Email Marketing Best Practices for the Inbox Era
By Brian Shelton — Founder of GrowPredictably.com
TL;DR: Email marketing best practices in 2026 start with reaching the inbox. Authenticate your domain with SPF, DKIM, and DMARC, send to the people who actually engage, keep your list clean and permission based, segment for relevance, and use AI as a co-thinker instead of a copy machine. The inbox now scores your sender reputation before it ever reads your copy.
Key Takeaways
- Since February 2024, bulk senders to Gmail must authenticate with SPF, DKIM, and DMARC and keep spam complaints below 0.30%, which makes deliverability the first best practice, ahead of copy.
- Engagement based sending protects your sender reputation because mailing only the people who open and click keeps complaint rates under the threshold that decides inbox placement.
- List hygiene, double opt-in, and one-click unsubscribe reduce spam complaints, and complaints are the number Gmail actually scores.
- Email still drives an average return of $36 for every dollar spent, higher than any other channel, but only for senders who reach the inbox.
- AI belongs in research, segmentation logic, and first drafts. A human point of view belongs in every send, because generic AI sameness suppresses the engagement your reputation depends on.
Marketing teams that watched open rates slide through 2024 and 2025 usually went hunting in the wrong place. They rewrote subject lines, redesigned templates, and A/B tested send times while the real problem sat in their DNS records and their unengaged list. The visible symptom is a slow bleed in opens.
The upstream cause is a sender reputation the inbox has been quietly scoring the whole time.
This guide is for the CMO, VP of Marketing, or head of demand generation whose email numbers are drifting down while the team swears the copy got better. It covers the reputation before copy order, from authentication through engagement based sending to AI-assisted writing, and it treats email as one lane of a larger system of AI powered marketing automation.
The work that follows is what diagnose-first looks like for email. Fix the layer the inbox scores before you polish the layer the reader sees.
Why did email best practices change in 2024?
Email best practices changed because Google and Yahoo changed the rules. In February 2024, both providers began requiring bulk senders to authenticate their email, offer one-click unsubscribe, and stay under a strict spam complaint threshold.
The practices that matter now are the ones the inbox scores, and Gmail grades your sender reputation before it ever reads your message.
The scale of the filtering explains why. Google reports that Gmail’s AI powered defenses stop more than 99.9% of spam, phishing, and malware and block nearly 15 billion unwanted emails every day. Your marketing email passes through that machinery on every send. The machinery does not read your subject line first.
It reads your authentication records, your complaint rate, and your history with the recipients you are mailing.
Yahoo enforces matching requirements on the same timeline, including SPF and DKIM, a DMARC policy, one-click unsubscribe, and a spam rate kept below 0.3%, which means the two providers that handle most B2B recipients now hold the same floor. The rules formally bind bulk senders, defined as 5,000 or more messages a day to Gmail, but the underlying signals predict placement for everyone.
A 2,000 person list gets scored by the same machinery, so the practices apply regardless of your volume.
That reorders the classic advice list. Good subject lines, clean design, and mobile friendly templates still matter, but they sit on top of a deliverability layer that decides whether anyone sees them at all. A brilliant campaign sent from an unauthenticated domain to a stale list is invisible, and no amount of copywriting recovers it.
I treat a sending domain the way a lender treats a credit score, and I push every marketing leader I work with toward the same frame. Every campaign either builds the score or spends it. Complaints, bounces, and ignored sends draw it down, while engaged opens, clicks, and replies build it back.
The rest of this guide follows that logic in order, starting with the part most teams have never audited.
How do you authenticate your domain with SPF, DKIM, and DMARC?

You authenticate your domain by publishing three DNS records that prove your email really comes from you. SPF lists the servers allowed to send for your domain, DKIM adds a tamper proof signature to each message, and DMARC tells receiving servers what to do when a message fails the first two checks.
For bulk senders, Google’s sender guidelines make all three mandatory.
What each record actually does
SPF is the guest list. It names the sending services, like your ESP and your CRM, that may send mail for your domain. DKIM is the signature. It lets Gmail verify the message was not altered in transit and genuinely came from your infrastructure.
DMARC is the enforcement policy. It tells receivers to quarantine or reject mail that fails, and it sends you reports on who is sending as you. Your ESP publishes the first two in an afternoon, and a starter DMARC policy can begin in monitoring mode.
The most common failure is silent. A team migrates to a new ESP, the DKIM key never gets updated, and authentication starts failing without an error anyone sees. Opens drop for weeks before someone thinks to check. The recovery is deterministic.
Publish the missing record, confirm alignment in Google Postmaster Tools, and let your reputation rebuild through consistent, wanted sends.
Staying under the spam rate line
Authentication gets you judged on your record instead of rejected at the door. The record that matters most is the spam complaint rate. Senders of 5,000 or more messages a day to Gmail must keep the rate reported in Postmaster Tools below 0.30%, and Google advises staying under 0.10% and never touching the ceiling.
