Most outbound teams track one reply rate. One number, one trend line, one argument every quarter about whether the channel is working. That number is an average of segments that behave nothing alike, and the moment you split it by department the strategy question changes completely.
Will Allred published Lavender's cold email benchmark report covering 231,818 cold emails sent across roughly 50,000 active inboxes as of February 2026, segmented by department, seniority and industry. It is one of the few outbound datasets large enough to say something specific rather than directional, and the department-level cut is where it stops confirming what everyone assumed.
The benchmark nobody segments
Here is the shape of the data across the personas Lavender broke out. The A grade column matters more than it looks: it is the share of emails in that segment that met the quality bar, which is a measure of how much effort the market is currently putting into each persona.
| PERSONA | SHARE EARNING AN A GRADE | REPLY RATE EFFECT AT A GRADE |
|---|---|---|
| Finance | 6.1% | 79% lift — highest of any persona |
| Operations | — | 5.4% reply rate, a 58% lift |
| HR | 12.3% | 3.4% → 4.3%, a 27% lift |
| Technical buyers (eng/product) | — | 5.2% reply rate |
Two columns, two very different stories. Finance and HR both underperform on raw reply rate, but for opposite reasons: almost nobody writes a good email to finance, and the ones who do get rewarded more than in any other segment. HR gets better emails on average and still converts them at a lower rate, because the thing being optimized is not the thing that moves HR.
Operations and technical buyers sit above both, at 5.4% and 5.2% respectively when the email quality is there. Those two segments are where the standard outbound playbook was developed, which is worth saying plainly: the tactics most teams inherited were tuned on the personas that were already easiest to reach. Applying them unchanged to finance and HR is not a strategy, it is a default. The same drift shows up when teams port a working sequence into a new vertical and watch it flatten, which is most of what 2026 cold email response rate data actually documents.
Finance: the highest upside and the worst execution
Just 6.1% of emails reaching finance buyers cleared the quality bar. That is the lowest of any persona in the set, and it is not hard to see why — finance is the segment reps are least comfortable writing to. The temptation is to lead with the product and let the CFO translate it into money themselves, and that is exactly the email that gets deleted.
The 79% lift is the largest quality premium anywhere in the data, and it exists because the bar is so low. When almost every competing email in the inbox is generic, a message that names a specific cost, a specific timeline and a specific decision the reader owns does not have to be brilliant to stand out. It only has to be better than the 94% that were not.
This is the same pattern we see when we audit client sequences: the segment with the worst average copy is nearly always the one with the most available upside, because the competitive bar is set by everyone else's laziness. It is the reason cold email reply rate data keeps pointing back at message quality rather than volume.
HR: a cold email reply rate problem of tone, not targeting
HR emails hit the A grade 12.3% of the time — double the finance rate — and still average a 3.4% reply rate, below the cross-department average. Getting to an A moves it to 4.3%, a 27% lift, which is the smallest quality premium in the set. More effort, less return. That combination usually means the effort is going into the wrong variable.
Lavender's read is that HR buyers respond to warmth and punish anything that feels transactional, which tracks with the job. People whose work is people are unusually sensitive to being processed. A tightly personalized email that still reads like it came out of a sequence builder fails here in a way it would not fail with an engineering buyer.
“Nearly nine out of ten emails landing in HR inboxes have clear, fixable problems. Almost none of them are targeting problems.”
Practically, that means the levers are different. Drop the jargon in the first two lines. Make the ask small and human. Let the sentence structure be slightly less optimized than your engineering sequence, because the polish itself is a tell. None of this is a personalization token you can merge in, which is why HR sequences resist the automation that works elsewhere in the stack.
It also means the usual quality checklist misleads you here. Most scoring tools reward brevity, specificity and a clear CTA, and those are the right instincts for four of the personas in this dataset. For HR they are necessary but not sufficient — an email can score well on every mechanical dimension and still read as processed. The tell is usually the second sentence, where a sequence-built email pivots to the value proposition and a human one does not. That gap between mechanical quality and felt quality is the reason the first email wins most replies in segments where trust is the constraint rather than relevance.
Why one playbook cannot serve both
Put the two side by side and the conflict is structural. Finance rewards compression, numbers and directness. HR rewards warmth, context and a softer ask. A single template tuned to the midpoint underperforms with both, and the blended reply rate hides it — you see a flat 4% and conclude the channel is saturated, when what you actually have is two segments pulling in opposite directions.
Reply rate at A-grade quality, by persona (Lavender benchmark)
The bars above mix two units deliberately — reply rates for three personas and the quality premium for finance — because that is how the data is published, and because the finance bar is the one worth staring at. It is not a reply rate. It is how much you gain by being one of the few people who bothered. Treat the chart as a priority ranking, not a scoreboard, and it tells you where to spend the next sprint of copy work.
Rebuilding your sequences by department
Freezing volume is the part teams skip and the part that makes the exercise worth doing. If you change the copy and the list size at once you learn nothing, which is the failure mode behind most of the outbound programs we inherit — and the reason list size and reply rate keep getting confused for each other. Our own cold email engagements start with the segmentation precisely because it decides what the rest of the program is allowed to conclude.
What to change this week
Pull your last ninety days of replies and split them by department. If finance is in your ICP and sitting at the bottom, you have found the cheapest available win in your outbound program: a segment where the market average is bad enough that competent beats clever. If HR is your softest number, stop adding personalization and start removing polish. Then hold volume still long enough to prove which one moved.
See where you are cited today
A free snapshot audit of your rankings and AI citations before we ever talk.
Tyler leads work at the intersection of SEO and generative engines at Something Inc., helping B2B brands get ranked and cited across every major AI engine.