Deliverability doesn't break. It decays. Your sender reputation rots one unengaged send at a time until Gmail quietly routes you to spam.
You can lose half your email revenue without a single error message. No bounce, no warning, no broken send. Your campaigns just quietly start landing in spam, opens slide, and by the time you notice the revenue drop, your sender reputation has already taken months of damage. That's the trap with deliverability: it fails silently.
Deliverability doesn't break. It decays. Sender reputation is a moving score, and it rots one unengaged send at a time: you keep emailing people who stopped opening, your complaint rate creeps up, an authentication record drifts out of date, and Gmail slowly decides you're not worth the inbox. Nothing snaps. It just degrades, which is exactly why it needs an SOP instead of a one-time setup.
This is the deliverability SOP we use with DTC brands on Klaviyo: how sender reputation actually works, the authentication that's now mandatory, the list hygiene and engagement segmentation that keep your reputation high, and the monitoring routine that catches decay before it costs you a campaign. Most of it applies to any sending platform; the specifics are Klaviyo.
Deliverability is a reputation you maintain, not a setting you configure. Inbox providers score you on how engaged your recipients are and how clean your list is. Keep sending to people who want your email and your reputation stays high. Keep blasting your whole list and it decays.
Every list degrades on its own. People change jobs and abandon addresses, interest fades, a subscriber who loved you in January ignores you by June. Do nothing and the share of your list that's dead weight grows every month. When you keep emailing that dead weight, inbox providers read it as a signal: this sender doesn't know or care who wants their mail.
Three forces pull deliverability down at once. Your list ages and engagement drops. Inbox providers raise the bar, with Gmail and Yahoo tightening sender requirements two years running and far less tolerant of partial setups than they used to be. And your own setup drifts: a new sending subdomain nobody authenticated, a DMARC record left on monitoring and forgotten, a send cadence that crept up. Any one of these erodes the score.
The fix isn't a heroic cleanup once a year. It's a routine that runs often enough to keep decay from compounding. That's what the rest of this SOP is.
You can't manage a score you don't understand. Inbox providers don't publish a formula, but they weigh a consistent set of signals, and every one of them is something your SOP can control:
Notice what these have in common: they all degrade quietly. None of them throws an error. The SOP exists to make the invisible visible, on a schedule.
Authentication is the one part of deliverability that's pass-or-fail, and in 2026 it's mandatory. If you send marketing email at any real volume, you need all three records in place and valid:
The trap here is that authentication looks like a one-time task and isn't. A new tool that sends on your behalf, a new subdomain, a DNS change during a site migration, any of these can silently break alignment. Verify authentication is still valid whenever you add a sending tool or touch DNS, and confirm it in your monitoring on a schedule.
Set up a branded sending domain and confirm SPF, DKIM, and DMARC are aligned.
A clean list is the cheapest deliverability win there is, because emailing people who never open is pure downside: no revenue, and a steady drag on your reputation. Hygiene is the routine that keeps the dead weight off your sends.
The instinct to email the whole list because they're on it is what kills deliverability. A smaller engaged list outperforms a big stale one on both revenue and reputation.
Your sunset flow is itself a flow that drifts; audit it alongside the rest.
Hygiene removes the dead. Engagement segmentation decides how you treat the living, and it's the lever that protects reputation while still reaching people. Group your list by how recently they engaged and send accordingly:
The mistake is sending every campaign to everyone because the list is big and the campaign feels important. Every send to a dormant segment is a small withdrawal from your reputation. Exclude them as a standing rule, not a one-off, and revisit the segments as engagement shifts.
Deliverability decays silently, so the only way to catch it early is to watch the leading indicators. These numbers move before your revenue does:
Put these on a dashboard and look at them on a fixed cadence. The point of monitoring isn't to admire the numbers; it's to catch the complaint-rate creep or the engagement slide while it's still a tweak, not a recovery project.
Deliverability falls through the cracks because it sits between marketing and ops and nobody clearly holds it. Name an owner and a cadence:
Deliverability rewards a monthly rhythm more than a quarterly one, because reputation moves faster than most operational docs. A complaint-rate problem caught this month is a settings tweak; caught next quarter, it's a reputation you have to rebuild.
The same owner-and-cadence discipline, applied to the legal side of your other big channel.
A deliverability SOP drifts on both sides. The providers keep moving the goalposts: Gmail and Yahoo have tightened requirements two years running and will again, so a setup that passed last year can quietly fall short. And your own program changes: a new ESP integration, a new subdomain, a higher send frequency, a campaign type that draws complaints.
Review the SOP every quarter and immediately after any change to your sending setup or any new provider requirement. This is the same documentation drift that breaks every operational doc, and on deliverability it shows up as the slow slide to spam that nobody notices until the revenue dips.
Why every operational doc, including this one, degrades within 90 days unless you catch it.
Don't try to fix everything at once. Do the two checks with the highest payoff first. Confirm SPF, DKIM, and DMARC are valid and aligned, because a broken record undermines everything else you do. Then pull your spam complaint rate and your engagement by segment, so you know whether you're healthy or already sliding.
If you're sending to everyone, the next move is the biggest one: build the engaged and dormant segments and start excluding dormant from regular campaigns. That single change does more for most brands' deliverability than any amount of subject-line tuning.
ReccordSOP turns a routine like this into a documented SOP with timestamped screenshots, and flags drift when your tools, domains, or provider rules change underneath it. Generate your first SOP free at reccordsop.com.
Usually decaying sender reputation, not a single error. The common causes are sending to unengaged subscribers, a spam complaint rate creeping above 0.3 percent, broken or missing authentication (SPF, DKIM, DMARC), or a list with too many dead addresses. Deliverability degrades quietly, so the fix is a maintenance routine, not a one-time change.
Stay under 0.3 percent to keep good standing with Gmail and Yahoo, and the strongest senders stay below 0.1 percent. A rate above 0.5 percent causes immediate problems, including spam-folder placement and possible blocking. Watch it as a leading indicator, because it moves before your revenue does.
If you send at any real volume, yes. DMARC is now required for bulk senders (around 5,000 messages a day to a provider), alongside SPF and DKIM. Even a monitoring policy has to be valid and aligned, and providers are pushing toward stricter quarantine or reject policies. Send from a branded domain so the reputation you build is your own.
A sunset flow automatically targets subscribers who haven't engaged in a set window, usually 90 to 180 days, with a short re-engagement series, then suppresses the ones who still don't respond. Yes, you need one: emailing people who never open drags down the inbox placement of everyone who does.
Monthly for the core metrics (complaint rate, engagement trend, sunset flow running), and immediately after any change to your sending setup like a new tool, subdomain, or DNS edit. Reputation moves faster than most processes, so a monthly rhythm catches decay while it's still a small fix.
I built ReccordSOP after watching too many DTC ops teams lose months to undocumented workflows. These SOPs are battle-tested with Shopify operators running $1M to $50M brands.
Last reviewed June 15, 2026
Most brands set up Klaviyo flows once and never revisit them. Here's the audit framework to keep them earning.
A mistimed text can cost $500 to $1,500. Here's the compliance SOP that turns 'we think we're fine' into 'we can prove it.'
Most SOPs are wrong within 90 days of publishing. Here's how to detect it before it costs you a customer.
We use essential cookies for sign-in and a small amount of analytics to improve the product. Privacy policy.