Case 01 · B2B wholesale

Auditing 288 conversions to find the 100 that were worth paying for

A wholesale business needed qualified buyers, not form fills. The account was reporting plenty of conversions. Two thirds of them were junk.

Client: B2B wholesale distributor (name withheld)
My role: Digital marketing intern, Adsplus
Timeline: Jun–Sep 2026
Stack: Google Ads · GTM · GA4 · Google Sheets

65% of recorded conversions were junk

Each dot is one conversion the account recorded.

188 retail enquiries, wrong numbers or duplicates100 real wholesale leads

Source: hand audit of all 288 recorded conversions, one row each.

Context

The client sells wholesale, so one genuine buyer is worth many times a retail customer, and a retail enquiry is worth nothing to them. The brief was lead generation with a budget of 100 million VND. Success meant buyers who could actually place a wholesale order.

What I did

1.Built the tracking first. A GTM container and GA4 events for form submissions, phone taps and the enquiry steps that mattered, so the account optimised toward a real action rather than a page view.
2.Structured Search around wholesale intent. Bulk, distributor and quantity-led queries, with aggressive negatives to keep retail traffic out.
3.Audited every recorded conversion by hand. Of 288 conversions, 188 were retail enquiries, wrong numbers or duplicates. 100 were genuine wholesale leads.
4.Fed the audit back into the account. Reported on qualified leads instead of raw conversions, and used what the junk had in common to refine keywords, negatives and the enquiry form.
5.Kept budget on what worked. Search top impression share held at 83% on the queries that produced qualified leads.

What happened

The campaign delivered more than 100 qualified wholesale leads at roughly 1 million VND each. The number that changed the conversation internally was the 65% junk rate: before the audit, the account looked like it was producing leads at a third of the real cost. Reporting the true cost per lead meant every budget decision rested on a number that was actually true.

The real cost per lead was about 3× the reported one

Cost per lead in VND, from about 100M VND of spend. Higher is worse.

Reported (spend ÷ 288 recorded)≈ 347k VND
Real (spend ÷ 100 qualified)1M VND

Source: account spend, recorded conversions and the hand audit.

What I learned

A conversion is only as good as its definition. The platform will optimise toward whatever you told it to count, and it has no way of knowing that two thirds of those records are worthless. Setting up tracking is not a chore that happens before the real work: it decides what the campaign will chase.

I would also automate the audit sooner. Going through 288 records by hand was the right way to learn what junk looks like, but by the second month I should have built a scoring sheet instead of repeating the work.