Four layers run end to end: ingestion, profiling, campaign generation and measurement. You connect the store once. OBV operates it from there.
Layer 01
A continuous feed of orders, line items, carts, sessions, refunds, loyalty events and subscription state, with full historical backfill on connection.
Layer 02
Per-customer models for purchase cadence, basket composition, price sensitivity, channel response and lapse risk. Profiles rebuild on every inbound event.
Layer 03
Campaigns are written, audienced, channelled and scheduled automatically against the moment each profile is most likely to convert.
Layer 04
Every send is attributed to orders and revenue, and the result is fed back into the models that choose what runs next week.
Pick a type to see the trigger, the audience it builds and the signal it reads.
{{ active.name }}
{{ active.body }}
Trigger
{{ active.trigger }}
Audience
{{ active.audience }}
Channel
{{ active.channel }}
Attributed revenue, incremental lift against holdout, cost per retained customer and spend against budget. One weekly summary, one live dashboard.
Weekly
Performance summary
Holdout
Incrementality, not last click
Per send
Orders and revenue attributed
Live
Budget and pacing
Week 1
We connect your store and backfill history, then walk you through the profiles that come out of it. You confirm guardrails: spend ceiling, send frequency, quiet hours, discount limits.
Week 2
First campaigns are generated and reviewed with your team. Cart recovery and reorder timing go out first because they pay back fastest.
Ongoing
Habit and frequency mechanics layer in, autonomy increases at your pace, and a weekly summary tells you what ran, what it earned and what changes next.
Thirty minutes, your numbers, the campaigns OBV would run this week.
{{ formNote }}