Most teams track the wrong metrics. Open rate is broken. Click rate is meaningless. Here is what to measure, what the benchmarks are, and how to diagnose problems when numbers go wrong.
Rees Bayba
Founder, Astra GTM
TL;DR
Most cold email dashboards show the same handful of metrics: open rate, click rate, reply rate, bounce rate, unsubscribe rate. Not all of them are equally useful. Some are actively misleading. The teams running the most effective outbound programs have converged on a short list of metrics that actually predict pipeline -- and a set of metrics they have deliberately stopped tracking.
Open rate became unreliable in September 2021 when Apple Mail Privacy Protection launched. MPP pre-loads email tracking pixels when an email arrives in an Apple Mail inbox, regardless of whether the recipient actually opened the email. Since Apple Mail has 46%+ market share on mobile, a significant portion of your reported 'opens' are phantom opens triggered by Apple's servers. Open rate is now a noisy, inflated signal. Optimizing your subject lines for open rate and then celebrating a 60% open rate is optimizing for a broken metric.
Click rate measures how many people clicked a link in your email. For cold outbound, this is almost always low (under 1%) and has weak correlation with meeting rate. Cold emails with links also flag spam filters at higher rates than plain-text emails. The better measurement question is not 'did they click' but 'did they reply.'
Sending 10,000 emails a month is not a performance metric. It is an activity metric. Sending 500 highly targeted emails to the right ICP with strong copy will produce more meetings than sending 10,000 generic emails to a bloated list. Volume is an input, not an output.
Formula: (total replies / emails sent) x 100. This is your primary copy and targeting signal. It measures whether your message was worth responding to at all -- positive, negative, or neutral. A reply means a human read your email and chose to engage. That is the first gate every outbound campaign has to clear.
| Reply rate | Interpretation |
|---|---|
| Under 2% | Targeting or copy problem. Do not scale until resolved. |
| 2-3% | Below average. Testable. Run copy variants and ICP refinement. |
| 3-5% | Good. Solid foundation. Optimize from here. |
| 5-8% | Great. Scale the playbook. |
| Above 8% | Exceptional. Very narrow or well-researched ICP, or unusually strong copy. |
Formula: (interested replies / total replies) x 100. Not every reply is a buying signal. This metric tells you whether the people replying are actually interested, or just replying to say they are not. A healthy cold email program converts 35-50% of total replies into positive ones. Below 30% means you are reaching people who can respond to email but cannot buy -- wrong seniority level, wrong company size, wrong ICP.
Formula: (meetings booked / emails sent) x 100. This is the metric that connects email activity to revenue. Everything upstream -- reply rate, positive reply rate -- exists to drive this number. For most B2B cold email campaigns targeting $10K+ ACV, 0.5-1.2% meeting rate is the standard range. Below 0.5% with a good reply rate usually means friction in the CTA or a mismatch between who you're reaching and who can agree to a meeting.
This is the aggregate benchmark across all campaigns, ICPs, and industries on the Instantly platform. Well-targeted campaigns with strong personalization consistently outperform this. Generic campaigns with purchased lists typically land under 1%.
Formula: (hard bounces / emails sent) x 100. Must stay under 2%. This is a list quality metric, not a copy metric. If bounce rate climbs, you have a data problem -- unverified emails, stale contacts, or catch-all addresses passing your verification gate. See the bounce rate remediation guide for the full diagnostic process.
Should stay under 0.5%. Higher than 0.5% suggests either the wrong ICP (people who have no reason to care about your product), overly aggressive follow-up cadence, or copy that reads as spam. An unsubscribe is not the worst outcome -- it is a signal that you reached someone who will never buy and you should stop spending sends on them.
