+12.4%
Urgency wins back lapsed customers.
Messages framed around what they’re missing lift win back revenue. The same framing doubles unsubscribes among new customers. Valotrace sends to one and never the other.
Loading…
Loading…
Causal measurement, built in
Valotrace runs your email, push, and SMS end to end. It writes every message in your brand voice, measures exactly how much revenue each one created, and learns from every result to make the next send better.
Now onboarding a small number of teams.
Campaign 3 · win back email
Lapsed 60–90 days
incremental revenue vs. holdout
$0
created by this send
Learned: urgency beats discounts for lapsed buyers - and the discount was paying people who’d buy anyway.
Next: novelty framing for the 90 day lapsed segment, no discount.
Runs on top of the tools you already use
The problem
AI made it free to produce endless variants, so teams now send more than ever with less idea of what’s working. Your dashboards measure attention, not revenue: a campaign can top the leaderboard while reaching only the customers who were going to buy anyway, and the message quietly driving unsubscribes never shows up as a problem.
Valotrace answers the only question that matters: how much revenue did this message create? Every campaign is compared against customers who didn’t receive it - the same honest test behind every real experiment. Messages that create revenue get more of your audience. Messages that don’t, stop.
Attention
Easy to count. Easy to game. Not the same as value.
Incrementality
Measured against a holdout, so the number is revenue you created.
The loop
Write, prove, learn. Each pass makes the next one sharper, and your team stays in control at whatever level you choose.
01
Valotrace drafts every message in your brand voice across email, push, and SMS, personalized to each customer's history - ready to review, or ready to send.
Subject: Your saved items are almost gone
Hi Maya, the three pieces you saved are down to final stock. No codes, no countdown clocks: just a heads up before they’re gone for the season…
The playbook
Findings like these accumulate with every send. They’re specific to your customers, owned by you, and impossible to see from opens and clicks alone.
+12.4%
Messages framed around what they’re missing lift win back revenue. The same framing doubles unsubscribes among new customers. Valotrace sends to one and never the other.
−2.1%
Redemptions jump 9%, but net revenue falls because people who were buying anyway just pay less. Without a control group, this looks like your best campaign of the quarter.
+7.9%
Evening sends beat mornings during a sale and make no difference the rest of the year, so effort goes where it actually pays.
The measurement
Every campaign holds back a small, random slice of its audience. Comparing the two groups is the same standard used in clinical trials and by the largest experimentation teams in tech, and it’s the only way to know what a message earned, rather than what it happened to touch.
Random holdout. A slice of the audience never receives the message.
Deterministic assignment. The same customer stays in the same arm, every time.
Propensity logged. Every send records its assignment probability, so lift stays valid.
Win back campaign · lapsed customers
Built by people who’ve done this at scale
Valotrace is built by data scientists and machine learning engineers who ran experimentation and personalization for products with billions of users. The science of controlled experiments, causal measurement, and models that improve with every send runs entirely in the background.
Your playbook is yours.
Findings from your customers never train models for anyone else, and your data never leaves the loop.
Holdout
Always
Data shared
Never
Get started
We’re onboarding a small number of teams and running the first measured campaigns alongside each one.
Or email us directly.
Connects to Braze, Iterable, or Klaviyo.