[ Company ]
We build rehearsals for commercial decisions.
Every price change and campaign is a bet you get to place once. Jonbar lets you play the versions against each other first, and tells you honestly how far to trust the result.
Each square is a simulated shopper, coloured by the version they bought in: A, B or C.
- Stage
- Private pilot
- Engine
- Simulation engine v0: price elasticity, promo lift, pull-forward, post-promo dip, cannibalization
- Backtest
- Blind backtest on public past decisions: being built
- Based in
- Istanbul
[ 01 ] Why we exist
You ship one version and never see the others.
The campaign week looks good; the weeks after decide whether it paid. We think that trade-off should be visible before you commit, and that a tool claiming to show it should be tested in public first.
[ 02 ] Principles
Four rules we hold ourselves to.
- [ 01 ]
Backtest before claims.
No accuracy number until the blind test is published, good or bad.
- [ 02 ]
Your data, not personas.
We model your market from your own sales history, not from asking AI what shoppers might say.
- [ 03 ]
Say what we don't know.
Thin history, summed channels, untested mechanics: the run says so.
- [ 04 ]
Recommend, don't operate.
Jonbar compares options. You decide and run the campaign.