As Neil Kumaran, Group Product Manager for Gmail Security and Trust at Google, framed the change when announcing the requirements:
These changes are like a tune-up for the email world, and by fixing a few things under the hood, we can keep email running smoothly. But just like a tune-up, this is not a one-time exercise.”
Neil Kumaran, Group Product Manager, Gmail Security & Trust, Google
Treat it the same way. Before your next campaign, confirm five things:
- SPF record published and passing for your sending domain.
- DKIM signing enabled in your ESP with a current key.
- DMARC policy published, at minimum in monitoring mode with reports on.
- Postmaster Tools connected, with the spam rate trending under 0.10%.
- One-click unsubscribe live on every marketing send.
Why should you send by engagement, not by the calendar?
Engagement based sending means your recipient list is defined by behavior, and it is the highest impact deliverability practice available to you. Recent openers and clickers get your regular cadence, while long-silent subscribers get a wind down path instead.
This one change lowers complaints and raises the signals Gmail weighs, because the riskiest sends never happen.
Why unengaged sends cost you the inbox
Every send to someone who no longer wants your email is a small withdrawal from your reputation. Some of those recipients mark you as spam rather than unsubscribe, and each complaint pushes you toward the 0.30% ceiling that decides placement for your whole list.
The calendar driven program, where everyone gets everything every week, maximizes exactly the sends most likely to generate complaints. Your most disengaged subscribers become your biggest deliverability risk.
A segmentation campaign that lets people self-select
The move I recommend to every marketing leader sitting on a cold or aging list is a segmentation campaign. Ask for one tiny action, because the act of replying is itself the segmentation. A working example you can adapt:
“We just rebuilt our email deliverability checklist for 2026. If you want it, reply with the word checklist and I will send it over.”
The people who reply have re-engaged, and replies are the strongest positive signal an inbox can record. The people who do not reply, across a few attempts, have told you something too. Move them to a lower cadence or a goodbye sequence, because continuing to blast them is how the complaint spiral starts.
Once complaints rise, placement drops for your engaged readers as well, so opens fall further and the temptation is to send even more. The recovery runs the other way. Pull back to the people who still open, rebuild the engagement trend, and then re-expand slowly.
You will mail fewer people and reach more of them, which is the trade that wins in the deliverability era.

Keep your list clean, permission-based, and easy to leave
A clean list is one where every address is valid, every subscriber opted in, and every exit is easy. That combination keeps bounces and complaints low, which protects deliverability for everyone who stays.
It also happens to be what Google’s 2024 requirements now enforce, including one-click unsubscribe processed within two days.
List hygiene and double opt-in
Hygiene is a standing job, and a quarterly pass works for most B2B programs. Remove hard bounces immediately, suppress addresses that have been invalid or inactive for a long stretch, and validate new addresses at the point of capture so typos never enter the list in the first place.
A validation step at the form catches the misspelled domains and disposable addresses that would otherwise become bounces, which means your future hygiene passes get smaller over time.
Grow the list only through real permission. Double opt-in costs you a little volume and pays it back in quality, because everyone on the list confirmed they want to be there, and people rarely hit the spam button on mail they asked for twice. Never buy or scrape a list.
Purchased addresses bounce, complain, and sit on spam traps, and one bad import can damage a reputation you spent a year building. Grow instead through lead magnets built for the AI era that trade real value for real permission.
One-click unsubscribe is a feature, not a leak
Marketing leaders sometimes resist easy exits because the list is a reported number. That instinct backfires. A subscriber who cannot leave in one click leaves by hitting the spam button instead, and the spam button is the one that costs you. An unsubscribe is neutral. A complaint is a scored strike.
Honor the exit fast, and your deliverability improves for the audience that remains. Google’s announcement names a two day processing window for those requests, so wire the suppression into your ESP rather than leaving it to a weekly manual export.
Segment so every email is relevant
Segmentation is how you earn engagement at scale. Group subscribers by what they did and where they are in the buying journey, then match the message to the group. Relevant email gets opened, and opens feed the reputation loop that keeps you in the inbox, so segmentation and deliverability reinforce each other.
Start with behavior, since it is the strongest signal you own, and resist the urge to build a dozen segments on day one. Over-engineered segmentation collapses under its own maintenance, so begin where the signal is loudest. Three segments cover most B2B programs at the start:
- New subscribers, who get a welcome sequence that sets expectations and delivers the thing they signed up for.
- Active engagers, who get your regular cadence and your best offers.
- Lapsing subscribers, who get the wind down path from the engagement section above.
Layer intent on top as your data matures. A subscriber who visited your pricing page behaves differently from one who read a single blog post, so your email should know the difference and speak to it. The pricing visitor is ready for a case study or a comparison, while the blog reader needs another useful idea before any offer makes sense.
Tools that enrich and segment your list can add firmographic depth to those behavioral segments, but behavior comes first because it never lies.
Keep the classic message types in their lanes as you go. Transactional email confirms actions, relational email delivers value between offers, and promotional email makes the offer. The ratio matters, because a program that only promotes trains its list to stop opening, and the engagement decline shows up in the numbers Gmail scores.
