The challenge
Two parallel problems, equally weighted:
- 01Lifecycle Framework — talabat pro (tPro). Increase adoption and usage intensity to drive LTV and ecosystem retention.
- 02Personalization Framework — Churn Prevention. Decide who to target, what to say, and whether to incentivize. Margin protection is the implicit constraint.
My read
I read “increase adoption” as “fix the funnel economics.” I read “personalization” as “margin protection.”
- —The real risk in subscription is the first 30 days. Habit loop, not signup.
- —Most CRM teams burn margin spamming discounts. The strategic question isn’t how to incentivize. It’s whether to.
- —tPro is a defensive moat, not a growth product. Different goal, different lifecycle.
If this is your problem space
Subscription economics, churn budgets, incrementality. 30 minutes, video, no agenda template.
Key frameworks
Framework 01
The Habit Loop, not a funnel.
The Habit Loop is a four-stage subscription adoption framework: Awareness & Rationalization, Onboarding & Activation, Value Realization, and Churn Prevention. Unlike a funnel, it treats the first 30 days after signup — the Golden Period — as the decisive stage: a subscriber who crosses breakeven on the fee in month one forms the habit; one who doesn’t cancels.
A subscription lifecycle has four stages — and the second one is the only one that matters for retention.
Stage 1
Awareness & Rationalization
Trigger Loss Aversion. Show what they’re losing by not subscribing.
Stage 2 · “Golden Period”
Onboarding & Activation
Drive 3+ orders in 30 days. Cross breakeven. If they don’t save more than the fee, they cancel.
Stage 3
Value Realization
Vertical cross‑pollination. Free delivery on Food drives tMart adoption. Share‑of‑wallet up.
Stage 4
Churn Prevention
Restate value with personalized savings data. "You saved AED 240 last quarter."
Framework 02
Behavioral propensity, not demographics.
Behavioral propensity segmentation divides a subscription audience into three groups by ordering behavior, not demographics: Whales, who already order enough and need a savings reframe; Calculators, who sit on the fence and need a trial; and Churn-risks, who subscribed but aren’t using the benefit. Each segment gets a different message and a different economic logic.
Three propensity segments. Each gets a different message — and a different economic logic.
Group A
Whales
Already order enough. Pitch isn’t "order more" — it’s "stop wasting money on fees."
Group B
Calculators
On the fence. Need a nudge — usually a free trial to prove the value.
Group C
Churn risks
Subscribed but not using it. The danger zone. 3 orders in month 1 or they cancel.
Framework 03
The 2×2: Risk × Value matrix.
The Risk × Value matrix is a 2×2 decision grid that allocates churn-intervention budget along two axes: user value and churn risk. High-value, high-risk users get aggressive, human-led saves; low-value, high-risk users get cheap automated nudges or are let go. It exists to prevent the default CRM failure mode — burning margin on blanket discounts.
Allocates intervention budget by user value and churn risk. Saves Whales aggressively. Doesn’t overinvest in low‑value users. Doesn’t insult anyone with an AED 5 coupon when they wanted service.
Cross‑sell verticals. Loyalty perks. Quiet — they’re happy.
Save aggressively
WhatsApp + senior CSM apology + service assurance. Never lead with a coupon.
Automated nudges. Don’t spend channel cost on these.
Email + push. Targeted discount only on a cart‑abandon signal. Otherwise — let them go.
Framework 04
Universal holdout. Incrementality, not opens.
A universal holdout is a randomized group of users permanently excluded from all CRM and lifecycle interventions. Comparing their revenue against the contacted population is the only way to measure true incremental GMV — what the program actually caused — rather than vanity engagement metrics like opens and clicks that would have happened anyway.
Open rates and CTRs tell you nothing about whether the user would have renewed anyway. A universal holdout group — randomized, persistent, excluded from all CRM intervention — is the only way to measure true incremental GMV. The methodology I pushed for at Uber and would carry into any subscription engine.
Full artifacts
12‑slide presentation + working doc showing the live thinking process. Available on request — typically shared during the fit call if relevant.