Email is the only audience you own outright. No algorithm sits between you and it. Most businesses treat it as a monthly obligation and wonder why it underperforms.
Key takeaways
- A few well-timed automated emails routinely outperform every campaign a small business sends manually.
- List size is a vanity metric. An engaged list of 800 beats a stale list of 8,000, and costs less to send to.
- Deliverability is decided by authentication and engagement, not by subject-line wording.
- Removing people who never open protects the inbox placement of everyone who does.
On this page
Automations earn more than campaigns
Emails triggered by something the recipient just did outperform scheduled broadcasts by a wide margin, because the timing is right by construction. They also run without anyone remembering to write them.
For most small businesses, four automations cover nearly all of the available return. Build them once, then improve them slowly.
| Automation | Trigger | Job |
|---|---|---|
| Welcome sequence | Someone subscribes | Set expectations, establish credibility |
| Abandoned cart or form | Started, did not finish | Recover intent that already existed |
| Post-purchase | Order completed | Reduce anxiety, prompt review, sell again |
| Re-engagement | No opens in 90 days | Win back or remove cleanly |
The welcome sequence is the highest-leverage of the four and the most commonly skipped. Someone who just subscribed is more interested in you at that moment than they will be at any point in the next year.
Why list size is the wrong goal
Mailbox providers judge you by how the people you send to respond, so uninterested subscribers actively damage delivery to interested ones. Every unopened send is evidence against you.
This inverts the usual instinct. Removing two thousand dormant addresses feels like destroying an asset; in practice it usually raises open rates, improves inbox placement and lowers your bill at the same time.
- Never buy a list. Beyond the legal exposure, purchased addresses generate spam complaints that damage your sending reputation for months.
- Use confirmed opt-in where you can. Fewer subscribers, dramatically better ones.
- Say plainly what people are signing up to and how often. Mismatched expectations produce complaints, not just unsubscribes.
- Remove people who have not opened anything in six months, after one honest attempt to win them back.
- Make unsubscribing easy. Someone who cannot find the link will use the spam button instead, which is far more damaging.
Deliverability is mostly a technical problem
If your email lands in spam, the cause is almost always authentication or engagement, not the words you chose. Rewriting subject lines to avoid imagined trigger words is folklore.
- Set up SPF, DKIM and DMARC on your sending domain. Major providers now effectively require all three for bulk sending, and misconfiguration is the single most common cause of unexplained spam placement.
- Send from your own domain, never from a free mailbox address. Sending as an address you do not control fails authentication outright.
- Warm up gradually on a new domain or platform. Ten thousand emails on day one looks exactly like a compromised account.
- Keep complaints low. Roughly one complaint per thousand sends is the threshold where providers start acting.
- Include a plain-text version and keep the HTML simple. Image-only emails are a long-standing spam signal.
What to actually send
Send things a subscriber would be mildly annoyed to have missed. That is a higher bar than “we published a blog post” and it is the only one that keeps a list healthy over years.
Useful categories: something you learned that changes what a reader should do, an honest account of something that did not work, early access or a genuine offer, and answers to questions your customers keep asking. Company news is almost never one of these.
Write to one person. The best-performing emails from small businesses read like a note from someone who knows what they are talking about, not like a designed template with four calls to action.
If you would not send it to a single customer you respect, do not send it to nine hundred.
Serhii Yelbaiev, Well Web Marketing
What to measure now that opens are unreliable
Open rates have been inflated and distorted by privacy features that pre-fetch images, so treat them as a rough trend and nothing more. Clicks, replies and revenue per subscriber survived intact.
Watch click-to-delivered rather than click-to-open, since the denominator is trustworthy. Track unsubscribes and complaints as a health signal. And measure revenue per subscriber per month, the one number that tells you whether the list is an asset or a habit.
Frequently asked questions
Consistently, at a frequency you set expectations for at signup. Monthly done well beats weekly done carelessly, and irregular sending is worse than either.
Only as a trend against your own history. Image pre-fetching by privacy features inflates them unpredictably, so never compare your rate to a published benchmark.
It is not required everywhere, but it consistently produces a healthier list and better deliverability. The subscribers you lose to it were rarely going to engage.
Check authentication first: SPF, DKIM and DMARC on the sending domain. That accounts for most cases. After that, look at engagement history and complaint rate.
Sources and method
Authentication requirements reflect the bulk sender rules published by major mailbox providers. Performance observations come from our own client email programmes. Last reviewed February 2026.