3. Feature-by-Feature Scrutiny (Your Stated Features)
Categories:
- Table stakes (TS): customers expect it. It earns nothing, but its absence costs you.
- Differentiator (D): a real reason to choose you over the alternatives.
- Distraction (X): built, but unlikely to move revenue or retention.
- Liability (L): creates support, compliance, legal or trust risk.
Some features sit in two buckets. The primary bucket comes first. Each is judged twice: for the idea as described (multi-platform) and for the Razorpay wedge.
3.1 Classification table
| # | Your feature | As described | Razorpay wedge | Reasoning |
|---|---|---|---|---|
| 1 | Smart retries | TS | TS / L | Stripe (Smart Retries), Paddle (7 retries over 30 days, algorithmic with Retain), Lemon Squeezy (4 over 2 weeks) and Razorpay ("Smart Payment Retries") all retry natively. Beating Stripe's model with no network data is unlikely. On Razorpay, the 24-hour pre-debit notice limits how many retries fit in a cycle, and whether a third party can trigger a retry is unverified. Promise it before you've tested it and it becomes a liability. |
| 2 | Card-update emails | TS | TS | Every platform sends them, with a hosted update page. You need them, and they win nothing. |
| 3 | SMS | D (weak) | TS → D only as WhatsApp | Outside India, SMS is a modest edge. Stripe doesn't do it natively, but Paddle Retain, Churn Buster and Stunning do. In India, SMS needs TRAI DLT template registration, and customers answer WhatsApp. WhatsApp with a UPI pay link is the differentiator. |
| 4 | Pre-dunning for expiring cards | TS | D (reframed) | Stripe sends expiring-card emails, and Paddle Retain and Churn Buster pre-dun. In India, the real "pre-failure" signals are different: mandate end dates, mandates cancelled in the UPI app, charges above the ₹15,000 extra-authentication limit, and tokenised cards being re-issued. Monitoring mandate health is new. |
| 5 | Cancel-flow saves | D, crowded | X in v1 | Churnkey, Churn Buster and Paddle Retain do this well. It needs a script in the merchant's app, which slows onboarding and adds support. It's voluntary churn, which is a different problem. Add it after month 12, if merchants ask. |
| 6 | Monthly "we recovered $X for you" report | D / L | D (with holdout) / L (without) | This is the retention engine and the sales pitch. But the platform retries anyway, so "recovered" must mean incremental over what would have happened. A 10% holdout group on default handling makes the number defensible. Without one, it's a liability: the first merchant who checks will cancel. |
| 7 | Multi-platform wedge (Paddle, Lemon Squeezy, Razorpay, Stripe without Chargebee, Shopify apps) | L | — | Five integrations for one small team, and four of the five platforms have a free native answer or an established specialist. This is positioning, not a feature, and it's the root problem with the idea as described. |
3.2 Scoreboard (as described)
| Category | Count | Features |
|---|---|---|
| Table stakes | 3 | Smart retries, card-update emails, pre-dunning for expiring cards |
| Differentiator | 3 | SMS (weak), cancel flows (crowded), recovered-$ report |
| Distraction | 0 | — |
| Liability | 2 | Multi-platform positioning, the recovered-$ report without a holdout |
About a third of what you described is a reason to buy, and each of those reasons is already offered by someone. In the Razorpay wedge the picture changes: WhatsApp recovery, mandate health and the holdout-proven report are all differentiators, because no competitor is there.
3.3 "Building to avoid selling": explicit call-outs
- Five platform integrations. Each one means webhooks, auth, edge cases, test accounts and permanent maintenance. Building them feels like progress and teaches you nothing about whether anyone pays.
- A retry-timing ML model. Stripe has network data from millions of merchants. You don't. Build rules first, and learn from your own holdout data later.
- Cancel flows. They're a second product with a second buyer conversation.
- A polished dashboard before the first report. The merchant wants one number a month. A PDF or email with the rupees and the proof is enough for the first 20 customers.
3.4 What a disciplined v1 looks like
Keep (Razorpay wedge):
- webhook ingest of failed subscription charges and mandate events;
- WhatsApp → SMS → email sequence with a UPI pay-now link and a "set up UPI Autopay again" link;
- mandate-health alerts before the renewal date;
- a 10% holdout group;
- a monthly ₹-recovered report.
Defer: retry orchestration (until API access is confirmed), multilingual templates beyond Hindi and English, a self-serve dashboard, analytics.
Cut: Paddle, Lemon Squeezy, Shopify and general Stripe integrations; cancel flows in v1; ML retry timing; human calling.