cold email personalisation that actually works

weekendhustle

i've been playing around with personalised cold email automation lately, and honestly, it's been a game-changer for reply rates. The stack I'm using is n8n for the workflow logic, Perplexity for research, OpenAI to draft the message, and Instantly AI for sending. No more copy-paste nonsense - every email feels like it's written just for that person.

what I focused on:

Scraping leads from Apollo and Apify (though I've heard Apollo's data quality has slipped recently - a colleague mentioned switching to Prospeo for fewer bounces)
Building the n8n workflow to pull in personalised details before the email gets drafted
Setting up Instantly AI with a proper follow-up sequence
Getting the domain infrastructure right (warm-up, SPF/DKIM, etc.) so emails actually land in inboxes

The secret sauce is the personalisation logic in n8n. Most guides skip that bit. Once you connect the lead source to the research step, then feed it into OpenAI for a tailored draft, the reply rates jump. my bounce rate is still decent with Apollo, but I'm curious if anyone else is seeing it rise lately.

Would love to compare notes if you're running a similar setup - no DM needed, just share your experience here. 😊

chatbox

interesting stack, but I've found Apollo's data quality has been slipping lately - lots of stale contacts hitting our lists. We swapped the lead sourcing to Prospeo a few months back and kept everything else exactly as you described. the personalisation layer through n8n still does the heavy lifting, but the bounce rate has dropped noticeably, which does wonders for deliverability scores.

What sort of bounce percentage are you seeing with Apollo leads? we were creeping towards 12 % before the switch.

ctrjunkie

are you still tracking open rates? Honestly, they're total vanity metrics. Do more harm than good in my experience. What's your actual positive reply rate?

budgetburner

81% open rate? Who actually buys that metric anymore? Everyone knows open tracking is broken-Apple's Mail Privacy Protection killed it dead. Reply rate and booked meetings are the only numbers worth your time. Stop chasing vanity metrics and start writing emails that make people actually want to respond.

olivia-roas

Open rates? Please. That's like celebrating impressions on a Meta ad that didn't convert a single pixel. Completely useless - and half the time the tracking's broken anyway from image blocking or privacy changes. The only number that keeps my lights on is positive reply rate, and even that gets polluted by autoresponders and meeting bots. Been running cold email alongside my ad campaigns for years - reply rate, cost per meeting, and pipeline velocity are the only things worth glancing at. Everything else is just noise that agencies sell to clients to justify their retainer.

copycat

I get the excitement-seeing high open rates on your first day feels like magic. But Apple's Mail Privacy Protection means those numbers are next to useless if your recipients use iPhones. The app auto-loads tracking pixels even if no one actually opened the email. without a single sale, it's worth asking: are you measuring real engagement or just a vanity metric? Open rates are a bit like a beautiful shop window-looks fantastic, but if nobody walks in to buy, it's just decoration. Keep experimenting, just make sure you're tracking the thing that actually pays the bills.

trafficcop

Oh, 81% open rate in 2026? Sweet summer child, that number is more fictional than my uncle's "I'll start the diet on Monday." Apple Mail Privacy Protection alone turns open rates into a fun game of "guess which pixel got triggered by a bot." Security gateways pre-clicking links? Also a factor. So either your list is made entirely of humans who never leave their inboxes, or you're selling snake oil with a fancy template.

vanity metrics like that are exactly why cold email gets a bad rep. Beginners see a shiny number and start optimising for opens instead of outcomes. The only numbers that matter: reply rate, positive reply rate, and meetings booked. Show me those, and I'll listen.

The stack itself - n8n, AI personalisation, Apollo - yeah, that's legit. Plenty of us run similar setups. But slap an 81% open rate on it and you've instantly lost credibility with anyone who's actually done this more than once.

Deliverability first. Then copy. then automation. Anyone chasing open rate optimisation is rearranging deck chairs while the spam folder swallows their ship. If the approach really works, the reply rate will prove it without needing a laughable claim.

Also, "template included"? Bold move for something that's supposed to be personalised. But let's not get started on that.

analyticsjunkie

Bounce rates have been kicking my arse lately - might actually give this a go.

jamesaso

yeah, open rate in the title made me twitch too. with Apple Mail privacy and bot opens, it's basically a "is my domain getting firebombed?" metric now. The only numbers that actually matter are positive reply rate and booked calls or demos. Would've been far more useful if OP had shared something like "X% positive replies and Y% turned into meetings" rather than flexing 81% opens

weekendhustle

That's a decent return - I've seen comparable numbers from personalised outreach like this 👌

ctrjunkie

honestly, 5% isn't terrible. the real trick is getting replies - focus on better lead quality or a stronger offer