A woman in India holding a tablet by a lake

Social commerce

Using Machine Learning and Real Time Data to Acquire Female Resellers via UACs

Meesho wanted to reach women resellers at scale, and to pay less for every sign-up while doing it.

Objective: More female resellers, lower cost per sign-up
Approach: Universal App Campaigns, machine learning, real-time data
Funnel focus: Female reseller sign-ups and transactions

The brief

Reach women resellers, for less.

Meesho's reseller model depends on women selling to their own networks. The brief was to maximise the acquisition of female resellers while bringing down the cost of each sign-up.

A woman at work in a packing and print workshop

The approach

Machine learning, fed with live data.

We ran Universal App Campaigns and fed them real-time data, letting machine learning find the audiences that produced female resellers and move spend toward them as the signal came in.

A reseller celebrating on her phone beside a rail of dresses

Where the work focused

Sign-ups and the transactions behind them.

The focus ran past the sign-up to the transaction goals behind it. A reseller who signs up and never sells doesn't grow the platform.

A colourful textile market stall on a stairway

What it delivered

Resellers, and the sales behind them.

The results are set out below, from the share of female resellers to the monthly transaction goals.

Tell the machine what a good customer looks like, then let it look.

Reported results

Female resellers: %; As stated in the campaign headline
Female reseller acquisition through ads: %; As reported in the campaign material
Overachievement of monthly female transaction goals: %; As reported in the campaign material