Amninder Singh

Google-certified · #1 on SuperProf · Top 2 on UrbanPro

Meta Ads Case Study: 5,787 WhatsApp Enquiries for a Video Editing Institute in Punjab, and ₹29,000 That Brought Nothing

Meta Ads Case Study: 5,787 WhatsApp Enquiries for a Video Editing Institute in Punjab, and ₹29,000 That Brought Nothing

Last Updated on by Amninder Singh

Industry: skills training institute (video editing, social media, AI courses) · Market: Ludhiana and Punjab, India · Channel: Facebook and Instagram ads to WhatsApp · Period: January to November 2026 · Client name withheld

An institute in Ludhiana teaches video editing, social media marketing and AI tools, in a classroom and online. Its sales happen on WhatsApp: a student sends a message, a counsellor replies, and the student visits or pays. So the ads had one job, which was to start WhatsApp chats with people who might enrol. Over eleven months the account started 5,787 chats for the video editing course alone. It also spent about ₹29,600 on two ideas that produced nothing measurable.

Who ran what: this is my review of the whole account. The campaigns from January to March 2026, including the overseas ads and the ₹99 webinar, were set up by another marketer working with the institute. I set up the campaigns of October and November 2026. The analysis, the lessons and the next steps are mine.

The numbers for the video editing course

Phase Setup Spend WhatsApp chats started Cost per chat
January, Punjab Engagement objective, interest targeting ₹6,481 1,349 ₹4.80
February, rest of India Engagement objective, interest targeting ₹5,147 866 ₹5.94
March, rest of India Engagement objective, interest targeting ₹2,160 344 ₹6.28
May, Ludhiana, 16 km radius Leads objective, customer list plus 1% lookalike ₹4,538 317 ₹14.32
June, Ludhiana city Leads objective, customer list plus 1% lookalike ₹5,532 380 ₹14.56
June onwards, all of Punjab Leads objective, customer list plus 1% lookalike, one video ₹36,502 2,531 ₹14.42
Total ₹60,360 5,787 ₹10.43
5,787 WhatsApp Enquiries for a Video Editing Institute in Punjab, and ₹29,000 That Brought Nothing (chart)

What worked

1. Targeting people who want freelance income

In January the earlier marketer ran two ad sets side by side in Punjab with the same ad. One targeted people interested in freelancing and freelance marketplaces. The other targeted people interested in Adobe Premiere Pro and After Effects.

Ad set Spend Reach Chats Cost per chat
Freelancing interests ₹4,242 129,913 895 ₹4.74
Editing software interests ₹2,137 62,421 434 ₹4.92

The cost was almost the same. The freelancing ad set took twice the spend and reached twice as many people without the price moving. People who want to earn from home do not need to know the name of an editing program yet, and a beginner course is for them.

2. One Punjabi video carried ₹36,502 without the cost rising

The largest campaign ran a single video to all of Punjab. The text was in Punjabi and made a specific argument: joining two clips is not editing, real editing is what keeps a viewer watching, and the course ends with how to find clients on Fiverr and Upwork. It promised recordings of any missed live class.

The Ludhiana tests in May and June cost ₹14.32 and ₹14.56 per chat on about ₹5,000 each. The Punjab-wide campaign spent seven times more and held ₹14.42. It reached 437,025 accounts, showed the ad 979,631 times, and each account saw it 2.24 times on average. A cost that stays flat while the budget grows seven times is the sign of an ad that can be scaled.

3. The November relaunch: ₹4.62 per chat

In November I set up a new campaign that went back to the Engagement objective with WhatsApp as the destination, aimed at all of Punjab, ages 20 and above, using a 1% lookalike of the institute’s past video editing students with video editing interests added. The text was three short lines in English.

It spent ₹947 and started 205 chats at ₹4.62 each, with a 1.93% click rate. That matched the January cost after five months of paying around ₹14. A week later a city-level campaign for the new Amritsar branch started 31 chats at ₹11.18 on ₹346, which is reasonable for one city.

4. Video editing was the cheapest course to advertise

Course advertised Typical cost per WhatsApp chat
Video editing ₹4.62 to ₹6.28 (Engagement objective)
Social media marketing ₹9.80 to ₹15.43
AI tools ₹9.50 to ₹14.82

The same account, the same period, the same style of ad. Video editing started chats at about half the cost of the other two courses. For an institute with several courses, that is a reason to lead with video editing in ads and offer the other courses inside the chat.