When every send earns its open, the relevance compounds into reputation.
Write for a human and use AI as a co-thinker
AI earns its place in your email program upstream of the send button. Use it for research, segmentation logic, subject line variants, and first drafts. Keep a human owner on the point of view, the offer, and the final pass, because the generic sameness of unedited AI copy suppresses the engagement your deliverability now depends on.
I have watched marketing leaders hire and fire digital marketing agencies for fifteen years, from the agency side and from the in-house side. The breaking point has shifted. Marketing leaders increasingly describe the same complaint about the work they pay for. It reads like it could have been written about any company in the category.
When your email sounds like everyone else’s email, the reader stops opening, and the reputation loop turns against you.
The fix is a division of labor, the approach behind treating AI as a co-thinker rather than a vending machine for copy. Three guardrails keep the quality bar in place:
- No unedited AI sends. A human reads every email aloud before it ships, and rewrites anything they would not say to a customer’s face.
- One real idea per email. Ground each send in a specific observation, result, or decision your reader cares about, then let AI help you say it faster.
- Cut what could be anyone’s. If a paragraph would work in a competitor’s newsletter without edits, it is not carrying your point of view. Delete it.
Run your program this way and AI becomes an accelerant for a voice you already own, instead of a dilution of it. Your reader can tell the difference, and so can your open rate, because engagement follows distinctiveness and the reputation loop follows engagement.
The teams that win this era are the ones that let AI raise their volume without letting it flatten their point of view.
Which metrics actually predict deliverability?
The metrics that predict deliverability are spam complaint rate, bounce rate, and your engagement trend. These three move before revenue does. An email program can look healthy on opens this month while complaints creep toward the threshold that gets next month’s campaign filtered.
Watch the leading indicators, and read your trend against your own history rather than someone else’s benchmark.
The first thing I check on any underperforming email program is the spam rate trend in Postmaster Tools, because it moves earliest and explains the most. Your early warning dashboard needs four numbers:
- Spam complaint rate in Google Postmaster Tools, tracked against the 0.10% guideline and the 0.30% ceiling.
- Hard bounce rate, which reveals list decay before engagement shows it.
- Engagement trend, meaning opens, clicks, and replies over time, where the direction matters more than the level.
- Unsubscribes relative to complaints, because a healthy program loses people through the front door, never the spam button.
Industry open rate benchmarks make poor targets because they blend every sector, list size, and program maturity into one number. A declining trend on your own list is the signal that predicts trouble, and it usually traces back to one of the practices above.
Falling opens with rising bounces points to hygiene. Falling opens with rising complaints points to cadence or relevance. The dashboard tells you which chapter to reread.
The numbers also tell you when to act. Once your Postmaster spam rate crosses the 0.10% guideline, treat it as a fire alarm rather than a data point, because the distance to the 0.30% ceiling closes fast on a large list. Pause the broad sends, retreat to your most engaged segment, and fix the practice the diagnostic pairs above point to.
Then re-expand gradually as the rate falls. Catching the slide early costs a short stretch of reduced volume, while letting it reach the ceiling typically costs far longer in rebuilt reputation.
How should you sequence these practices?
Order matters more than effort here. Authentication opens the door, engagement based sending keeps it open, hygiene and permission keep complaints low, segmentation earns the opens that feed the loop, and AI assisted craft multiplies whatever the layers beneath it produce.
A team that runs the sequence backwards, polishing copy on top of a broken reputation layer, works hard and stays invisible.
The sequence is also a diagnosis you can run this week, and my rule for running it is bottom up. Start at the base of the stack and check each layer before touching the one above it. If authentication fails, nothing else you fix will register.
If authentication passes but engagement is falling, the list needs the wind down treatment before the copy needs a rewrite. Most teams discover their real constraint sits one or two layers below where they were working, which is why the effort felt wasted.
The payoff justifies the discipline. Litmus pegs email’s average return at $36 for every dollar spent, higher than any other channel, and that return accrues to the senders who actually reach the inbox. Make each practice a standing system rather than a one-time fix, so the gains survive busy quarters.
Authentication gets a quarterly check, hygiene gets a quarterly pass, and your welcome and re-engagement flows run on automated lead follow-up so consistency stops depending on memory.

If your email numbers have been drifting and you now suspect the cause sits deeper than copy, test that suspicion against your whole growth engine, because email is usually one constraint among several candidates. A diagnostic beats another quarter of guessing, and it tells you whether email is even the right place to spend the next ninety days.
Find your growth gap and see which stage is capping growth, email included.
Frequently Asked Questions
About the author

Brian helps B2B founders install marketing + automation engines powered by Co-Thinking with AI. With 15+ years building predictable revenue systems, he's worked with SaaS, agency, and service businesses on 90-day done-with-you growth accelerators.
Ready to install your predictable revenue engine?
Book a free strategic growth session. Walk away with a tailored 90-day blueprint and 3 quick wins you can use this week.