(The exact audit protocol I personally run on every new customer's first 14 days of campaigns, built from nine-hour audit days across 2,000+ customers and a quarter million inboxes)
The Single Red Flag That Tells You Your Audit Is Overdue
before you read anything else, answer this one question honestly.
when did you last change your hook or your copy?
i'm not talking about tweaking a subject line or swapping a bullet.
i mean a genuine rewrite of the opening hook.
if the answer is 'not recently' and your reply rate has dropped, you're already looking at copy fatigue.
i had a prospect last week tell me his reply rate had gone from 5% to sub 1%.
he was convinced the infrastructure was broken and asked me to audit it.
first question i asked him was when he last changed his copy.
his answer was 'i haven't really'.
that's the audit done in 30 seconds.
we come up with about 10 hooks a week internally, and he'd been running the same one for months.
HERE'S THE 60-SECOND TRIAGE:
→ reply rate down, copy unchanged in 30+ days → copy fatigue, not infrastructure
→ reply rate down, bounce rate up → infrastructure problem, start at chapter 1
→ reply rate down, open rate flat → targeting or offer problem, skip to chapter 4
→ reply rate down, everything else looks 'fine' → this is the invisible one, run the full audit
write your diagnosis here: ______
now keep reading.
the rest of this sop walks you through the exact same checks i run on every 30-day audit call, nine hours a day.
Why This Audit Exists (and What Happens Without It)
i'll be straight with you.
i built this audit because i got tired of watching people change four things at once and then have no idea which fix actually worked.
when emails are declining in performance, there are only four reasons that can be the cause.
those reasons are infrastructure, targeting, the offer, and the copy.
there's no magical fifth pillar that just makes it work or not work.
but most people don't look at it through a logical lens.
they panic and change all four at the same time.
then they've got zero idea which variable actually caused the net positive or negative result.
THIS AUDIT FORCES DISCIPLINE.
you test one variable at a time, find the actual root cause, fix that thing, then move on.
here's the context for why this matters so much right now.
we've just launched onboarding calls for every new customer.
30 days after they join, they book an audit call with me personally.
i'm doing those calls nine hours a day at the moment, every day.
i go through their entire infrastructure, the first 14 days of campaigns, what's working, what's not, whether the setup they've got is the right size, whether they should upsell, downsell, or hold.
this sop is literally what i'm doing on those calls, written down.
if you run the audit properly, you'll catch at least one thing your current setup is getting wrong.
most teams catch three or four.
here's the thing about cold email that most teams underestimate.
it stacks.
when emails get filtered, engagement drops because fewer people see them.
lower engagement means more emails go unreplied and complaint signals go up, and your sender reputation takes a hit.
by the time your reply rate visibly drops, the reputation damage has been building for weeks.
and it's very, very hard to reverse.
the 30-day audit is how you catch it before it gets expensive.
yeah?
Infrastructure Sizing Check
most teams i speak to don't actually know if their infrastructure is the right size for what they're trying to do.
they've just bought what they could afford or what a provider recommended two years ago.
that's fine for day one, but it's not fine for day 30.
this is the first check i run on every audit call.
THE SIZING QUESTIONS I ASK EVERY CUSTOMER:
- how many emails a day are you actually trying to send right now?
- how many inboxes do you have live?
- what's the volume per inbox per day?
- are you running warmup in parallel with campaigns, or did you stop warmup when you launched?
- how many domains are underneath those inboxes?
if you were running accounts the way people did two years ago, pushing 30 to 40 emails a day out of a single inbox, you'd be getting instant bans.
we see this all the time.
i jump on a call and the customer tells me everything's broken.
second we look at volume per inbox, it's right there on the wall.
THE CURRENT SAFE RANGE:
across the quarter million inboxes we've got in circulation, the behaviour that extends inbox lifespan from the old 3-4 month ceiling to 8-12 months is lower volume per inbox plus parallel warmup.
the trade-off is real.
you stop maximising output per inbox and start maximising the lifespan of each inbox instead.
you keep volume within the range providers consider normal.
FILL THIS IN NOW:
target daily send volume: ______
current inbox count: ______
current volume per inbox per day: ______
parallel warmup running? yes / no
if your volume per inbox is above what most modern setups are running, or you stopped warmup at launch, your infrastructure is under-configured for the volume you're trying to hit.
THE THREE SIZING OUTCOMES:
→ RIGHT-SIZED: your volume per inbox is conservative, warmup is still running alongside campaigns, and reply rates are stable. keep going.
