Fix your marketing data.
Grow on the right numbers.
No fancy dashboards. I find what’s breaking in your tracking across Adjust, AppsFlyer, GA4 and your CRM, fix it, and show you which campaigns really bring customers who stay.
Your dashboard looks fine.
Your customers are still leaving.
Most brands don’t have a data problem. They have a tracking problem, and every decision built on top of it is a guess.
Every platform claims the same sale
Meta, Google and your agencies together report far more orders than your backend actually received.
Offline & influencers show as “organic”
QR codes, hoardings and creators send users with no source, so the campaigns that worked get cut.
Fake installs look like growth
Bots and click spam make cheap networks look cheaper, and teach your ad algorithm to find more fakes.
Optimising for installs, not buyers
The lowest cost per install often brings the most expensive customers.
No view of retention
You know this month’s installs, but not who is still active on Day 7 and Day 30, by channel.
CRM campaigns you can’t prove
Push, email and WhatsApp journeys get credit for orders that would have happened anyway.
Understand. Audit. Fix. Grow.
A clear, four-step process. You always know what I’m doing, why, and what you get at the end of each step.
Understand
Your business, goals and growth model, before touching any tool.
- Business goals & core actions
- Your marketing ecosystem
- Teams, agencies & partners
Audit
Every data source, checked end to end against your real orders.
- MMP: Adjust, AppsFlyer, Branch
- Web: GA4, GTM, UTMs
- CRM: MoEngage, Insider, CleverTap
- Ad platforms & backend
Fix
Repair the broken tracking mechanisms so every number can be trusted.
- Event tracking plan & QA
- Deep links, postbacks, naming rules
- Fraud checks (e.g. mFilterIT)
- One source of truth
Grow
Decision-ready reporting, and a clear view of what to scale and what to cut.
- Cost per meaningful action
- Retention cohorts by channel
- CRM lift with holdout tests
- Weekly growth review
Follow one brand from messy data
to clear decisions.
Meet Sample App, a fictional food-delivery app spending ₹8 lakh a month across Meta, Google, influencers and affiliates. Here is exactly what happens, week by week.
One call and a short workshop. No tool logins yet.
Every source checked against real backend orders.
- HIGHPlatforms claim 1,600 orders. The backend shows 1,000. Meta and Google both count the same buyers.
- HIGHThe purchase event fires twice on the latest iOS build, so revenue looks higher than it is.
- HIGH3,750 of 12,500 installs from Network X are invalid (bots and click spam).
- MEDIUMInfluencer QR codes have no tracking link, so 1,800 users show up as “organic”.
- MEDIUMMoEngage push campaigns have no holdout group, so nobody knows what they really caused.
- LOWSix different spellings of “instagram” in UTMs split one channel into six rows.
Repair the tracking, then test it before trusting it.
| Area | Before | After |
|---|---|---|
| Orders | 4 platform reports | 1 de-duplicated view (Adjust + backend) |
| Purchase event | Double-firing on iOS | QA’d, matches backend |
| Influencers | Counted as “organic” | One Adjust link + QR per creator |
| Fraud | Paying for 3,750 fakes | Excluded via fraud check, ₹1.5L recovered |
| CRM | Push “drove” 800 orders | 10% holdout: real lift 200 |
| UTMs | 6 spellings | 1 naming rule for every team |
Now the numbers can make decisions.
Sample App is fictional. All numbers are sample data to show the method.
Reports built around decisions,
not decoration.
A preview of the dashboards I build. Click through the tabs. These use sample data for a fictional app.
Orders: claimed vs real
Real cost per order
| Channel | Spend | Orders | Cost |
|---|---|---|---|
| Meta | ₹3.0L | 450 | ₹667 |
| ₹2.5L | 300 | ₹833 | |
| Influencers | ₹1.5L | 150 | ₹1,000 |
| Affiliates | ₹1.0L | 100 | ₹1,000 |
Weekly cohorts: % still active
Day 30 retention by channel
Journey funnel
Valid vs invalid installs by network
Red flags this week
- Installs within 10 seconds of click1,920
- Activity between 2–5 am spikeNetwork X
- Installs with zero in-app events2,410
- Single-source spike vs last week+180%
Conversion: treated vs control
Journey performance
| Journey | Credited | Real lift |
|---|---|---|
| Welcome series | 1,240 | 410 |
| Cart reminder | 960 | 380 |
| Weekend discount | 800 | 200 |
| Win-back | 310 | 40 |
All dashboards above are illustrative mock-ups with sample data for a fictional app.
Three numbers most teams get wrong.
Cheap installs, expensive customers
₹5L on two channels. One costs ₹50 per install, the other ₹100.
The ‘cheapest’ network
12,500 installs at ₹40, then a fraud check flags 3,750.
The push that “drove” 800 orders
A holdout group shows most would have bought anyway.
Illustrative examples to explain the method.
Fix your ad data.
Then grow on it.
Tracking & Attribution Audit
A full check of how your data flows, and exactly what’s broken.
- MMP, GA4/GTM & CRM review
- Platform vs backend reconciliation
- Fraud & data-quality scan
- Prioritised fix list
Fix & Setup
I fix the tracking and set up one source of truth.
- Event tracking plan & QA
- Adjust / AppsFlyer setup & postbacks
- Deep links & UTM naming rules
- Core dashboards
Growth Analytics Partner
Ongoing analysis so every week’s decisions use the right numbers.
- Weekly growth review
- Retention & cohort analysis
- CRM lift & holdout tests
- Budget recommendations
I don’t believe in fancy dashboards. I believe in the right numbers.
I’m Harish Kumar, a data analytics expert with more than eight years in marketing analytics across India and the UAE. I’ve worked on growth measurement in consulting, e-pharmacy, streaming, travel, mobility and large-scale app launches, and I’ve seen the same problem everywhere: teams making big decisions on numbers they can’t trust.
“If you can’t trace the journey, the data will invent one for you. Usually it calls it organic.”
- NOW
Data Analytics Lead, Interesting Times · Founder, Explain Data - DUBAI
Almosafer · marketing & CRM analytics - BENGALURU
Disney+ Hotstar · subscription & attribution analytics - BENGALURU
PharmEasy · cohort retention & growth - NEW DELHI
ZS Associates · business intelligence consulting
Questions, answered.
Who is this for?
App and e-commerce brands, D2C companies, startups and marketing teams in India, the UAE and the wider Gulf who spend on ads and want to know what’s really working.
Which tools do you work with?
Adjust, AppsFlyer and Branch for apps; GA4, Google Tag Manager and BigQuery for web and data; MoEngage, Insider and CleverTap for CRM; plus Meta, Google and other ad platforms. If you use something else, we’ll check it together.
How do I get started?
Fill the short form on this page or email explaindataofficial@gmail.com. Harish reads every request personally and replies within 24 hours to fix a time.
What happens on the free audit call?
30 minutes. You share your goals and current setup; I point out the likely gaps in your tracking and what to fix first. No obligation, no sales pressure.
Do you need access to our accounts?
For the audit, read-only access is enough. I’m happy to sign an NDA, and I only look at what’s needed to check the tracking.
How long does it take to see results?
The audit takes 1–2 weeks. Most tracking fixes go live within 2–6 weeks, and you’ll see clearer numbers from the first clean week of data.
How much does it cost?
It depends on your setup and scope. After the free call, you’ll get a clear proposal with a fixed price for the audit or project.
Let’s find what’s breaking in your marketing data.
A free 30-minute data audit call. Send a short request and Harish will reply personally by email to fix a time.
- Free, 30 minutes, online
- Clear next steps, even if we don’t work together
- India, UAE & worldwide