Each metric failure points to a specific upstream problem. Work backwards from what is low.
| Symptom | Likely cause | First thing to test |
|---|---|---|
| Reply rate under 2% | Wrong ICP, weak copy, or both | Run 100 sends to a tighter ICP segment. If reply rate improves, the original targeting was too broad. |
| Positive reply rate under 30% | Reaching people who cannot buy (wrong title, wrong company size) | Check the seniority and buying authority of the accounts replying negatively. Tighten title targeting. |
| High reply rate but low meeting rate | CTA friction or wrong seniority level | Simplify the CTA. Try 'Worth a 15-minute call?' instead of asking them to book a link. |
| Bounce rate above 3% | List quality -- unverified or stale contacts | Pause sends. Run the full list through verification at 97+ threshold. |
| Good reply rate, good positive rate, low meeting rate | Handoff friction or follow-up timing | Check how fast positive replies are being followed up on. Same-day follow-up on positive replies converts significantly higher than next-day. |
Cold email A/B testing requires discipline on two points: sample size and variable isolation. Most tests in the wild fail on both.
| Metric | Benchmark | Check frequency | Action if below benchmark |
|---|---|---|---|
| Reply rate | 3-5% | Weekly | Run copy variant test. Tighten ICP targeting. |
| Positive reply rate | 35-50% of replies | Weekly | Review which titles and company sizes are replying negatively. Adjust targeting. |
| Meeting rate | 0.5-1.2% | Weekly | Audit CTA and handoff speed on positive replies. |
| Bounce rate | Under 2% | Daily during active campaigns | Pause campaign immediately. Run full list re-verification. |
| Unsubscribe rate | Under 0.5% | Weekly | Check ICP relevance and follow-up cadence. |
Why is open rate still shown in my sequencer if it's unreliable?
Email platforms still report open rate because it is technically measurable and some recipients -- particularly on non-Apple mail clients -- do generate real open events. The problem is that Apple MPP has made it impossible to distinguish real opens from phantom opens in aggregate. Some platforms now offer 'adjusted open rate' that attempts to filter MPP-triggered opens, but even these estimates are rough. Treat open rate as directional at best and never use it as a primary optimization signal.
What counts as a positive reply?
Any reply expressing interest in learning more, asking for information, agreeing to a call, or requesting a demo. This includes 'yes, let's connect', 'can you send more info?', 'we might be interested, tell me more', and 'let's set up a call.' It excludes out-of-office replies, referrals to a different person, hard nos, and unsubscribe requests. When in doubt, classify conservatively -- a reply that requires multiple back-and-forth exchanges to determine if they are interested is not a positive reply.
How do I increase reply rate without increasing send volume?
Tighten your ICP, improve your copy, or both. Tightening the ICP means removing companies or contacts from the list that are unlikely to buy -- lower employee count, different industry, wrong tech stack. Improving copy means making the opening line more specific to the recipient, shortening the email, and simplifying the call to action. The highest-leverage single change is usually the first sentence: if it describes their specific situation accurately, everything else works harder.
What is a good meeting rate for enterprise targets?
Enterprise ICPs (companies with 500+ employees, $25K+ ACV deals) typically produce lower meeting rates than SMB ICPs -- often 0.3-0.7% -- because buying committees are larger, gatekeeping is heavier, and decision cycles are longer. A 0.5% meeting rate from an enterprise campaign is strong performance. Do not benchmark an enterprise campaign against SMB numbers.
Should I track metrics per campaign or across all campaigns?
Both, but separately. Per-campaign metrics tell you what is working in a specific ICP segment or copy approach. Aggregate metrics across all campaigns tell you the health of your overall outbound program. Mixing them together obscures both signals. A strong campaign can mask a weak one in aggregate. A weak campaign can make a strong program look average.
How do I know if my sequencer's reporting is accurate?
Cross-check reply counts against your actual inbox -- the raw number of replies you received should match what the platform reports. Bounce rates can be validated by checking your sending domain's postmaster data in Google Postmaster Tools or Microsoft SNDS. If your platform reports 1% bounce rate but Postmaster Tools shows a spike in invalid addresses, trust Postmaster Tools. Meeting rate should be validated against your calendar system, not just the platform.
We implement these systems end-to-end. First sends within 14 days.