My Framework for Email Deliverability and B2B LinkedIn Outreach

A message landed in a WhatsApp group a few days ago. Someone wanted ideas for using email marketing and LinkedIn to bring in new clients. It’s a fair question, and one most B2B founders ask themselves at some point, usually right after a quarter that didn’t go the way they hoped. The honest answer takes longer than a WhatsApp reply, so I recorded a video about it instead, and this article is the longer, written version of that answer.

Two things decide whether email marketing and LinkedIn outreach bring you new clients, or just quietly damage your reputation. The first is email deliverability, whether your messages reach an inbox at all. The second is how you approach people once you can reach them. Get the second right while ignoring the first, and you end up with a beautifully written email that never left the spam folder. Get the first right and skip the second and you have a technically flawless email nobody asked for.

This piece covers both, in the order they actually matter, and goes further into a few things the video didn’t have time for.

Why deliverability decides everything else

No subject line, offer or LinkedIn connection request matters if the message never reaches a human being. Even well-run programmes lose a meaningful slice of their list to spam folders and quiet filtering. The global average inbox placement rate across 2025 sat at 87.2% [Validity, 2026 benchmark report, vendor data, worth reading with that in mind]. That means roughly one message in eight from an average sender never reaches a primary inbox, whatever the copy inside it says.

Mailbox providers decide that placement, and they’ve got stricter about it. Since February 2024, senders who send more than 5,000 messages a day to personal Gmail addresses have had to authenticate outgoing mail with SPF and DKIM, publish a DMARC record, and keep spam complaints below a set threshold [Google’s own sender guidelines]. Google confirmed in November 2025 that it had started tightening enforcement, and non-compliant traffic now faces temporary and permanent rejection rather than a warning. Outlook brought in similar rules for the same volume of senders in 2025. None of the big mailbox providers treat authentication as optional anymore.

Authentication used to be a nice-to-have for careful senders. Now it decides whether Gmail, Yahoo and Outlook let your email through the door in the first place.

The three records that decide whether you’re trusted

Three technical records do the actual work, and each answers a different question a receiving mailbox provider asks about your email. SPF, Sender Policy Framework, answers “is this server allowed to send on behalf of this domain?” You publish a list of the IP addresses and platforms permitted to send as your domain, and any message from an unlisted source fails the check. It’s the simplest of the three records, and the one most people set up first.

DKIM, DomainKeys Identified Mail, answers “has this message been tampered with, and does it really come from where it claims?” Your sending platform attaches a cryptographic signature to each message, generated from a private key only it holds, and the receiving server checks that signature against a public key published in your DNS. A mismatch means the message was altered in transit or isn’t genuinely from you.

DMARC, Domain-based Message Authentication, Reporting and Conformance, is the policy layer tying the other two together. It tells receiving providers what to do when a message fails SPF or DKIM, whether to deliver it anyway, quarantine it, or reject it outright, and it requires the domain used in SPF or DKIM to align with the domain the recipient actually sees in the From field. A DMARC record set to a policy of “none” only monitors and reports; moving to “quarantine” or “reject” is what actually protects your domain from being spoofed. BIMI works as a reward for having the other three fully and correctly in place, a small brand logo displayed next to verified messages in some inboxes, unlocked only once DMARC is enforced at quarantine or reject rather than just monitoring.

Reputation is slow to build and fast to lose

Authentication proves you’re who you say you are. Reputation is a separate, ongoing score mailbox providers keep on your domain and sending IP, built from how recipients actually behave when your email arrives: whether they open it, click it, reply, mark it as spam, or ignore it entirely. New domains and new IPs start with no reputation at all and need a gradual warm-up period of consistent, well-received sending before providers trust them with any real volume.

This is exactly why the automated cold outreach tools spinning up fresh subdomains for every campaign tend to underperform, whatever they promise. The logic behind buying a new domain for every push is to protect the main domain if something goes wrong, but every new domain starts back at zero reputation, and mailbox providers now use their own detection systems to spot a business rotating through disposable domains at scale. What looks like protection from the sender’s side often reads as a coordinated spam pattern from the receiving side. A domain and IP with a genuine sending history, built up over months rather than swapped every few weeks, is worth more than any amount of clever domain rotation.

List hygiene is the boring part that actually moves the needle

Clean lists outperform large ones. A double opt-in process, where a new subscriber confirms their email address before joining your list, produces fewer sign-ups but a meaningfully more engaged and deliverable list than a single-click form. Removing hard bounces immediately and pruning subscribers who haven’t opened or clicked in a long stretch protects your sender score, because mailbox providers read persistent disengagement as a signal your list is stale, or was never properly permissioned in the first place.

Consent matters here too, particularly if you operate under EU rules. On my own newsletter signup form, new subscribers see a separate, optional checkbox to consent to individual-level tracking inside my emails, distinct from the subscription itself. More than 70% tick it voluntarily, which tells me people are willing to trade a little visibility for a newsletter they think is worth reading, as long as the ask is explicit and the value is clear rather than buried in a privacy policy nobody opens.

AI is useful for parts of this and not for others. It can walk you through configuring a DMARC record or explain what a bounce code means, but it can’t decide which contacts on your list deserve to stay, and it has no way to verify someone actually consented to what you’re about to send them.

