Retention audit and CRM audit for D2C brands and apps in India

Every tool is on. None of them is listening.

A Compound Lab retention audit reads your CRM, store, checkout, WhatsApp and app the way a customer meets them. We shop your brand, follow every event and every message, and show you what is in use, what is switched off and what is broken. Then we give you the plan, one problem at a time.

Run by Raviraj Ramavarapu, who reads every account himself.

Five tools, one customerA store, a checkout, a CRM, a chat tool and an app are all switched on and each one is linked to the same customer. The links between the tools exist but are switched off.OFFOFFOFFOFFOFFSTOREONCHECKOUTONCRMONCHAT TOOLONAPPONONECUSTOMER
Illustrative. Each tool sees the customer. None of them sees the others.

What is a retention audit?

A retention audit is a read-only review of what happens to a customer between their first visit and their second order: which actions your tools record, which messages they try to send, what actually arrives, and who speaks after the first order. It is measured against your own targets for repeat rate, channel revenue and discounts, and it ends with a numbered list of problems and the plan to fix each one.

Where you stand

The waiting room.

  1. 01

    Orders, buyers and the repeat curve, cohort by cohort.

  2. 02

    Everyone who came back, and how long they took.

  3. 03

    Everyone who bought once, more than 30 days ago.

  4. 04

    The targets you set, and the gap to each one.

CAME BACKBOUGHT ONCE, OVER 30 DAYS AGONEWER BUYERS

Illustrative. A hundred buyers, sorted the way the audit sorts them.

They trusted you once. Nobody asked them twice.

Act one: the shop

We shop your brand. Then we walk away.

  1. 01

    A real cart on our own phone, timed to the second.

  2. 02

    Every event your tools record about us, in the order they fire.

  3. 03

    Every message they try to send, and what actually reaches the phone.

  4. 04

    What happens the next day, on the app, and in the week after.

THE VISIT0:00VIEWED0:04CART0:12CHECKOUT0:37PHONE GIVEN0:41LEFTTHEN, THE MESSAGES+21 MINCART REMINDER, WITH AN OFFERNEVER ARRIVED+31 MINA CALL PROMISING A CODEANSWERED+32 MINTHE CODE, ON WHATSAPPNEVER ARRIVED+40 MINA GENERIC WHATSAPPREFUSED+2 HRSAN SMS TO THE HOMEPAGEDELIVERED

Illustrative. Built the way every audit starts.

Every tool said sent. The phone said otherwise.

Act two: why it happened

One tap. Four names.

  1. 01

    How your tools pass data to each other, and the links that exist but are switched off.

  2. 02

    One customer action arriving under several event names, and the one each flow listens to.

  3. 03

    Flows racing for the same customer, and the generic message that wins.

  4. 04

    Who can be reached at all, and who is messaged anyway.

  5. 05

    Two tools claiming the same order, with no control group to settle it.

ONE TAPADD TO CARTStore - Add To CartFROM THE STOREadd_to_cartFROM THE APPvisitor_addtocartFROM A VISITOR TOOLAdd To CartFROM THE SITE TAGTHE ONLY NAME THE CART FLOW LISTENS FOR

Illustrative. Bars drawn to scale, as they appear in a real event list.

Your customer did everything right. Your tools just didn’t hear it.

Act three: after the order

Some moments have three owners. Others have none.

  1. 01

    Every message from the order to day 90, from every tool that sends.

  2. 02

    Who owns each moment, and the moments nobody owns.

  3. 03

    What support is answering that a message could have.

  4. 04

    Features already paid for, and still switched off.

CRMCHAT TOOLSTORE EMAILCART LEFT3ORDER2SHIPPED1DELIVEREDNOBODYDAY 201DAY 60NOBODYOWNERS

Illustrative. Filled squares are tools that send something at that moment.

The second order is nobody’s job.

Retention strategies, checked

Every strategy. Scored against your stack.

  1. 01

    Is it built, or does it live on a slide?

  2. 02

    Does it start on the right event, for the right people?

  3. 03

    Does the message reach the phone or the inbox?

  4. 04

    Does it earn, measured against a control group?

WELCOMEIN USECART AND CHECKOUTBROKENPOST-PURCHASEUNUSEDREPLENISHMENTUNUSEDCROSS-SELLBROKENWIN-BACKUNUSEDLOYALTY AND REFERRALUNUSED

Illustrative. For apps the list is onboarding, activation, trial to paid, renewals and win-back.

A strategy on a slide isn’t a strategy.

Every tool, sorted

In use. Unused. Broken.

Each tool in your stack is read against your targets and every setting, flow and feature lands in one of three states.

01

In use

Set up and doing real work.

  • The checkout recovery that earns
  • Broadcasts that deliver
  • Data that already flows into the CRM
02

Unused

In the account, switched off.

  • A link between two tools, never turned on
  • Web push, in-app and send-time AI, left idle
  • Events that arrive and start nothing
03

Broken

Running, but failing.

  • A flow waiting on an event that rarely fires
  • Messages logged as sent that never arrive
  • Order updates sent as marketing

A platform limit, like Meta’s cap on marketing messages per person, is shown as a limit. It is never counted as a problem.

How we audit

Five rules. Every audit.

  1. Read-only. Nothing is changed.

    We read your tools with view access or exports. Nothing is edited, paused or switched on while we audit.

