Deliverability has a lot of moving parts. Authentication, reputation, content, volume, list hygiene, monitoring, warmup, segmentation. Reading about it is overwhelming; doing it is exhausting.
Most founders, trying to find the starting point, end up doing nothing. Analysis paralysis. The counter-intuitive answer to "what should I do first" is not any of the technical items you'd expect. It's this:
Set up weekly seed testing. Three minutes every Monday morning. Paste your campaign into Inbox Check. Record inbox placement at Gmail, Outlook, Yahoo. Look at the per-provider breakdown. That's it.
Everything else — authentication fixes, volume ramps, content audits — flows from the data you collect here. Without this, you're guessing. With this, you have a compass.
Why this beats "fix SPF" as a first action
Fixing SPF is a perfectly reasonable piece of advice. It's also often irrelevant. About half the domains we audit have clean authentication and still poor placement. For those senders, hours spent on DNS produce zero improvement. The time would have been better spent finding the actual cause.
Weekly placement measurement tells you what the actual cause is. If placement is 80%, your deliverability is fine; stop worrying. If placement is 40%, you have a problem; the test result points at where. If placement dropped 20 points week-over-week, you changed something recently; the change is the cause.
In all three cases, the measurement is what makes the right next action obvious. Without measurement, any advice you follow is an educated guess.
What the weekly test actually does for you
Catches drift before it kills a campaign
Placement doesn't usually collapse overnight. It drifts. 75% one week, 68% the next, 61% the week after. By the time your reply rates visibly drop, you've lost 4–6 weeks of pipeline. A weekly test spots the trend after one data point.
Lets you correlate change to effect
You switch ESPs on March 1. If placement drops the next week, the cause is obvious. Without baseline measurement, you can't distinguish "ESP switch broke it" from "it was already broken."
Gives you per-provider signal
A seed test shows Gmail, Outlook, Yahoo, Mail.ru, Yandex, GMX, ProtonMail. If Outlook drops while Gmail stays flat, the diagnosis is Microsoft-specific (SmartScreen, SNDS flagging, ATP). You'd never see that in your ESP dashboard, which aggregates everything.
Tells you when to stop optimising
Founders waste hours tweaking subject lines and tracking setup when their placement is fine. The weekly test lets you say "placement is 80%, the problem is copy." Or "placement is 40%, copy doesn't matter yet." Both answers are time-savers.
The 3-minute Monday routine
- Open Inbox Check. (30 seconds)
- Paste in the current week's primary campaign. If you're running multiple, pick the one with the highest volume. (30 seconds)
- Hit send. Wait 2 minutes for results.
- Record in a spreadsheet: Gmail %, Outlook %, Yahoo %, date. (30 seconds)
- Glance at the trend. Is it stable? Up? Down? (30 seconds)
That's the entire routine. No DNS work. No theory. No jargon. At the end of four weeks, you have a chart. At the end of twelve weeks, you have pattern-recognition.
What to do with the data
Three rules for interpreting the weekly number:
- Stable 70%+: do nothing. You're healthy.
- Dropped 10+ points in a single week: something changed. Audit what changed (ESP, volume, content, DNS, payment). Roll back if possible.
- Slow drift downward over 4+ weeks: reputation decay. Reduce volume, warm the domain, clean the list.
For each scenario, there's a specific playbook. Without the weekly measurement, you don't know which scenario you're in, and the playbooks aren't actionable.
What this does NOT replace
To be clear: weekly seed testing is the one highest-leverage thing. It's not the only thing.
- You still need authentication set up. Most ESPs do this by default now.
- You still need to ramp new domain volume slowly.
- You still need list hygiene.
But those are setup tasks that happen once or at clear trigger points. The seed test is the ongoing discipline. If you do the setup tasks once and then the seed test weekly, you've covered the 90% case.
Why founders skip this
Three common objections:
"My ESP already shows me a delivery rate."
ESP delivery rate measures SMTP acceptance, not inbox placement. The gap between the two averages 40 points. ESP dashboards report 95% delivered on campaigns where real inbox placement is 55%. The seed test is the only way to know the real number.
"I don't have time."
Three minutes a week is 2.6 hours a year. Most founders spend more than that debugging a single broken campaign.
"My volume is too low to matter."
Low-volume senders are more affected by placement issues, not less. When you only send 500 emails a week, losing half to spam folder cuts your effective reach in half — which is the difference between "our outbound is working" and "our outbound is dead."
If you do this one thing for one quarter
At the end of 12 weeks of disciplined weekly testing, you'll know:
- Your true baseline placement %.
- Whether the "macro" slump you're hearing about is affecting you.
- Whether your copy iterations moved placement, or just reply rate on already-inboxed mail.
- Which provider is your weakest link.
- When it's time to ramp volume, and when it isn't.
That's more deliverability insight than most founders accumulate in their entire career. From three minutes a week.