Why outbound stops working after month two and how to fix it
Outbound not working after a strong start is almost always one of four things: a dry list, a burned domain, a stale message or a team that cannot keep up.
ON THIS PAGE · 8 SECTIONS
THE SHORT VERSION
- Outbound stops working when one of four layers breaks: the list runs dry, sending domains lose reputation, the message goes stale or reply volume outruns the team handling it. It is almost never all four at once.
- Gmail requires bulk senders to stay under a 0.3% spam complaint rate and recommends under 0.1%, per its own sender guidelines published in 2026. Cross either line and inbox placement drops before you notice in reply rate.
- Saleshandy's 2026 analysis of 40,006 sequences found 57% of positive replies come from follow-ups, not the opener, but reply rate per step falls from 0.80% on the first follow-up to 0.37% by the fifth.
- B2B contact data decays at roughly 2.1% a month according to Apollo's research summary, so a list built in January is already down double digits in accuracy by month two without a refresh.
- Diagnose the layer before you touch anything. A new message sent into a dead list or a burned domain fails for the same reason the old one did.
Outbound stops working after month two because one of four things degraded: the account list ran out of fresh fits, the sending domains lost reputation under real volume, the message went stale after the easy wins or reply volume passed what the team behind it could handle. Month two is when all four finally show up in the numbers, because month one runs on momentum that papers over all of them.
This piece breaks down each of the four, the exact signal that tells you which one broke and how long each one takes to fix once you know. If you want someone checking these four layers every week instead of guessing from a dashboard, this is the weekly review we run inside every outbound system we build.
The four things that actually break
A founder watching reply rate fall usually reaches for a new subject line first. That fixes the problem maybe a quarter of the time. The other three quarters, the message was never the issue.
The four are the list, the domains, the message and the team's capacity to handle what comes back. Each one fails on its own schedule and leaves a different fingerprint in the numbers. Confusing one for another wastes the weeks it takes to fix the wrong thing.
Your list runs dry
A list is not a fixed asset. It decays the moment you stop maintaining it and empties the moment you have sent to everyone who fits. Apollo's research summary on B2B contact databases puts average data decay at roughly 2.1% a month, which compounds to over 20% a year even before anyone leaves a job or a company gets acquired.
The faster version of the same problem is simpler: you ran out of accounts. A market of 3,000 to 4,000 good-fit accounts gets fully touched in four to six weeks at a normal daily volume. By month two, a team that has not re-segmented or widened the filter is sending to the same names for the second or third time. Those names have already decided.
THE USUAL WAY
- Build one list at kickoff and run it until it stops replying.
- Treat a quiet list as a copy problem.
- Re-send the same accounts with a new template.
THE SYSTEM WAY
- Track how much of the TAM has been touched, not just how many emails went out.
- Refresh and re-verify contact data on a schedule, not when it visibly breaks.
- Re-enter old accounts only on a new signal, never on a blank re-send.
Our piece on building a TAM list covers how to size the market correctly before the first send, which is the step that prevents this exact failure. If the market was sized right at the start, month two is a data refresh problem, not a market size problem. Waterfall enrichment, covered in our piece on that process, is the usual fix for the refresh side.
Your domains lose reputation
New sending domains get a grace period. Inbox providers have no history on them yet, so they place mail generously while they build a profile. That grace period runs out right around month two, which is exactly when a lot of teams are also increasing volume because month one "worked."
Gmail's own sender guidelines, published in 2026, set two numbers: keep the spam complaint rate reported in Postmaster Tools under 0.1% as a safety margin and never let it reach 0.3%, which triggers automatic spam classification. Both thresholds are measured on Google's side, invisible to you until deliverability already dropped.
Source: Google, Email sender guidelines, 2026
The symptom is specific: open rate and reply rate fall together across every segment and every message at once, not just on your weakest offer. That is a domain problem, not a message problem. No amount of rewriting fixes it. Our piece on outbound infrastructure architecture covers the warmup and monitoring setup that keeps this from happening in the first place.
FROM THE YARD
The clearest tell we have found: pull open rate by domain, not by campaign. If one domain out of five is dragging the average down while the other four hold steady, that domain took a reputation hit and needs to be paused and rewarmed, not blended into a company-wide number that hides it.
Your message goes stale
Every message has a shelf life. The easiest replies come first regardless of how good the message is. Saleshandy's 2026 analysis of 40,006 cold email sequences, run across 2,906 sending accounts between July and October, found that 57% of positive replies came from follow-ups rather than the first email, but reply rate per step still falls steadily through the sequence.
| Sequence step | Reply rate | Share of all replies |
|---|---|---|
| Follow-up 1 | 0.80% | 29.0% |
| Follow-up 2 | 0.57% | 15.3% |
| Follow-up 3 | 0.56% | 7.3% |
| Follow-up 4 | 0.46% | 2.9% |
| Follow-up 5+ | 0.37% | 2.1% |
Follow-ups still carry most of the volume of replies. The decline is the warning: the same angle sent again and again earns less every time it goes out.
