Email Deliverability for Sales Teams: Why Your Emails Land in Spam
- No invented stats
- Written for UK readers
- 6 min read
- Published 19 September 2026
You can write a perfect cold email and never find out, because it went to spam. Deliverability is the least glamorous part of outbound and the one that silently invalidates everything else. The good news is that most of it is a setup problem you solve once, plus a small set of habits you maintain.
The three things every sales domain needs
These are DNS records. Someone technical can add them in under an hour; if nobody has, your email is being treated with suspicion before anyone reads it.
SPF
A record listing which servers are allowed to send email for your domain. If you send through Microsoft 365 and an outreach tool, both need to be in it. A common mistake is having more than one SPF record, or exceeding the lookup limit by stacking every tool the company has ever trialled — both cause failures.
DKIM
A cryptographic signature proving the message wasn't altered and genuinely came from your domain. Every sending platform gives you a key to publish. If you use two platforms, publish both.
DMARC
A policy telling receiving servers what to do when SPF or DKIM fails, and where to send reports. Start at p=none so you can see what's happening without blocking legitimate mail, read the reports for a few weeks, then move to p=quarantine and eventually p=reject once you're confident everything legitimate passes.
Major mailbox providers have tightened enforcement of these in recent years, particularly for bulk senders. Treat all three as mandatory rather than advanced.
Don't send cold email from your main domain
This is the single most valuable structural decision. If your company operates on yourfirm.co.uk, buy a close variant — yourfirm-mail.co.uk or getyourfirm.co.uk — and run outbound from there.
The reason: reputation damage from cold outreach is cumulative and hard to reverse. If your primary domain gets flagged, your invoices, support replies and contract emails start landing in spam. That's a business problem, not a sales problem. Keeping the two separate means a bad outbound month costs you an outbound domain, not your company's email.
Set up the new domain with a real website (even a single page), the full SPF/DKIM/DMARC set, and matching branding. A sending domain with no website attached looks exactly like what spammers use.
Warm up before you send volume
A brand-new domain that sends 400 cold emails on day one will be filtered. Mailbox providers assess sender reputation partly on history, and a domain with no history sending at volume is the classic spam pattern.
A sensible ramp over four to six weeks (our suggested pacing, not a published rule from any mailbox provider):
- Week 1: a handful of emails a day, mostly to people who will reply — colleagues, friendly customers, your own other accounts.
- Week 2–3: build to 20–30 a day, mixing real outreach with conversational mail.
- Week 4–6: increase gradually towards your target, watching bounce and reply rates.
Then hold. As a rule of thumb rather than a published limit, keeping each mailbox to something like 30–50 cold emails a day is a conservative place to sit. If you need more volume, add mailboxes rather than pushing one harder. Automated "warm-up networks" that exchange fake replies between accounts exist; providers have got better at detecting them, and they're a shortcut with real downside risk.
List quality is a deliverability issue
Bounces are one of the strongest negative signals there is. If you're sending to guessed addresses from a scraped list, you will bounce heavily and your reputation will fall, regardless of how good your setup is.
- Verify every address before sending. Aim to keep hard bounces to a very low fraction of a percent.
- Avoid catch-all domains where you can't confirm the individual mailbox exists.
- Purge old data. A two-year-old list is full of people who've left, and abandoned mailboxes sometimes become spam traps.
- Respect opt-outs across every tool. Complaints hurt far more than bounces.
There's a legal dimension here too, not just a technical one — see GDPR for sales teams for what you can and can't do with UK B2B contact data, including the requirement to offer a working way to opt out.
Content habits that trigger filters
Filters look at far more than keywords, but some patterns reliably hurt:
| Habit | Why it hurts | Do instead |
|---|---|---|
| Links in the first email | Unknown sender plus a link is the spam signature | Save links for a reply, or describe the thing in text |
| Attachments | Heavily scrutinised from unknown senders | Offer to send on request |
| Tracking pixels | Adds an invisible image and a third-party domain to every message | Judge by replies, not opens |
| HTML templates, logos, signatures with images | Marketing-shaped mail gets marketing-shaped treatment | Plain text, plain signature |
| Identical body text at volume | Bulk fingerprinting | Genuine per-recipient variation in the first lines |
| Long messages | Correlate with low engagement, which feeds back into reputation | Under 100 words on a first touch |
Open tracking deserves a specific mention. Because it requires loading a remote image, it makes your email fractionally more suspicious, and modern privacy features make the data close to meaningless anyway. Most teams would be better off switching it off and measuring replies.
Subject lines matter for filtering as well as for opens — avoid all-caps, excessive punctuation, currency symbols in the subject and words like "free" or "guaranteed". Our cold email subject line checker flags the common offenders.
Engagement is the thing filters actually measure
Modern filtering is largely behavioural. Replies, moves out of spam and messages marked as important are positive; deletions without reading, marks as spam and total silence are negative. This has a strategic implication that most teams miss: sending to people unlikely to reply damages your ability to reach people who would.
Practically, that means a tighter list beats a bigger one, first emails should ask a question that's easy to answer, and you should cut contacts who have ignored a full sequence rather than adding them to the next campaign. It also means that improving your follow-up sequence so it earns replies is a deliverability improvement, not just a conversion one.
How to check where you stand
- Send a test to accounts on the major providers — Outlook/Microsoft 365, Gmail, and a UK ISP address — and note where each lands.
- Check your DMARC reports for authentication failures you didn't expect.
- Check whether your sending domain or IP appears on the well-known public blocklists.
- Watch your bounce rate weekly. A sudden rise usually means a bad data import.
- Watch your reply rate by domain. If one mailbox's replies fall off a cliff while others hold, that mailbox is the problem.
If you're already in trouble
Stop sending from the affected mailbox for a week or two. Fix authentication, purge the list, then restart at very low volume with messages likely to get replies. Recovery is slower than damage, typically weeks rather than days. If the affected domain is your primary business domain, prioritise the fix — that's an operational risk, not a marketing one.
The honest limitation
Nobody outside the mailbox providers knows exactly how their filtering works, and it changes. Anyone quoting precise thresholds — a specific number of emails per day that is always safe, or an exact spam-score cutoff — is guessing. What's reliable is the direction of travel: authenticate properly, send from a separate domain, keep volumes modest, send to people who might genuinely reply, and write like a person rather than a marketing department.
Spotted a mistake or a number we can't back up? Tell us and we'll fix it.