which of your ads will work?
travenlabs
ABOUT
Traven Labs combines statistics, machine learning, and performance marketing to help companies put their ad spend behind the users worth keeping. Based in New York and Boston, Traven Labs works across rewards, subscription, CPG, e-commerce, agencies, and talent.
Traven Labs was founded by Harvard statisticians and got its start in rewards apps, where campaign quality decides every partner relationship. Our team of statisticians, machine learning researchers, and growth marketers is working toward a moonshot goal of bringing predictive intelligence to every company that runs ads. We want to democratize access to the modeling capabilities that, until now, only the biggest brands have had.
Trusted by
WORK
We build models that tell growth teams which campaigns are bringing in users worth paying for, within the first few days: campaign quality scores, LTV and ROAS forecasts, and custom models built around each client's economics.
Our models run across platforms serving more than 10 million users and power $80 million in annual ad-spend.
Clients use them to cut failing campaigns early and move that budget to the campaigns that pay back.
APPROACH
Because every business operates differently, each model is trained on your own history: your campaigns, your cohorts, your users. It learns to predict the outcomes you care most about, whether that's purchases, subscriptions, retention, or ad views. We tie its recommendations to your payback target and margins, and run it alongside your team every day.
Each cohort that matures makes the model sharper, and we retrain it as your channels, creative, and economics change. We stay on as part of your growth team.
CAREERS
If you are interested in reducing waste in our prodigal society and helping companies focus on building better products, Traven Labs might be the place for you. Our models decide where real acquisition budgets go, and every one is built for a specific business. If you want to do statistics and machine learning with that kind of consequence, we would like to hear from you.
Our team members have backgrounds in marketing, statistics, computer science, mathematics, economics, and archaeology!