All articles
Foundational Fixes Drive Retention Growth Faster Than New Tools
Brian Campbell, Founder and CEO of CrucialPoint, makes the case that most retention gains come from repairing basic lifecycle infrastructure before any new technology enters the stack.

Make The CMO Wire one of your go-to sources on Google
People are going after these cool ideas, infusing AI into the system or integrating a new piece of tech that's supposed to drive growth, when the real problem is they have a broken segment.
Retention has moved up the priority list for consumer brands facing rising acquisition costs and a crowded competitive field. The spending that follows tends to go toward new capability, whether that means an AI layer on top of the existing stack or another piece of software promising incremental revenue. Audits of established email and CRM programs often surface a different picture, where the constraint is a misconfigured segment or a trigger that stopped firing months earlier. Repairs like these stay off most roadmaps, and they frequently return more revenue than the new tool that took priority over them.
Brian Campbell is Founder and CEO of CrucialPoint, a lifecycle engineering firm that builds and operates email, CRM, and loyalty infrastructure for brands across ecommerce, quick-service restaurants, hospitality, and retail. He spent seven years at Overstock.com across email and push marketing, marketing technology, and personalization, managing product and development teams that built real-time personalization systems. Today, Campbell offers free audits to prospective clients, which has given him a repeated view of the same failures across dozens of companies.
"People are going after these cool ideas, infusing AI into the system or integrating a new piece of tech that's supposed to drive growth, when the real problem is they have a broken segment," says Campbell. The failures repeat across accounts and categories. Subscribers fail to enter flows, sign-up forms stop passing data, and abandonment sequences serve content that broke during a template update. Each one's small enough to escape a quarterly review but large enough to move revenue.
Broken segments and missed flows
Campbell uncovers failures by clearing away everything else competing for his attention. He turns off notifications and examines one area of an account at a time, starting with flows and reaching segments only afterward. The limiting factor, in his description, is attention, and a narrow scope surfaces problems that a broader review passes over. "Set a timer on your phone. Just spend 30 minutes and try to find as much as you can," says Campbell.
The same restraint shapes what a team buys once the basics are repaired. Campbell frames any remaining gap as a "build-or-buy" question, and he lands on "buy" in most cases. The appeal of building has grown as code generation tools lowered the barrier to a working first version. "If you've never built a SaaS product, it's not as easy as you think," he explains. "Even though you can vibe code, it doesn't work the way you think, and upkeeping that code is a nightmare if you don't know what you're doing."
Offers past the welcome series
Email, SMS, CRM, and loyalty programs get filed under retention on most org charts, and Campbell argues that filing is expensive. A subscriber who joins a list and leaves without buying already cost the brand money to acquire, putting the economics of loyalty in play before a first purchase. Treating the first seven days as the only conversion window writes off that spend. "These are always seen as retention channels, but they can also be used as acquisition channels," says Campbell.
His correction is a continuous offer strategy across the lifecycle. A brand willing to spend ten extra dollars in coupon value to convert someone in week one is generally willing to spend it in month three. Campbell recommends carrying that offer into broadcast sends and abandonment sequences, with a split between buyers and non-buyers. The offer can be a free add-on or time with a sales representative in place of a discount. He also points to programs that lose track of where a customer stands, including clothing brands still serving toddler product to a buyer whose child reached middle school. "It's easier to get an email out with an offer than to think it through and map out the whole lifecycle," he notes.
Frequency caps by cohort
Campbell spent years in a high-volume environment, where Overstock sent multiple emails a day and the strategy rewarded more of them. Work with quick-service restaurant clients has revised that position over the past three years. Frequency tolerance varies by industry and by brand, and the point where extra sending turns counterproductive comes earlier than most teams assume. "We're finding there's a detrimental layer when you hit four emails a week," says Campbell. "You start causing reverse loyalty metrics. Users start dropping out of the funnel, they start deleting the app."
Daily sending remains viable in his view, provided the limit is applied to each subscriber. A program can send every day while capping any one person at three messages a week, which requires frequency testing broken out by cohort and segment. Segment-level reporting is what makes the damage visible, and recent industry research points to relevance as the variable that determines tolerance. "If you look at the wrong metrics, it will show you that what you're doing is correct, when in fact it may not be," he adds.
Thirty days and outside review
Fixing a retention program used to be scoped as a multi-month project. Campbell now structures engagements as 30-day turnarounds. He opens with a technical audit, resolves a list of quick wins in the first two weeks, then works through segmentation and deliverability issues before handing the program back to the internal team. The same sequence is available to any team willing to run it in-house. "Thirty days is the new 90 days," says Campbell. "Timelines are just shrinking."
Turnarounds that fast raise an obvious question about automation. Campbell has tested whether general-purpose AI tools can run the diagnostic and found they surface little, because the data they read is the same broken data that created the problem. His team uses those tools for mapping and heavier repetitive tasks, while the diagnosis stays with a human. He describes accounts where repairing one broken segment lifted revenue by half in the first month, which he counts as recovered budget.
His last point is aimed at the people who sign off on the program. Delegation is where he sees the gap, since a retention program can look healthy in a status report while the account underneath it degrades. "You want to trust your team and you want to trust your agency, but you just need to spend time in that account and look for those issues," Campbell concludes. "If you don't know what you're looking for, have somebody else do it."




