Start a project

Setu Classes Sample case study

Twice as many “leads” as the counsellors could find.

Industry
Education, coaching
Location
Ballabgarh and NIT, Faridabad
Year
2026
Duration
Ongoing since April 2026
Services
Analytics, Tag Manager, WhatsApp journey, Reporting
Setu Classes SampleOne week in April 2026: enquiries reported by Meta and Google combined, against the numbers the counsellors could actually find and call.
  1. The platforms counted taps, duplicates and page loads.

Context

Setu runs foundation courses for classes 8 to 10 across two centres. Admissions happen in a six-week window, and Meta and Google both run during it.

The platforms reported plenty of leads. The counsellors said many numbers never answered or had already enrolled.

What they wanted
One number for enquiries that the admissions team trusts.
Constraints
  • Admissions window already under way
  • Counsellors work from WhatsApp on their phones
  • No CRM

What we found

  1. A tap on the WhatsApp button counted as a lead, even if no message was sent.
  2. The same parent often tapped three or four times across a visit.
  3. The form “submit” event fired on page load on one template.
  4. Meta and Google both claimed the same enquiries, so the combined total was double-counted.

The work

01

An event plan

Named the five events that matter, removed the rest, and fixed the form trigger. Every event now has a plain-language definition the counsellors signed off.

Setu Classes SampleThe event plan. Before and after counts are for the same week in April 2026.
  1. A tap on the WhatsApp button no longer counts as an enquiry.
  2. Reported enquiries fell by 42%, and now match the counsellors’ list.

02

A WhatsApp first reply

An automatic first message asks for the student’s class and centre. A reply to it, not a tap, is what now counts as an enquiry.

Setu Classes SampleThe automatic first reply. A parent’s answer to it is what counts as an enquiry.
  1. One question, answered in a few words.

What happened

Early results, April to June 2026

42%

Reported leads that were not real enquiries

Duplicates, accidental taps and page-load events

1

Weekly sheet everyone uses

Enquiries, callbacks and admissions by source

Early

Cost per admission

Visible for the first time. We will publish it after the window closes.

What we learned

Nobody needed a new dashboard. They needed the old number to mean one thing.

Aditi Rao, search and measurement Sample profile

Recognise any of this?

If your situation sounds similar, tell us what you’re working on. We’ll start with the business, then talk about channels.