The RingBack numbers, honestly: our first 14 days of real data

Every startup pitches its wins. This post does something less common: it publishes all of RingBack's early numbers — including the zeros, the gaps, and the honest smallness — from our own platform's ledgers, with the methodology and the raw data linked. We are a two-week-old missed-call recovery service for Indian salons, gyms, clinics, detailers and barbershops. Here is exactly what our funnel looks like on day 14, so you can judge us on data instead of adjectives.

Numbers below were extracted on 1 October 2026 from the same platform ledgers our customers' dashboards read. Next update: first week of November 2026.

What we can already say, in one paragraph

Two weeks in, our own phone line has captured only a handful of missed calls — and we will not pretend otherwise. What the data does show already: a qualified-lead pipeline of 171 businesses across 6 cities, scored 0–100, with every phone number present, built entirely from public listings by our own lead agent; a pre-registered 8-message WhatsApp outreach experiment with 0 replies that we are publishing anyway; and a product whose entire design exists because the industry-standard benchmark — roughly 1 in 3 calls to small businesses goes unanswered — matches what every salon and gym owner tells us anecdotally. The honest story of a two-week-old funnel is a story of supply and patience. We would rather earn your trust with small true numbers than with big unverifiable ones.

The qualified-lead pipeline: 161 → 171 in 14 days

Our lead agent builds its pipeline from public business listings only — OpenStreetMap-derived, name, city and phone, no scraped private data, no purchased lists. Every lead carries a 0–100 score of listing completeness. The agent was deliberately throttled to ~10–15 new leads per day (quality caps, city rotation across 6 cities) to keep the pipeline honest rather than inflated.

Two honest footnotes. On Sep 23, 25, 26, 29 and 30 the pipeline recorded 0 — not because leads vanished, but because our own monitoring probes could not reach the ledger on those days (Cloudflare free-tier incidents and maintenance windows). We show the zero days rather than smoothing them out. And the count decreased from 172 to 171 because deduplication removed one lead — the pipeline self-cleans, and we show that too.

The WhatsApp outreach experiment: 8 sent, 0 replies, fully pre-registered

On Sep 18 we sent our first 8 cold WhatsApp pitches — and registered the hypothesis, the cohort balance and the decision rules before the outcome window opened:

• Design: a balanced 2×2×2 — 4 dental clinics vs 4 gyms (niche), SHORT (~46–50 words) vs LONG (~102–105 words) pitch, score-matched pairs 85–100, all sent manually by the founder from the business number.
• Result at 14 days: 0 replies from 8 sent. Not one. The trajectory was recorded every 30 minutes by an automated watcher, and it never moved.
• What we conclude: almost nothing, statistically — and that is the point of pre-registration. 8 pitches cannot distinguish "cold WhatsApp does not work" from "this batch was unlucky"; the registered decision rule says silence at n=8 routes to a 25-per-cohort batch, not to a conclusion. What we CAN say: founder-sent cold WhatsApp at n=8 produced no replies in 14 days, with delivery unverified to the recipient side.

We are publishing the zero because "we sent 8 messages and got 8 customers" posts are the reason nobody trusts startup numbers. The 30-minute trajectory, the exact texts, and the pre-registered decision rules all exist in our research repo, and the monthly updates will keep publishing whatever comes next — including more zeros if that is what the data says.

Methodology (so you can check us)

• Where the numbers come from: the same platform ledgers that render customer dashboards — the missed-call capture log, the qualified-lead store, and the outreach ledger with its 30-minute trajectory watcher. Every number in this post was machine-extracted from those stores, not hand-typed.
• What a "lead" is: a business with a public listing, a phone number, and a 0–100 completeness score — a sales prospect for us, not a customer of RingBack. We label it that way on purpose; "leads" is the most-abused word in startup posts.
• What we do NOT yet claim: any recovery rate, any revenue figure, any "calls answered" statistic. Our own missed-call sample is single-digit small; the honest place for a recovery-rate claim is a future update with enough sample to matter. The industry benchmark we cite (about one in three calls missed) is from published small-business phone studies, not our data.
• Update cadence: monthly, first week of the month, appended below with the date. If a month is flat, the flat month gets published.

Why publish this at all?

Because the entire product thesis is that unanswered moments are where business is lost — and a startup that goes silent about its own funnel would be a strange ambassador for that idea. These numbers will move. When the recovery rate becomes measurable, this page will carry it with its sample size. When something breaks (you read above that it already did), the break gets a date and an explanation. Press, customers and the curiously-minded: everything above is re-derivable from the raw data, and we answer questions about methodology at the contact link below.