→ UNDERSIZED: you need more inboxes to hit target volume safely. buy more, warm them up properly, don't rush.
→ OVERSIZED: you're paying for infrastructure you're not using. consolidate, consider a downsell, put the savings into copy volume.
most people fall into the undersized bucket.
they're trying to hit agency-scale volume on a solo-operator infrastructure setup, and the math never works.
First-14-Days Campaign Performance Analysis
this is the meat of the audit.
the first 14 days of campaigns tell you more about the health of your setup than any dashboard metric in isolation.
the reason i personally sit down with new customers 30 days in and look at this specific period is because 14 days gives you enough data to spot a real trend without being so short that you're reading noise.
THE 14-DAY REVIEW I RUN:
i look at five things.
- volume actually sent vs volume planned.
if you planned 500 a day and you're sending 280, something's throttling you somewhere, which is a signal.
- bounce rate trend across the 14 days.
flat and low is healthy. climbing is a problem. a single spike is usually a bad list import.
- positive reply rate.
that means real replies that could book a meeting, not total reply rate.
60% of your reported opens are security scanners anyway, so opens don't tell you much. positive replies do.
- copy performance across multiple hooks.
if you launched with one hook only and it's declining, you don't have a data set. you've got a single data point dressed up as a trend.
- inbox-level performance split.
are ALL inboxes underperforming, or is it concentrated in 2-3? concentrated problems are usually inbox-specific. universal problems are infrastructure or copy-level.
THE FILL-IN TEMPLATE:
days 1-7 total sent: ______
days 8-14 total sent: ______
bounce rate day 1-7: ______
bounce rate day 8-14: ______
positive replies day 1-7: ______
positive replies day 8-14: ______
number of distinct hooks tested: ______
number of inboxes under 50% of mean reply rate: ______
THE DECISION TREE:
→ bounce rate climbing AND positive replies flat → infrastructure or list quality issue. start with list verification, then check authentication.
→ bounce rate flat AND positive replies declining → copy fatigue or targeting drift. change the hook, check that you're still emailing the right icp.
→ bounce rate flat AND positive replies climbing → don't touch anything. document what's working and scale slowly.
→ concentrated underperformance on 2-3 inboxes → rotate those inboxes, don't burn the fleet.
once you've filled in the numbers above, you know which variable to test first.
that's the whole point.
you're no longer guessing.
Upsell, Downsell, or Hold Decision Framework
this is the part of the audit most people skip.
they want to fix what's broken.
they don't want to think about whether they've got the wrong amount of infrastructure in the first place.
on every 30-day call i run, the third question after 'is the infrastructure right-sized' and 'what's the 14-day data showing' is whether you should add capacity, cut capacity, or hold.
THE 3 DECISION PATHS:
add more infrastructure (UPSELL) when:
- positive reply rate is stable or climbing and you're already running at target volume per inbox
- you're turning down leads because you can't follow up fast enough
- you want to run 2+ separate campaign tracks with different icps and you're trying to share inboxes across both (don't do that, split them)
- you've got a tested, working hook and you want to scale volume without breaking what works
cut infrastructure (DOWNSELL) when:
- you're paying for inboxes that are sat idle or sending under 30% of their safe range
- your copy or targeting is broken and adding more inboxes won't fix it (this is the most common one)
- you've over-bought because a provider convinced you you needed X when you really needed half of X
do nothing to capacity (HOLD) when:
- 14-day data is trending positive but you don't have enough volume to be statistically confident yet
- you've just changed copy or targeting and you want to see that variable play out before you add the infrastructure variable on top
- a provider just pushed a policy change and the whole market is finding its new normal. sit still for 2 weeks.
THE HARD QUESTIONS I ASK MYSELF BEFORE RECOMMENDING AN UPSELL:
- would this customer hit the same reply rate on half the infrastructure with better copy?
if yes, i'm not selling them more inboxes. i'm telling them to fix copy first.
- is their problem a supply constraint or a demand constraint?
if you're demand-constrained, more inboxes solve nothing. you need more hooks, more audiences, or a better offer.
if you're supply-constrained, you need more inboxes, simple as.
most people can't tell which one they're in.
that's half the reason this audit exists.
YOUR DECISION:
demand-constrained or supply-constrained? ______
upsell, downsell, or hold? ______
if you got upsell, the next question is which track you need more inboxes for, prospecting volume or follow-up capacity.
those are different builds.
if you got downsell, the next question is what minimum viable infrastructure hits your actual current need, not your aspirational need.
if you got hold, the next question is what one variable you're going to isolate and test over the next 14 days.
write that one variable down and stop changing anything else until that test runs.
