Message testing · on Starter and up
Find out which line lands before you send it.
Paste the subject lines and openers you are already weighing. A panel built to match your buyer ranks them and tells you which segment moved. Nothing here drafts the copy.
A test is one study. Starter is 20 a month. See pricing. Your first study is free on a work email.
The problem
The same stack writes the same email.
Tens of thousands of operators run the same GTM stack, and it all writes like this. The prospect has read it a hundred times. They delete it on sight.
A panel matched to your buyer can tell you whether your line reads like these before the list does.
Your lines to a ranked answer, in one pass
Paste the lines you are weighing
Two subject lines for a head-to-head, or one message you want read cold. You write the copy. Nothing here drafts it for you.
Pick who reads it
A national panel, one of 51 state panels, or an occupation panel. Respondents are built from census margins, so the read is for your buyer and not for the internet at large.
The panel reacts, then it tabulates
Each respondent answers in their own words, and the run scores the set. A/B ranks two variants head to head. Monadic scores one message on its own. Same engine as every other study, with a published 7.47% mean absolute error on 404 ex-electoral questions of a 460-question benchmark vs real polls (Pew, Gallup, UT/Texas Politics Project, PPIC).
Ship the winner, with the why
You get the ranking, the split by age, region and politics, the verbatims behind it, a confidence tier, and a PDF you can forward.
Your variants, ranked
You paste the lines you already wrote. They go to a panel and come back ranked for your segment.
An audience with splits
Respondents are built from census margins, so you get the split by age, region, and politics rather than one number for everyone.
Calibrated to real data
Published accuracy: 7.47% mean absolute error on 404 ex-electoral questions of a 460-question benchmark vs real polls (Pew, Gallup, UT/Texas Politics Project, PPIC). See the proof.
Straight answers
How is this different from Instantly, Clay, Gojiberry, or Smartlead?
Lewsearch does not write copy or send email. You bring the lines you are already considering, and a panel matched to your buyer ranks them and tells you which segment moved.
Does it guarantee more replies?
No. Copy is one lever; your list, offer, and deliverability matter too. What you get here is a read on your copy before you spend the list on it.
Do you write the variants for me?
No. There is no generator on this page and the model does not draft your outreach. You paste the variants, the panel ranks them.
What do I get out the other end?
The winning subject line or opener, a ranked list of every variant with a confidence tier, the segment splits, verbatims, and a PDF. Copy-paste ready.
Which plan includes it?
Starter and up. Starter is the survey and message-testing tier: 20 studies a month, and an A/B or monadic test is one study. The free study on signup can be a message test too, so you can see the output before you pay.
Stop guessing which line works.
Your first study is free on a work email, and a message test is one study. Bring two lines and see the output for yourself.