Blanket discounts give away margin
Site-wide promotions and coupon codes discount visitors who would have paid full price anyway—margin given up with no incremental sale to show for it.
The right price, decided in real time.
Real-time, visitor-by-visitor decisioning—the existing experience, or a targeted incentive for visitors least likely to convert at full price.
Live policyHealthy
+3.2%estimated lift
Standard54%
Free shipping27%
10% off19%
The cost of guessing wrong on price.
No discount leaves revenue on the table
Visitors unlikely to buy at full price simply leave—without ever seeing the incentive that could have converted them.
Decide per visitor. Prove it with a holdback.
Three principles keep the model honest and the experience untouched until you're ready.
Signals
Ad data + on-site behavior
Campaign, ad set, placement, and creative performance combine with on-site behavior and customer-journey telemetry to identify visitors unlikely to convert at full price.
Proof
Randomized holdback
Within the eligible, lower-intent audience, half receive the existing experience and half receive the incentive—isolating the incremental effect instead of crediting sales that would have happened anyway.
Control
Nothing changes until you say so
No customer-facing changes during data review and modeling. When an offer does launch, it's displayed in your own page and visual language—no pop-ups, banners, landing pages, or countdowns. You can pause or disable it at any time.
Standard data inputs
Website & journey
Site events, customer-journey telemetry, and historical website, promotion, and customer data
Context
Device, location, time, referral information
Ad platform
Campaign, ad set, placement, creative, and acquisition performance (e.g. Meta Ads)
Outcomes
Purchases, plan selection, cancellations, refunds, payment failures
Three phases, your approval before anything ships.
Phase 1 · 0–8 weeks
Data Review & Collection
Shortened if historical data is usable. No customer-facing changes during this phase.
Phase 2 · 1–2 weeks
Modeling & Experiment Proposal
Once validated, we propose a targeting plan: what share of visitors to include, and what incentive to test.
Phase 3 · ~6–12 weeks
Targeting & Measurement
Controlled rollout with regular reporting on contribution profit and revenue impact.
Contribution Profit per Eligible Visitor
The primary metric, accounting for incremental conversions, margin given up, plan mix, variable costs, payment/refund/cancellation effects, and customer value over the measurement window.
Supporting metrics
Revenue per visitor, conversion rate, plan mix, average order value, cancellation/refund behavior, and LTV where estimable.
In every report
Probability of positive incremental profit, estimated annualized profit and revenue improvement, whether company-wide improvement hits ≥1%, and uncertainty ranges.
Illustrative revenue impact
Annual revenue
1% lift
2% lift
3% lift
4% lift
$5M
$0.05M
$0.10M
$0.15M
$0.20M
$10M
$0.10M
$0.20M
$0.30M
$0.40M
$25M
$0.25M
$0.50M
$0.75M
$1.00M
$50M
$0.50M
$1.00M
$1.50M
$2.00M
Illustrative only—actual lift depends on traffic, margin, and signal quality.
Before anything launches
Agreed in writing, up front
- ✓Eligible visitor population
- ✓% of traffic included
- ✓Offer & presentation
- ✓Max margin exposure
- ✓Brand / CX restrictions
- ✓Contribution-profit calculation
- ✓Pause / stop conditions
What Adaptly provides
End to end, with your team
- ·Technical & data feasibility review
- ·Website data collection & integration
- ·Historical data analysis
- ·Model development & validation
- ·Targeting & offer analysis
- ·Experiment design & implementation
- ·End-to-end website integration
- ·Monitoring & financial analysis
- ·Working sessions & direct team access
- ·Final results & deployment recommendations
Find out what it's worth before you change anything.
A short data review tells us whether your traffic and signal quality support this. No customer-facing changes, no commitment beyond that review.