What did not work

1. Ads to Italy, the UK and Australia: ₹12,043 for almost nothing

Country Spend Cost per 1,000 impressions Link clicks Cost per click
Italy ₹6,765 ₹160.83 375 ₹18.04
UK ₹4,652 ₹457.39 134 ₹34.72
Australia ₹626 ₹783.86 4 2 chats at ₹312.76 each

The idea was to reach Punjabis living abroad for the online course. Three things went wrong. The Italy and UK campaigns were optimised for link clicks, so Meta looked for people who click and not people who message. The Italy ad set was named for people who had lived in India, but the targeting inside it was a list of sports interests such as volleyball, fishing and surfing. And impressions in these countries cost 5 to 25 times more than in Punjab. A click in the UK cost more than seven chats did at home.

2. The ₹99 webinar: 10,316 clicks and no recorded sale

Three sales campaigns promoted a live social media webinar priced at ₹99. The ads did their part: click rates between 4.2% and 4.5%, the highest in the account, and 10,316 link clicks for ₹17,595. The webinar page loaded 5,982 times. Meta recorded zero purchases.

Two problems are visible in the ad data. About 42% of the clicks never became a page view, which points to a slow page. And a campaign told to find buyers received no purchase signal at all in three attempts, so either the purchase event was never set up on the payment page or nobody paid. Without a working purchase event, Meta was guessing for the whole ₹17,595.

3. Retargeting cost more than cold traffic

In August a campaign showed a “limited seats” video to people who had engaged with the institute’s Instagram in the last 90 days, within 40 km of Ludhiana. It spent ₹5,800 and started 234 chats at ₹24.78. Cold audiences were costing ₹14.42 at the time. The warm audience was small, many of them had already messaged, and urgency alone gave them no new reason to write.

4. Copies and relaunches of winning campaigns

Original Cost per chat Copy or relaunch Cost per chat
Punjab-wide video, June ₹14.42 Same campaign copied in October ₹52.80
November Punjab campaign ₹4.62 Copy made two days later ₹36.90

A copied campaign starts with no history. On a small budget it often never gets enough results to settle, and both of these were stopped under ₹600.

5. Instant forms and customer lists used as the audience

These were my own tests. An October instant form campaign for the AI and social media courses spent ₹1,187 and collected 1 lead. Two campaigns that targeted the uploaded student list directly, instead of a lookalike of it, paid ₹131 and ₹235 per thousand impressions and started chats at ₹22.61 and ₹104.95. A list of past students is a small audience, and small audiences are expensive to reach. The list is more useful as the seed for a lookalike, which is how the ₹4.62 campaign used it.

What I learnt

  1. The objective changed the price by three times. With WhatsApp as the destination, the Engagement objective started chats at ₹4.62 to ₹6.28. The Leads objective started them at about ₹14.40. Leads campaigns are built to find fewer, more serious people, so the cheaper chat is not automatically the better one. The counsellors’ enrolment record is the only way to settle it, and that is the next thing to measure.
  2. Sell the income, then the software. Freelancing interests matched software interests on cost while taking twice the budget. The winning Punjabi video talked about clients and standing on your own feet before it mentioned any tool.
  3. Stay where impressions are cheap until the home market is full. Punjab cost ₹20 to ₹40 per thousand impressions. The UK cost ₹457.
  4. Match the optimisation to the action you want. The link-click campaigns abroad delivered clicks and almost no chats. The conversation campaigns at home delivered chats.
  5. Never run a sales campaign before one test purchase shows up in Events Manager. This single check would have saved ₹17,595.
  6. Raise budgets on the winner and leave copies alone. The June campaign proved a good ad can absorb seven times the spend at the same cost.
  7. Use the student list as a seed. Lookalike of past students plus course interests gave the lowest cost in the account.

What I would set up next

  1. One always-on video editing campaign for Punjab on the Engagement objective, and one on the Leads objective, with separate WhatsApp greeting messages so the counsellors can tell the chats apart.
  2. A weekly count from the counsellors: chats, visits and enrolments per campaign. Cost per enrolment can then decide the budget.
  3. City campaigns for Ludhiana and Amritsar only when a new batch has a start date to announce.
  4. A fresh video every four to six weeks for the main campaign, since frequency on the June ad had reached 2.24.
  5. The ₹99 webinar only after the payment page loads fast on mobile data and the purchase event is confirmed.

Figures are from the institute’s two Meta ad accounts and were not adjusted. “Chats” are what Meta reports as messaging conversations started.

More Meta Ads case studies

If you want to learn to run and read campaigns like these, see my Meta Ads course. If you want them run for you, see my Meta Ads agency page. All of my Meta Ads case studies are on the Meta Ads case studies page. For enquiries across Google and Meta together, see my lead generation service. Related: video editing with Claude for digital marketing. You can look up any advertiser’s live ads in the Meta Ad Library.

Leave a Comment

Your email address will not be published. Required fields are marked *