Analytics

2 Million Users. Same Six People.

April 18, 2026 5 min read

You have a WhatsApp group. Everyone in India has at least four.

And in every single one of them, the same six people exist. The uncle who sends good morning images at 7am with the reliability of a Swiss watch. The cousin settled abroad who disappears for six months and resurfaces only when they need a gift recommendation. The person who only reacts with thumbs up and has never once typed a full sentence. The self-appointed admin. The one who last seen-ed in 2022 and is somehow still in the group. The one who replies to a message from three days ago as if it just came in.

Here is what I want to tell you: your user base is this group. Every single product, every single app, every single platform. Same cast, different names.

THE WHATSAPP GROUP · EVERY USER BASE EVER CHARACTER WHAT THEY DO USER EQUIVALENT Forwarding Uncle RFM: High F, Low M Good morning images, fake news, motivational quotes. 7am without fail. Opens everything. Clicks everything. Buys nothing. Great DAU, terrible revenue. The NRI RFM: Low F, High M Silent for months. Appears at Diwali: "what's a good gift under 5k?" Seasonal, high value when active. Your retention team cannot explain them. The Reactor RFM: High F, Zero M Only sends thumbs up and laugh emojis. Never types. Always watching. High session frequency, zero conversion. Loves your app. Has never paid for it. The Group Admin RFM: High F, High M Posts rules. Pins messages. Adds people without asking. Thinks they run things. Power user. High LTV. Refers others. Treat them well. They notice everything. The Ghost RFM: Zero everything Last seen March 2022. Still in the group. Nobody removes them. Hope dies last. Dormant user. Probably churned. Still inflating your install numbers. The Late Replier RFM: Unpredictable Responds to a 3-day-old message urgently. Irregular cadence. Breaks your cohort logic. Real, just noisy.

The mistake most teams make is sending the same message to all six of them. The same push notification. The same discount. The same reactivation email. It goes out to 2 million users and lands differently for each of them, which is to say it mostly does not land at all.

This is what segmentation is actually about. Not fancy ML models or data warehouses. Just the basic acknowledgment that your users are not one person.

RFM: the framework your WhatsApp group already taught you

Recency, Frequency, Monetary. Three questions about every user.

When did they last show up? How often do they come? And when they do come, do they spend?

That is it. Everything else is refinement.

Score each user on these three dimensions, usually on a scale of 1 to 5, and suddenly your 2 million users become five or six meaningfully different groups of people. Each group needs a different conversation.

RFM SCORE MAPPING · WHERE DO YOUR USERS FALL? SEGMENT RECENCY FREQUENCY MONETARY WHAT TO DO Champions Recent Often High spend Reward. Don't over-contact. Loyal Recent Often Medium spend Upsell. Build habit. At Risk Not recent Was frequent Was high Win-back now. Clock ticking. One-Order Wonder Recent-ish Once Low Second order is everything. Dormant Long gone Rarely Minimal Low ROI. Accept the loss.

The Champions are your Group Admin. High on all three. They know the product better than most of your team does. The instinct is to flood them with communication because they engage. Resist this. They are already yours. Over-contacting them is how you lose them.

The Loyal users are solid. They show up, they spend, just not at the Champion level. The job here is habit-building. Make the next visit slightly easier than the last.

The At Risk segment is the one that keeps retention teams up at night. These are people who were once frequent, once valuable, and have quietly stopped coming. The window to win them back is short and closing. Every week you wait, the probability of return drops.

And then there is the one I want to talk about properly.

The one-order wonder

Every app has them. Nobody talks about them enough.

The one-order wonder installs your app, usually right after a campaign, often because there was a discount involved. They place one order. They have a fine experience, nothing exceptional. And then they never come back.

They look great in your acquisition dashboard. New user, first order, ticks every box. Then they vanish and spend the rest of their existence silently inflating your install numbers while your retention cohorts look increasingly grim.

The reason nobody talks about this enough is that it is uncomfortable. It implies the acquisition campaign that everyone celebrated brought in a lot of people who were never really your users. They came for the discount, not for you. And there is a difference.

This is not a reason to stop running campaigns. It is a reason to track what happens after. The second order is the real signal. A user who orders twice is meaningfully more likely to order a third time than a user who has ordered once. That gap, between first and second order, is where the actual retention problem usually lives.

If you are looking for somewhere to focus, look there.

The philosophical bit

Here is what changes when you think in segments rather than in aggregate numbers.

Your definition of growth shifts.

“We grew 20% this month” means something very different if that growth came from Champions ordering more often versus a flood of one-order wonders who will be gone by next month. The number looks the same on a slide. The business reality is completely different.

Segmentation does not give you a better dashboard. It gives you better questions. Who actually came back? Who did we acquire versus who did we retain? Are we growing the right users or just growing the count?

The WhatsApp group does not get smaller when the Ghost stops showing up. The member count stays the same. But anyone who is honest knows the group already lost that person a long time ago.

Your install numbers work the same way.

The hard question is not how many users you have. It is how many of them are actually in the conversation.