Julien Paltrinieri
What went wrong here
The opener claims the prospect mentioned a specific product on their profile. He never had. Everything after that is the sender trying to keep an automated sequence moving while apologising for the thing that started it.
This is the failure mode of buying intent data you cannot inspect: when the prospect asks where the signal came from, there is no answer, and the sequence keeps sending anyway.
The opener is false, the sender cannot explain it, and the sequence continues for eight days after the prospect disengaged.
- Personalization5
- Conversational tone30
- Practical call to action20
- Achieved goal10
Scored on the same six criteria Emailx applies to every message it sends.
The message, as it arrived
Y.Enterprise technology consultant1st
Random one JulienFiller opener - I saw Heavy.AI mentioned on your profile.Untrue Are you using it regularly?
I don't think I've ever mentioned Heavy.AI, where did you see me mention it?
That's fair, we reach out based on LinkedIn and software signals, so it may have been a false lead on our end. My apologies for the confusion. Have you used Heavy.AI within the last 12 months?Sequence continues
No I did not. What did you use to pick up the signal?
No worries, sorry for the mix-up. We use LinkedIn and software signals to reach out, so false positives can happen. My apologies for the confusion.
We're also collecting feedback on other tools professionals commonly use. You can check the full list here.Question dodged
Unfortunately you didn't reply to my question asking which tool you used to pick up the software signal.
You're right, and I should have been upfront. I honestly don't know the exact tool that flagged Heavy.AI.The one good message We use a few data services that scan LinkedIn and software usage signals, but I can't see the source for each person. Appreciate you calling that out.
Hey Julien, just following up, any tools from the list you've used lately?Automation resumes
Hey Julien, just wanted to check in one last time. If any tools match your experience, happy to help with a quick review!Ignores the conversation
What went wrong
- 01The personalization was fabricated
The whole message rests on one claim about the prospect, and the claim is false. Unlike a generic opener, which is merely ignored, a false one is actively disproved by the reader in the first two seconds. Everything after it is discounted.
- 02Nobody could explain the signal
Asked twice where the data came from, the sender eventually admitted they cannot see the source for any individual prospect. If you cannot defend your reason for writing, you do not have a reason for writing.
- 03The sequence outran the conversation
The sender had a real human exchange, apologised well, and then the automation sent two more scheduled check-ins that ignored all of it. That is worse than never replying, because it proves the apology was not connected to anything.
- 04The ask changed halfway through
It opens about the prospect's software stack and ends asking him to write a product review for a gift card. Those are different transactions, and switching between them mid-thread reads as a bait.
The rewrite
Julien PaltrinieriFounder, EmailxJulien, you're building an outreach tool, so you're probably evaluating data and enrichment vendors constantly.Defensible inference
I run a review programme for exactly that category. Practitioners record a short review of a tool they actually use, and it gets published where buyers compare products.What it is
Is there a tool in your stack you have strong opinions about?One question
The rewrite gives up the fake precision and keeps the part that was always true: this person builds an outreach tool, so he evaluates data vendors. That is an inference from a public fact rather than a claim about a fact, and it cannot be disproved in one line.
It also stops pretending the message is about the prospect's software usage when it is really a request for a review. Naming the transaction immediately is less clever and much more likely to get an answer, because the reader can decide in one read instead of three messages.
One follow-up, then stop
Julien PaltrinieriFounder, EmailxJulien, closing this one out.Clean exit If reviewing tools isn't something you'd want your name on, no problem at all.
If it ever is, the programme is open.Door left open
The part that transfers
Intent data is only useful if you can inspect it. The moment you send a claim you cannot source, you are betting your credibility on a vendor you have never audited, and the prospect finds out before you do. Use signals you can point at: a hire, a launch, a post, a funding round. If the only reason you have for writing came out of a black box, that is not a reason.
Send the version that gets answered.
Emailx researches each prospect, drafts both channels and stops on reply.
Verified August 18, 2026. Senders are anonymised. Written by Julien Paltrinieri for Emailx.
