Cold email conversion to meeting booked averages 0.5-1.0%. Best-in-class sits at 2-3%. If you're below 0.5%, your list is the problem. Not your subject lines. Not your follow-up sequence. Your list.
We've watched hundreds of campaigns run through MiraReach. The pattern is consistent: fixing list quality produces a 200% lift in meetings booked. Fixing copy produces 20%. That's not a typo.
Conversion rate to meeting booked is the only metric that matters
Open rates are noise. Google and Microsoft's DMARC crackdown made sure of that. Reply rates are better, but they conflate "thanks, not interested" with "let's talk." Click rates depend on whether you included a link, which half the time you shouldn't.
Meeting booked is the number your board cares about. It's the number that maps to pipeline. It's the number you can defend.
Here's the benchmark we see across B2B SaaS and services:
- 0.5-1.0% — typical. Most campaigns land here. Some weeks you're at 0.3%, some at 1.2%. It averages out.
- 1.0-2.0% — good. You're doing something right on targeting or offer.
- 2.0-3.0% — best-in-class. Rare. Usually means tight ICP definition and a genuinely useful first touch.
- Above 3.0% — either you've found a niche nobody else is working, or you're counting meetings that don't happen.
If you're at 0.5% and want to get to 1.5%, you have two levers: list and copy. They are not equal.
List quality sets the ceiling. If your ICP definition is loose — say, "B2B companies with 50-500 employees" — you're emailing thousands of people who will never buy, and no amount of copywriting rescues that. Copy determines how close you get to the ceiling. A sharp list with mediocre copy might hit 1.0%. A sharp list with sharp copy hits 2.0%+. A bad list caps you at 0.3% regardless.
This is why we push teams to instrument the funnel at the meeting-booked stage rather than the reply stage. Replies are a leading indicator that lies. A prospect who replies "not right now" is not a reply you can forecast against. A meeting on the calendar is. When you track meeting-booked as your primary conversion metric, you stop optimizing for vanity signals and start optimizing for the thing that actually moves revenue.
The practical implication: audit your list before you rewrite your copy. Most teams at 0.5% don't have a copy problem. They have a targeting problem they've been avoiding because rewriting copy feels more productive than admitting the list is wrong.
List quality is a 200% lever. Copy is a 20% lever.
We ran a test last quarter. Same sequence, same offer, same sender. Two lists.
List A: 2,400 contacts pulled from a general B2B database. Job titles matched, company size matched, industry matched. Standard stuff.
List B: 2,400 contacts from the same database, but filtered down to companies that had posted a relevant job listing in the last 60 days. Same titles, same size, same industry.
List A booked 14 meetings. List B booked 41. That's a 193% lift from changing nothing but the filter.
Then we rewrote the copy for List A. New subject line, new opening, new CTA. Same list. It booked 17 meetings. A 21% lift.
Copy matters. But it's a rounding error compared to list.
The reason is simple: a great email to the wrong person is still a bad email. A mediocre email to the right person gets a reply.
What the numbers actually expose is a difference in intent signal. A job listing is a public, dated admission that a company has budget, a mandate, and a timeline for a specific problem. That's not a demographic match — it's a trigger event. When you filter on it, you're not narrowing your list, you're reordering it by probability of need. The same 2,400 contacts stop being a sample and start being a queue.
This is why list work compounds and copy work doesn't. A better subject line lifts one send. A better filter lifts every send that follows, because the underlying qualification logic persists across sequences, senders, and quarters. Copy is a variable you tune per campaign. List criteria are infrastructure.
Practically, that means the highest-leverage filters are almost always behavioral or temporal, not firmographic:
- Hiring signals — job posts, department headcount growth, new role types.
- Funding and filing events — raises, 10-K risk disclosures, regulatory registrations.
- Tech and process changes — stack migrations, vendor churn, tooling replacements.
- Leadership transitions — new VP or founder hires inside the first 90 days.
Firmographics tell you who could buy. Trigger events tell you who is currently buying. Most teams spend their optimization budget on the first category and wonder why the second outperforms it by an order of magnitude.
What "good list" actually means
Most founders think list quality means "correct email address." That's table stakes. A verified email to someone who will never buy is still a wasted send.
