Expansion program
Some customers get more value from your product than they're paying for. That's a problem worth solving, gently.
Expansion is the program that names what they're already doing and asks if they want to do more of it. Plan upgrades, seat adds, usage-cap increases. Not a sales pitch. A nudge that's accurate enough to feel like good service.
When it fires
When usage crosses a threshold you set. The Cartographer watches the event stream and flags contacts whose pattern has outgrown their plan: hitting the seat cap, brushing against rate limits, using a feature their plan only partially includes.
Tenured customers also trigger expansion on a slow cadence. If a customer's been on the same plan for six months and their usage has roughly doubled, Mara has something to write about.
What's in it
One or two emails per trigger. Never more. Expansion is the easiest program to overdo and the hardest to recover from when you do.
The first send names the specific usage pattern. "You've added five teammates in the last month. The team plan is built for this." Concrete. Not "you might want to consider upgrading."
The second send (only if there's room) offers a check-in. The Reply Analyst routes questions to your inbox.
How she writes it
The Brand Analyst supplies the voice. The Cartographer supplies the usage signal. The Copywriter writes the connection.
What Mara avoids is the upgrade-pitch tone. No "unlock the full power of," no "take your team to the next level." The customer's already at the next level. The email just names the plan that matches where they are.
If you've connected your billing provider, Mara reads the pricing page directly. The numbers in the email are the real numbers from your plan tiers, not approximations.
What gets tested
Framing. The bandit picks between achievement framing ("you've outgrown the X plan"), limit framing ("you're at 95% of your seat cap"), and offer framing ("here's what the Y plan adds"). Each one works better for some segments than others.
CTA copy gets less variation than welcome or activation. "Upgrade" is hard to beat; the bandit mostly tests the modifier ("Upgrade now" versus "See the team plan" versus "Compare plans").
What you control
Trigger thresholds are yours. The Cartographer's defaults are reasonable, but you know your customers better. If a customer hitting 80% of their seat cap shouldn't be nudged toward an upgrade, set the threshold higher.
You can also exclude segments. Some customers are on annual contracts that lock their plan; Mara should not write expansion emails to them mid-term.
Related programs
Expansion overlaps with feature-adoption when the feature you shipped is plan-gated. Mara's good at separating the two: the feature-adoption send tells active users the feature exists; the expansion send tells the ones who'd benefit that their current plan doesn't include it.
If a customer hits a usage cap hard and bounces, that's different. Churn-save takes over if they cancel as a result.