have legal or similarly significant effects (Article 22). Most marketing offers fall short of that bar. Check each use with counsel.
The same care applies to email. The FTC says you must honor an opt-out request within 10 business days (FTC). Build the objection and the opt-out into the data model, so every send checks them first.
How do loyalty programs run one-to-one marketing?
A loyalty program holds the history this approach needs. Starbucks says it can tailor special offers to what a member buys and prefers. To get them by email, a member must opt in to promotional and marketing offers (Starbucks terms). The opt-in is part of the design. Starbucks says special offers can reach members in the app, with push notices. Email and physical mail carry them too. The channel is part of the one-person choice.
McDonald’s lets customers get tailored offers and order in its app (McDonald’s 10-K). It shows the scale. Across 70 loyalty markets, sales to loyalty members rose 20% in 2025, to nearly $37 billion, with nearly 210 million 90-day active users (McDonald’s release).
Ulta says its loyalty programs show what guests prefer. That helps it tailor promotions and recommendations (Ulta 10-K).
For illustration, a program tests a segment offer against an individual offer on matched groups of 20,000 members. The segment offer lifts purchases by 4.0 points. The individual offer lifts them by 4.6 points. The extra 0.6 points is 120 members. If the individual build costs more than those 120 members are worth, the segment offer wins.
How GRAVTY handles one-to-one marketing
GRAVTY holds a patent on AI-Driven Mass Individualization (PR Newswire). The firm says it brings personal member experiences to large scale. SHARE, the Majid Al Futtaim program on GRAVTY, sends 4.5M offers a month, each one tailored to a member (Loyalty Juggernaut clients).
Member 360 pulls behaviors, influences and transactions into one view. That view is the data a one-person offer needs.
On its retail page, lji.io says AgenticTest watches lift in flight. It pulls budget from weak arms automatically.
A full guide covers the method. A glossary entry covers the software.
Frequently asked questions
Who coined one-to-one marketing?
Don Peppers and Martha Rogers launched the idea in their 1993 book The One to One Future (Don Peppers). They set out the case in Harvard Business Review in 1995, with Joseph Pine, as a learning relationship with each customer (HBR, 1995).
What is the difference between one-to-one marketing and personalization?
One-to-one marketing is the strategy. Personalization is a tactic inside it. The strategy treats each customer as a market of one. A tactic tailors a message or offer for one person. Add real-time context and you get hyper-personalization.
Is one-to-one marketing worth the cost?
It depends on the channel and the category. The extra payoff of one-person offers was small in general, especially offline (Zhang and Wedel). Test against a segment offer first.
Does one-to-one marketing break privacy law?
Not by itself, but it must honor privacy rights. The GDPR gives people a right to object to direct marketing, including related profiling (Article 21). After an objection, the data can no longer be used for that purpose.
Can small programs do one-to-one marketing?
Yes, with a narrow scope. Pick one decision, such as the next offer. Test it against a segment offer. Use the purchase history the program already holds.