Good list means three things:
1. They have the problem you solve. Not "they're in the right industry." Not "they have the right title." They have a specific, current, painful problem that your product fixes. If you can't articulate that problem in one sentence, your list is too broad.
2. They know they have the problem. A company with a broken outbound process might not realise it's broken. A company that just hired three SDRs definitely knows they need to make those SDRs productive. The second company is a better prospect.
3. They can buy. Title matters less than authority. A VP of Sales at a 50-person company can sign a $5K contract. A Sales Manager at a 5,000-person company cannot. Filter for buying authority, not seniority.
There's a fourth dimension most founders skip: timing. Even a perfect-fit prospect with budget authority converts poorly if the problem isn't urgent this quarter. This is why trigger-based filtering matters more than static firmographics. A company that just raised a round, posted a job for the role your product replaces, or switched tools in the last 90 days is signalling active intent. A company that merely matches your ICP on paper is not. The difference between a 0.5% reply rate and a 3% reply rate usually lives here, not in the copy.
This is also why list size is a vanity metric. A 50,000-row list built on industry and title filters will underperform a 2,000-row list built on problem, awareness, authority, and timing. The smaller list costs less to work, produces cleaner signal on what's actually resonating, and doesn't burn your domain reputation on prospects who were never going to reply. Volume doesn't fix a targeting problem — it just makes the failure more expensive.
We built MiraReach's prospect scoring around these signals. It's not magic. It's just filtering for people who are actually in-market.
The copy still matters, just not as much as you think
Once your list is tight, copy becomes the difference between 1.5% and 2.5%. That's worth chasing. But don't start there. The reason copy gets over-weighted in cold email advice is that it's the only variable most people can see. You can't A/B test a bad list against a good one in a single afternoon, but you can rewrite a subject line in five minutes. That asymmetry produces a lot of confident advice about a second-order lever.
What works on a good list:
Short. Under 100 words. Under 75 is better. The first sentence should tell them why you're emailing. The second should tell them what you do. The third should ask for something small. Length is a proxy for confidence. A long email reads like someone who isn't sure the recipient will reply, so they pre-empt every objection in advance. That instinct is backwards — objections are better handled in the reply, where you have their attention and a reason to keep it.
Specific. "I noticed you're hiring SDRs in London" beats "I saw your company is growing." Specificity signals you did the work. It also signals you're not a bot. The second point matters more than it used to. Recipients now pattern-match for automation, and generic praise reads as a tell. A detail that could only come from reading their job board, their changelog, or their last funding announcement does double duty: it proves relevance and it proves a human was involved.
No links in the first email. Links trigger spam filters. They also give the reader an escape hatch. Ask for a reply, not a click. A reply is a commitment signal; a click is a maybe. You want the former in your pipeline, because replies convert to meetings at a rate clicks never approach.
One ask. "Worth a 15-minute call?" is fine. "Let me know if you'd like to see a demo, or a case study, or a pricing sheet" is not. Multiple asks force the reader to make a decision about which ask, which is a decision they'll resolve by not replying.
We've written about why solo founders are ditching the $150 stack for simpler tools. The same principle applies here: fewer moving parts, better results.
What we'd do next
If your conversion rate is below 1%, stop rewriting your emails. Pull your last 500 sends and look at who you sent them to. Find the 20% that got replies. What do those companies have in common? That's your new filter. Look for structural signals rather than surface ones: headcount band, funding stage, whether they've hired into the function you're selling to in the last two quarters, whether their tech stack overlaps with what you integrate. Those attributes are durable. A subject line that worked once is not.
If you're already above 1.5%, then copy is your lever. Test one thing at a time. Subject line, opening sentence, CTA. Give each test 200 sends before you call it. Below that threshold you're reading noise, not signal — a 2% swing on 40 sends tells you nothing you can act on. Keep a control in every test and resist the urge to stack changes, because a combined win tells you which variable moved the number.
Two process notes that matter more than most teams admit. First, define "meeting booked" before you measure it. A calendar hold that never gets confirmed is not a booked meeting, and counting it inflates your rate against a benchmark you can't actually hit. Second, track reply-to-meeting separately from send-to-reply. A high reply rate with a low booking rate usually means the CTA is asking for too much, not that the list is wrong.
And if you want to see how MiraReach handles list scoring and inbox prioritisation, give MiraReach a try. We never send without you pressing the button.
— Mira