The DataDelivers Playbook: Lifting Retention Across a Multi-Brand Portfolio
Kahala Brands runs one of the hardest versions of the guest-data problem in the restaurant industry.
The problem: a franchisor approaching 30 brands, built by acquisition. Every brand has its own POS, its own providers, and a loyalty view that only sees the guests who joined.
The playbook: unify each brand's data under one team, identify the guests loyalty can't see, put offers where they change behavior, and aim at the inactive majority.
The result: working with DataDelivers and the brand's loyalty platform, retention lifted from about 15% to between 20% and 25%, depending on the brand.
About this playbookDataDelivers built this playbook from the published Kahala Brands case study. It covers one lane of that story: lifting retention across a multi-brand portfolio. For the full story, read the Kahala Brands case study.
The company is a franchisor approaching 30 brands, with Cold Stone Creamery, Baja Fresh, Planet Smoothie, and Taco Time among the largest, and it grew by acquisition. Every acquisition arrives with its own technology: its own POS, its own email and SMS providers, sometimes its own loyalty platform. A small loyalty team operates as an in-house agency across the whole portfolio, supporting every brand that runs a program.
So before Kahala could ask any of the interesting marketing questions, it had to answer an unglamorous one first: how do you find partners that can integrate across every POS platform and every communication service, at every brand?
Underneath the systems problem sat a knowledge problem. Brandon Hodgins, Kahala's Director of Loyalty Marketing, puts it plainly: "If you ask a traditional marketer who their customer is, and compare that to what the CDP says, it's normally two different things." Loyalty tells you about the guests who joined, and that is a smaller group than most brands assume. DataDelivers' 2026 Restaurant Guest Engagement Report describes the ceiling this way: participation in loyalty programs tops out at around 15% to 20% of a customer base, leaving the rest unknown and unreachable. At a franchisor's scale, that's most of the guest base at every brand in the portfolio.
That gap is what the playbook below goes after. Not every guest in it can be matched, and the honest version of this work is about the portion that can. Here's how Kahala ran it.
Move 1 — Unify Each Brand's Data, Run It With One Team
DataDelivers is Kahala's customer data platform. POS data, online ordering data, and loyalty data from across each participating brand's stack are pulled into one place and married together into a picture the team could never assemble before. The views stay brand by brand. The unlock is who runs them: one loyalty team, operating every participating brand's program the same way. Everything sitting in one place is what makes a portfolio manageable by a single team, and what lets that team act on the whole picture instead of one brand's slice of it.
Move 2 — Identify the Guests Loyalty Can't See
Start with what credit-card tokenization actually does: it turns each card into a persistent, anonymized token. Visits and spend become trackable, but a token isn't a person. It carries no name and no permissioned channel, and a guest who pays with two cards looks like two strangers. Every card-tracking CDP gets that far.
The step that changes the marketing is matching. The CDP resolves those tokens against DataDelivers' embedded consumer database, and two cards become one known guest the brand is allowed to reach. Not every token resolves. The ones that do come out of the part of the base a marketer otherwise can't see at all. Two things follow.
First, advertising starts working harder. Upload a list of phone numbers to an ad platform and it matches whatever share it matches. Run the same audience through the CDP first and the blanks get filled in. The match rate climbs, the ads reach far more real guests, and return on ad spend goes with it.
Second, the loyalty program gets a way to grow. A guest the brand can reach is a guest the brand can invite, so the program can finally be promoted to the people who aren't in it. That isn't something a brand can do for a guest it can't see.
"DataDelivers allows us to do much better ad targeting, so our return on ad spend skyrockets. It also allows us to do campaign orchestration and build audiences based on past behavior in ways we couldn't before. We can drive specific offers to different segments, getting the right offer at the right time into the hands of the right guests."
Brandon Hodgins, Director of Loyalty Marketing, Kahala BrandsMove 3 — Put Offers Where They Change Behavior
With unified data and real audiences, campaign orchestration becomes the craft: the right offer, at the right time, to the right guest, and, just as important, no offer to the guest who was coming anyway. The already-frequent guest might get a modest nudge toward one extra visit. The guest who hasn't been in for months gets the aggressive incentive.
The same logic runs seasonally at a brand like Cold Stone. In the busy season, discounts get scarcer and more campaigns are built to grow the check. In the slow season, the priority flips to visits, and a discount that buys an incremental trip pays for itself.
Move 4 — Aim at the Inactive Majority
Left alone, first-time guests don't come back. The 2026 Restaurant Guest Engagement Report found that without direct marketing engagement, 88% of new guests never return after their first visit.
Set that against the cost of winning them, which Boston Consulting Group put at about $30 a guest in 2025, roughly ten times what keeping one with weekly email costs, and the economics are unforgiving.
"If you've spent $30 to get me to come in one time, you've lost money on that endeavor. You end up continuously in this loop of losing money. You acquire a customer, they don't return, and you don't make enough off that single visit to pay for what it cost to acquire them."
Pat Riley, VP of Sales, DataDeliversKahala's sharpest use of the platform is the least glamorous audience: guests who stopped coming. It compounds the trap above. Guests who visit once and disappear never enter the returning base at all, so the inactive pool ends up the largest audience any brand has, and re-engaging even a slice of it is revenue that otherwise doesn't happen.
The CDP is what makes those economics legible. Once a brand can see what a first visit costs to buy and how few of those guests return on their own, bringing a known guest back through the door is plainly the cheaper path.
What It Produced
Working with DataDelivers and the brand's loyalty platform, Kahala has lifted retention from about 15% to between 20% and 25%, depending on the brand, by focusing on inactive guests. Campaign orchestration across the unified data has raised loyalty-member lifetime value, and ad campaigns routed through the CDP return more per dollar than campaigns run without it.
"With the CDP, we've been able to focus on inactive guests and bring our retention rates up to 20% or 25%, depending on the brand. If you're getting one or two extra visits out of an individual person in a year, that lifetime value really skyrockets."
Brandon Hodgins, Director of Loyalty Marketing, Kahala BrandsThe Portfolio Lesson
A loyalty program's points and rewards are not what produce the return. They're the price of admission. The guest gets recognition and the occasional reward; the brand earns the data. The return comes from what you do with that data: the targeting, the segmentation, the orchestration.
Across almost every brand in Kahala's portfolio, loyalty members spend five to six times what non-members do. That gap isn't something a program conjures out of nothing. The guests who sign up tend to be the ones who were already coming most often. What a program does is make them legible, and what a CDP does is extend that legibility to everyone else it can match.
That's the portfolio lesson. At two brands, the guests loyalty can't see are a blind spot. At thirty, acquired one at a time with a different stack behind each one, they're most of the guest base, and no amount of program design reaches them. What does reach them is the unglamorous work at the front of this playbook: each participating brand's data in one place, the matchable guests given names, and one team running it with the whole picture in front of them.
Read the Full Case Study
The complete Kahala Brands story: the portfolio, the program, and the numbers, in their words.
Read the Kahala Brands Case StudyAnd if your guest data lives in as many places as your brands do, the first question is the same one Kahala started with: what would one unified view show you that your loyalty dashboard can't? That's a conversation we have with restaurant brands every week.