This article outlines:
What a McDonald's guest actually found when he requested his own loyalty data
Why the discomfort in that story isn't really about data collection
Why personalization at this scale used to require McDonald's-level resources, and doesn't anymore
This month, WIRED published an article from a reporter who was shocked when he requested a copy of the data that McDonald's stored on him from its loyalty program and received 515 pages back. It detailed every Monopoly prize he'd ever won, a predicted 2.16 visits over the next six weeks, an average order of $13.49, and a churn score of zero. In the end, he asked McDonald's to delete his data and stopped eating there to see if he could prove the algorithm wrong.
Plenty of takes since then have focused on how much McDonald's knew about the reporter. But most brands with a loyalty program are sitting on some version of this data. What’s more interesting is how restaurants can turn those data points into meaningfully personalized experiences that benefit both the guest and the restaurant, and can now accomplish this even if they don’t have the scale of McDonald’s.
Is guest data a trade or a form of surveillance?
Loyalty is a transaction, and data is the currency the guest spends. It's the only currency a brand wants more than money, because it's what makes better business decisions possible.
That's not a controversial take, even among the article's own readers. Several commenters shrugged at the report: tracking and forecasting purchases is a legitimate business case, and companies have been profiling regulars since the department store credit card. (One reader turned the question back on the outlet, asking whether WIRED could say what it stores on its own subscribers.)
The reporter's 515-page file is really a receipt for years of purchases. In exchange for reporting what he ate and when, he got perks and points that added up (and let’s not forget those Monopoly prizes). Both sides paid. Both sides collected.
The reporter only felt uneasy once he requested the file and saw the backend laid bare, with terms like "attrition likelihood" and "RFM-derived top product." He wasn't reacting to the fact that McDonald's has data on him. He's sensationalizing the 515 pages of data when that is in fact a fairly normal amount of information about regular transaction details over several years. Restaurants need that amount of data to market effectively (with personalized, relevant offers to guests at scale), make informed business decisions, and quite frankly survive in today's market.
Creating personalized loyalty
I worked in restaurants for 15 years before building software, and I lived the analog version of loyalty rewards. We’d show regulars to a corner booth, and have their favorite drink waiting for them. In return, guests gave us their feedback and their business. They didn't feel watched. They felt known.
Recreating that kind of personalized experience at scale used to require an entire team of analysts. Most restaurant brands, even successful multi-unit ones, never had the resources to know a single guest this well, let alone act on it.
What's changed is the technology. A modern guest engagement platform like Olo Engage builds one profile per guest across every ordering channel, then segments guests automatically by how recently, how often, and how much they've ordered. That segmentation is what turns "hasn't ordered in 60 days" into an automatic win-back offer, or "always orders the same wrap" into a birthday reward that isn't generic. None of that has to be built anymore. AI-powered predictive attributes and features like dynamic content to deliver more relevant content based on what you know about the guest come with the platform.
A known guest should be getting things a stranger cannot buy:
- Order preferences that carry across channels, so the guest never has to re-enter them
- A pickup order that's labeled with their name, thanking them for being a loyal guest
- First access to a limited item before it hits the full menu
- A comped item that shows up the day after a bad visit, unprompted, because the brand noticed
Every one of these requires knowing the guest. Every one is worth more to that guest than another 15% off, and most cost the brand less to deliver.
How does guest data benefit the brand?
Repeat guests carry the restaurant business, so the data that helps a brand keep them coming back feeds the forecast, the staffing plan, and the marketing budget. Personalization built on that data is what turns "we have 515 pages on this guy" into an offer he'd actually want. A brand that knows a large Diet Coke is someone's top item doesn't need to guess what to put in a win-back offer.
And, it lets a brand prove the personalization is working rather than assume it. Holding out a control group that doesn't get an offer shows whether it actually drove a visit, or whether that guest would have shown up anyway.
Know them, then prove it
If you run a loyalty program, ask one question this week: What does a known guest get from us that a stranger can't get? If the honest answer is a discount and some emails, you don't have a loyalty program. You have a coupon list with a database attached.
Guests hand over an extraordinary amount of information about themselves, every day. Creating a more personal system used to require McDonald's-level resources, but not anymore.
Guests just want to be treated like someone who's been there before. Do that, and no one writes an article about your data. They’ll write one about how you always seem to know what they need.