Building an email that survives contact with a real inbox

Asking an AI model to generate an email’s HTML from scratch, with no review from someone who understands how email rendering works, is a reliable way to produce something that looks fine in your own inbox and breaks in someone else’s. Older versions of Outlook still render using a stripped-down engine that handles modern CSS and JavaScript badly, and an email built without that constraint in mind can arrive looking nothing like it was designed to.

This is why a proper email builder, one built specifically for email rather than for the web, earns its place in the stack. I use Stripo for this reason: it’s a drag-and-drop builder that renders consistently across the major clients, supports dark mode properly so an email doesn’t turn into a glowing white rectangle when someone’s phone switches to night shift, and lets you preview exactly how a layout will look on a laptop versus a specific phone model before you send. A handful of technical details separate an email that performs from one that quietly frustrates people: a working web version link for anyone who wants to share the email outside their inbox, a visible unsubscribe link, and a clear statement of who is sending and why. Skipping any of these costs more than a professional impression. It actively works against your deliverability, because mailbox providers read the presence or absence of these signals as part of how they score a sender.

A short story about charging for a newsletter

For a while, my newsletter was a paid subscription at €76 a year, and it brought in a genuinely useful amount of extra revenue, close to €20,000 a month at its peak. It was also quietly working against me. The people paying to subscribe were mostly other email marketing professionals and competitors, rather than the founders and marketing leads who might actually hire one of my agencies or come to a workshop, because a paid newsletter about email marketing selects for people already deep in the subject, not people who need help with it. The side income was real, but it was pulling attention and content decisions away from the newsletter’s actual job, which is turning readers into clients or students. I dropped the paywall. The lesson generalises past newsletters: a lead-generation channel optimised for a secondary revenue stream will quietly underperform at its primary job, and it’s worth checking whether any of your own top-of-funnel content has drifted into that trap.

The second half of the framework, earning attention before you ask for anything

Deliverability gets your email into an inbox. What happens next is a separate skill, and it’s where most B2B outreach quietly falls apart. The broad, unsegmented approach, often nicknamed spray and pray, treats an entire market as equally interested and equally ready to buy, then measures success by volume sent rather than relevance to any individual recipient.

Account-based marketing takes the opposite approach: instead of a wide net, you build a specific, researched list of the companies you actually want as clients, then focus your attention inside each one on the people who genuinely influence the decision.

That last part matters more than most outreach plans account for. B2B purchases are rarely made by one person acting alone. Gartner’s research puts the typical buying group at somewhere between five and sixteen people spread across as many as four different functions, each arriving with their own research and their own priorities [Gartner, May 2025]. A message aimed at a single job title, even the right one, is talking to one voice in a room full of them. Mapping that buying committee, the champion who’ll advocate internally, the technical evaluator, the person who signs the budget, before you ever send an outreach message changes who you write to and what you say.

The outreach sequence that doesn’t feel like outreach

Once you know who the buying committee actually is, the sequence that works is slower than most people expect and considerably less pitchy. Start by following and genuinely engaging with what those people post, for weeks before you ever send a connection request tied to a pitch. When something relevant comes up, share something you’ve already built, an article, a specific answer to a question they raised publicly, rather than a generic introduction to your company. The goal at this stage is to become a familiar, useful name rather than a stranger with an agenda.

Cold pitches sent the moment a connection is accepted rarely work, because nothing has been earned yet. A slower approach, personal emails sent from an actual person’s inbox rather than an automation platform, referencing something specific and useful, tends to produce replies precisely because it doesn’t read like outreach. Only once that warmth is genuinely there, not immediately, but when the relationship can carry it, does pitching your services belong in the conversation. It’s a longer road to a client than a templated sequence blasted to a purchased list, but it protects the same sender reputation the first half of this piece spent so long building, and the data backs the patience: organisations running account-based marketing report an average programme ROI of 137%, based on a survey of 771 marketers [Outcomes Rocket, State of ABM 2025].

What to actually measure

Two sets of numbers matter here, and they rarely get tracked together. On the deliverability side, watch inbox placement rate rather than delivery rate, keep spam complaints under Google’s 0.3% threshold, and check your authentication status through a tool like Google Postmaster Tools rather than assuming it’s still correctly configured months after you set it up.

On the outreach side, track engagement-to-reply rate within your target accounts specifically, not connection requests sent or messages fired, and track meetings booked per named account rather than per campaign. A framework that’s working shows up as fewer, better conversations with the right people, not more email sent.

Where to start this month

If none of this is in place yet, the order matters. Audit your SPF, DKIM and DMARC records first, because everything else here assumes your email is actually landing somewhere. Then look honestly at your list: how many contacts haven’t opened anything in the last six months, and what removing them would do to your numbers.

Only after that is it worth turning to the outreach side, building a realistic target account list rather than a market-wide one, and mapping the buying committee inside two or three of those accounts as a starting point rather than all of them at once. Sustainable growth in B2B, the kind that doesn’t put your domain reputation or your client relationships at risk to hit a short-term number, gets built in that order, not the reverse.

If this kind of thing is useful, I write one case study, one insight and one actionable idea like this every Monday in my newsletter. Subscribe here.

Frequently asked questions