Pick my brain!
Fill out a few details so I can prepare before we chat.
Making cold traffic pay for itself by predicting who is worth acquiring
ROAS vs benchmark campaign
Cold traffic does not have to be a loss-leader. Tell the algorithm what a good customer is worth and it will go and find more of them.
Generate incremental sales from Meta Ads cold traffic, cost-efficiently, and grow the retargeting pool while doing it.
For most brands cold traffic is a loss-leader: the sales come later, from retargeting. Ciaté needed prospecting that paid its own way rather than being written off as the cost of filling the funnel.
Lifetime value measured to date punishes your newest customers for being new. That’s recency bias.
A customer who joined last month looks worthless next to one who joined three years ago. Feed that to Meta and it learns to find more old customers. Predicting lifetime value instead of measuring it corrects the recency bias and changes who the algorithm goes after.
Every order pulled from Shopify into a data warehouse, customer by customer.
Machine learning scores each customer on what they bought, at what price, and estimated household income from third-party enrichment.
Emails and CLV scores go to Meta, which finds the prospects who look like the most valuable customers.
The pipeline. Predicted value, not value to date — which is what corrects the recency bias.
Smart audience against the benchmark campaign, on the three measures that were tracked.