Source: Saleshandy, cold email statistics, 2026
At the campaign level the same decay shows up over months, not just within one sequence. A 2026 benchmark report from Belkins, covering 7.5 million emails sent across 2025, found the average reply rate fell from 0.50% in the first half of the year to 0.40% in the second, a 20% drop industry wide as markets got more saturated with outbound in general. Your own message is competing against a steadily more crowded inbox, not just against itself.
The fix is not a better version of the same offer. It is a different offer, aimed at a different problem or a different signal. Our piece on outbound offer design covers how to build the next angle instead of polishing the first one past the point it can still convert.
Your team cannot keep up
The fourth failure is the one nobody blames first because it looks like success. Volume worked, replies came in, then they sat in an inbox for two days because nobody was watching for them at the pace they arrived. A reply that waits does not count less in the dashboard, but it closes less. A prospect who replied and got nothing back for 48 hours rarely replies to the next email either.
This failure shows up specifically in month two because month one volume is usually light while domains warm up. By month two, volume has ramped and reply flow has grown, but the review process is often still the one built for month one's trickle. Our piece on lead routing for outbound replies covers the qualification and assignment rules that keep a reply from waiting once volume outgrows what one person can track by hand.
A reply that waits two days closes at half the rate of one answered in five minutes. Slow outbound and no outbound fail the same way, just more expensively.
How to tell which one broke
Each failure leaves a different fingerprint. Check them in this order, because the first two are faster to rule out than the last two.
- Check open rate by domain. If one domain dropped while others held, that is reputation, not message or list. Pause and rewarm it.
- Check reply rate by segment against how much of the TAM you have touched. If the segments you hit hardest replied least, the list ran dry there first.
- Check reply rate by message variant, holding list and domain constant. If every variant fell at the same rate on fresh accounts, the message itself is not the problem; something upstream is.
- Check response time to replies against your own history. If replies are piling up or getting answered slower than they used to, capacity broke, not acquisition.
- Only then rewrite the message. A new angle into a healthy list and healthy domains, reviewed by a team that can respond fast, is the only version of a message test that tells you anything.
The fix and how long it takes
Once you know which layer broke, the fix is specific and the timeline is predictable. None of these are instant, which is the real argument for catching the signal early instead of waiting for month three to admit something is wrong.
| Layer | Symptom | Fix | Time to recover |
|---|---|---|---|
| List | Reply rate falls hardest in your most-touched segments | Re-segment, refresh data, widen or re-enter on a new signal | 1 to 2 weeks |
| Domains | Open rate falls on one domain across every message | Pause, rewarm, fix the authentication and sending pattern | 30 to 90 days |
| Message | Every domain and segment is fine, only reply content is weak | New angle tied to a different problem or signal | 1 to 2 weeks to test |
| Capacity | Response time climbs, replies sit unanswered | Add routing rules or a person watching the inbox | Days, once staffed |
Domain reputation is the slowest fix by far, which is why catching it in week one of the decline matters more than for any other layer.
Reputation recovery is the outlier on this list. Return Path's research, cited in a deliverability analysis from Mailtester, found senders who cleaned their lists and held bounce rates down saw real improvement in inbox placement within 60 days, with full recovery typically by day 90. A domain problem caught in week one of a decline is a month of reduced sending. The same problem caught in month three is a quarter with no reliable inbox at all.
Source: Mailtester, email reputation recovery timelines, 2026
If you are tracking the five numbers from our outbound KPI scorecard every week, this diagnosis takes minutes, because the scorecard already separates the symptoms by segment and channel. If you are only looking at one blended reply rate, it takes a lot longer to even notice something broke.
Questions about outbound not working
Why did my cold email campaign stop working after a few weeks?
Usually one of four things: the list ran out of fresh accounts, a sending domain lost reputation as volume ramped, the message went stale after the easiest replies were taken or reply volume outgrew the team's capacity to respond fast. Check which one by domain, segment and message variant before changing anything.
How long does it take to fix a burned sending domain?
Mailtester's review of Return Path research found real improvement within 60 days of cleaning the list and holding bounce rates down, with full recovery typically by day 90. That makes domain reputation the slowest of the four failure types to fix and the one most worth catching early.
Is it normal for reply rates to drop over time?
Yes, to a point. Belkins' 2026 benchmark report found average reply rates fell 20% industry wide between the first and second half of 2025 as outbound volume rose across the market. A steady, market-wide decline is normal. A sudden drop on one domain or one segment is not. That points at a specific fix.
Should I write a new cold email template if replies slow down?
Only after ruling out the list, the domains and the team's response time. A new message sent into a dried-up list or a reputation-damaged domain will fail for the same reason the old one did. Rewrite the message last, not first.
How often should I refresh my outbound contact list?
On a fixed schedule, not when it visibly breaks. B2B contact data decays at roughly 2.1% a month according to Apollo's research summary, so a list left untouched for two months has already lost meaningful accuracy before anyone notices in reply rate.
Sources: Google, Email sender guidelines, 2026, Saleshandy, cold email statistics, 2026, Belkins, cold email response rate benchmark, 2026, Apollo, B2B contact database decay rate, 2026, Mailtester, email reputation recovery timelines, 2026