The 4-Pillar Diagnostic (Infrastructure, Targeting, Offer, Copy)
when someone tells me their reply rate has dropped, i don't guess.
i run the four pillars in order: infrastructure, then targeting, then offer, then copy.
that's the logical lens.
the reason most people don't diagnose properly is they change all four at the same time.
if you change four variables simultaneously, you have zero attribution on which one fixed or broke the thing.
you've got to isolate.
PILLAR 1: INFRASTRUCTURE CHECK
- are SPF, DKIM, DMARC all configured and green? yes / no
- is authentication passing on every sending domain? yes / no
- is warmup running in parallel with campaigns? yes / no
- is volume per inbox within safe range? yes / no
- is bounce rate under 1.5%? yes / no
if all five answers are yes, infrastructure is fine and you can move to pillar 2.
if any are no, fix that first and don't touch anything else.
PILLAR 2: TARGETING CHECK
- is the list actually your icp, or did you broaden it recently? _______
- are you emailing the right role at the right company size in the right industry? _______
- when was the list last verified or refreshed? ______ (if older than 30 days, re-verify)
- is the list too big for one hook to work across all of it? ______ (if yes, segment)
if targeting is drifting, you'll see flat opens and dropping reply rates.
no amount of infrastructure tuning will fix this.
PILLAR 3: OFFER CHECK
most people don't touch this one because it feels like the biggest change.
it usually isn't. it's often a one-line edit.
- is the offer still relevant to this icp? ______
- has the market moved and your offer hasn't? (example: a year ago, getting brands on tiktok shop affiliates was cutting-edge. now everyone knows to do it. if your offer is 'i'll put you in touch with affiliates,' you're a year late. the market knows.) ______
- does the offer promise a tangible outcome, or is it vague? ______
offer problems look like copy fatigue but they're deeper.
changing the hook won't fix them.
PILLAR 4: COPY CHECK
this is where copy fatigue lives.
- when did you last change the hook? ______ (if over 30 days, you probably have fatigue)
- how many distinct hooks are you actively testing right now? ______ (we run about 10 a week internally. if you're running 1, you don't have a test, you've got a prayer)
- is the first email link-free? ______ (link patterns are one of the strongest structural signals AI filters use)
- does the copy read like a real person wrote it, or like an engineered opening? ______
- are you varying template structures across inboxes? ______ (30 inboxes using identical templates is a behavioural fingerprint that looks automated)
THE ORDER OF OPERATIONS:
- run infrastructure check first. fix any no's.
- if infrastructure is clean, run targeting. fix drift.
- if targeting is clean, check the offer. rewrite if the market has moved.
- if offer is clean, rotate copy. add hooks until you've got 10 fresh options.
only change one pillar at a time.
give it 14 days.
then reassess.
that's the entire science of cold email diagnosis.
it's not clever.
it's just disciplined.
The 30-Day Audit Scorecard
print this page.
tape it to your monitor.
use it every 30 days.
INFRASTRUCTURE SIZING
- target volume: ______
- inbox count: ______
- volume per inbox per day: ______
- parallel warmup running? yes / no
- right-sized / undersized / oversized? ______
14-DAY CAMPAIGN DATA
- volume sent vs planned: ______ / ______
- bounce rate day 1-7 vs day 8-14: ______ / ______
- positive replies day 1-7 vs day 8-14: ______ / ______
- number of distinct hooks tested: ______
- inboxes under 50% of mean reply rate: ______
UPSELL / DOWNSELL / HOLD
- demand-constrained or supply-constrained? ______
- decision: ______
- one variable to isolate next: ______
THE 4 PILLARS
- infrastructure check: pass / fail
- targeting check: pass / fail
- offer check: pass / fail
- copy check: pass / fail
- first pillar to fix: ______
KEY BENCHMARKS WORTH MEMORISING
- bounce rate safe line is under 1.5%
- 60% of reported opens are security scanners, not real humans
- average inbox lifespan with parallel warmup is 8-12 months
- we run about 10 hooks a week internally
- the 4 pillars of decline are infrastructure, targeting, offer, and copy
- there is no fifth pillar
ONE-LINE DECISION RULES
- if copy hasn't changed in 30 days and replies are down, it's copy fatigue.
- if bounce rate is climbing, fix infrastructure or the list first.
- if 2-3 inboxes are underperforming, rotate those. don't burn the fleet.
- if you're changing 4 things at once, you're not auditing. you're guessing.
speed. quality. communication.