  2. Your targets, not a checklist.

    Your repeat rate, your channel goals and your rules on discounts. Every finding is measured against what you said you want.

  3. A problem is something you can already fix.

    If your tools can do it and are not set up to, it is a problem. If a platform caps it, it is a limit, and we say so.

  4. Every number has a source.

    Each figure names the tool, the screen and the day it was read. What we could not confirm is listed as an open question, never guessed.

  5. Fix what you already pay for.

    The plan starts with switched-off features and broken settings, before it suggests any new software.

How the audit runs

Seven steps. From your brief to your plan.

The same order for every brand and app. How long it takes depends on your stack, and we tell you on the first call.

  1. 01

    Brief

    Your targets in your words: repeat rate, revenue from each channel, and what you will not discount.

  2. 02

    Access

    View logins or exports for your CRM, store, checkout, messaging tools and app. Either works.

  3. 03

    Shop

    We visit, add to cart and leave like a customer, and log everything that follows.

  4. 04

    Read

    Every tool, flow, event and setting, read on one day so the numbers line up.

  5. 05

    Register

    Problems numbered and tied to evidence, with limits listed apart.

  6. 06

    Plan

    Fixes in order, one problem at a time, with what each one should move.

  7. 07

    Walkthrough

    Both decks on a call with your team, and the documents to keep.

Proof

Results that compounded.

What happened after the audit: the missing checkout event at PNG Jewellers, and the retention work at Healthify and Mygate.

All case studies
D2C / 01
Funnel automation

Six journeys, built from scratch. 77% of CRM revenue came from them.

Read the case
77%CRM revenue from automations
CONSUMER APP / 04
In-app funnels

The app already had the traffic. Every screen became a way in.

Read the case
11Surfaces. One connected funnel.
D2C / 03
Repeat-led category growth

Supplements grew from 37% to 45% of the website's revenue. That's the category customers reorder.

Read the case
39%Store revenue driven by CRM
CONSUMER APP / 06
Churn reduction

Residents took a cab to the airport. More of them booked the ride home.

Read the case
+43%Airport return pickups
Raviraj Ramavarapu, founder of Compound Lab
Raviraj RamavarapuFounder. Ex Head of Ecommerce and CRM, Healthify. Ex Senior Manager, Retention, Cumin Co.

Why Compound Lab

A retention agency that reads it like an operator.

01

We run what we audit.

Raviraj has run CleverTap, MoEngage, WebEngage and Kwik Engage with revenue on the line, so settings are read the way an operator reads them.

02

We shop it ourselves.

The audit starts with a real visit on a real phone, not a dashboard. You see what your customer saw.

03

The audit ends in a plan.

Every problem comes with its fix, in the order it should happen, inside the tools you already pay for.

Questions

Before you book.

What does a retention audit include?

Where you stand against your targets, a real visit followed event by event and message by message, how your tools are wired, what happens after the first order, delivery by channel, how revenue is counted, a scorecard for every tool, a numbered problem register and the plan to fix it.

What is the difference between a retention audit and a CRM audit?

A CRM audit checks one tool: its events, flows, channels and settings. A retention audit follows the customer across every tool they meet, from the store and checkout to WhatsApp, the app and support, and reads the repeat curve. Ours includes a full CRM audit as one part of it.

How is customer retention analysis different from an audit?

Customer retention analysis gives you the numbers: cohorts, repeat rate and the gap between orders. The audit adds the why behind them, by reading the tools, events and messages that produce those numbers. See cohort analysis for the numbers side.

Do you need logins to our tools?

View-only logins work best, and exports and screenshots work too. Either way the audit is read-only. Nothing in your accounts is edited, paused or switched on.

Which tools can you audit?

Customer engagement platforms such as MoEngage, CleverTap, WebEngage and Kwik Engage, your Shopify store or app, your checkout, your WhatsApp provider, chat and helpdesk tools, and visitor identification tools. If your stack is different, we start by reading what it can do. A martech stack audit is the same method applied to every tool you pay for.

Which retention strategies does the audit check?

Every customer retention strategy a brand should run: welcome, cart and checkout recovery, post-purchase, replenishment, cross-sell, win-back, and loyalty and referral. For apps: onboarding, activation, trial to paid, renewals and win-back. Each one is checked for whether it is built, starts on the right event, reaches people and earns against a control group.

How long does a retention audit take?

It depends on how many tools you run and how quickly access comes through. We share the timeline on the first call, once we know your stack.

Does the audit include the fix plan?

Yes. Every numbered problem gets its fix, in order, starting with what your tools can already do. You can run the plan with your own team or with us.

What do we receive at the end?

Two decks, one showing where you stand today and one going tool by tool, plus the event list and the source notes behind every number. We walk your team through both on a call.

Is a retention audit useful for consumer apps?

Yes. For apps the shop becomes an install, an onboarding and a trial, and the audit reads app events, push opt-in, paywalls and renewals the same way.

How is a retention audit priced?

After the first call, scoped to the tools you run and the questions you want answered. There is no fixed price list.

Do you work with brands outside Bengaluru?

Yes. Compound Lab is based in Bengaluru and audits brands and apps across India. The work happens on calls and inside your tools.

Book a call

Your tools are already on. Let’s see if they’re listening.

Thirty minutes with Raviraj, not a sales team. On the call we look at three things:

  1. Your targets for repeat and channel revenue
  2. The tools you run and how they connect
  3. What the audit would read first
Book an audit call