Email marketing operations, run by people who own deliverability.

Most marketing operations are judged on what is visible, the design, the copy, the calendar, and assume the underlying plumbing just works. It often does not. Emails land in spam, authentication is misconfigured, sending reputation erodes, and the team has no visibility into any of it because it sits a layer below the platform they use. A campaign that does not reach the inbox is indistinguishable from one that was never sent, no matter how good the creative was.

The cost of that gap has grown sharply. Mailbox providers spent two years turning email best practice into enforced rules, and the penalty is no longer a soft dip in performance. Compliant senders average around 89 percent inbox placement, while non-compliant ones see roughly a fifth to a third of their mail routed to spam, a three to seven times penalty on the audience a campaign actually reaches. The gap between a well-run programme and a careless one has never been wider, and it is mostly invisible from inside a marketing dashboard.

We come at marketing operations from that hidden layer. The same capability that runs production sending infrastructure for deliverability clients runs your campaigns, which means the unglamorous part, getting the message delivered, is owned by people who do it for a living rather than assumed to be someone else's job.

The operation covers the full engine room. Campaign and newsletter production with real QA, so broken links and rendering faults do not ship. Segmentation and list hygiene, so you are sending to people who want to hear from you rather than to a list that quietly poisons your reputation. Lifecycle and automation flows that do the repetitive work, welcome series, re-engagement, post-purchase, without a person rebuilding them each time. Deliverability monitoring with remediation when a signal dips. And reporting that connects sends to outcomes rather than to a vanity metric.

Underneath all of it sits infrastructure we control, PowerMTA, dedicated IP ranges, and SPF, DKIM and DMARC configured properly, so reputation and authentication are things we tune rather than things we hope a third-party platform is handling. You keep strategy and brand; we keep the machine running. That division is the whole point of outsourcing operations rather than hiring another marketer: you are buying a working, compliant, monitored sending capability, not a pair of hands to operate a dashboard you still have to understand. Bringing a new sending domain or IP online is part of the same operation rather than a separate scramble, with a measured warm-up and seed testing before volume ramps, because a careless launch damages the reputation everything else depends on.

Three providers, Gmail, Yahoo and Microsoft, now enforce a common baseline for anyone sending at volume, broadly five thousand or more messages a day to their consumer inboxes. The requirements are similar in principle: authenticate with SPF and DKIM, publish a valid DMARC record that passes alignment, keep the spam complaint rate below 0.3 percent and ideally under 0.1, and offer one-click unsubscribe on marketing mail with opt-outs processed within two days. Authentication applies to all your mail, including transactional; the one-click unsubscribe requirement applies to marketing and promotional messages, not to order confirmations and password resets.

The scale behind the enforcement is easy to underestimate. Gmail alone blocks something in the order of fifteen billion unwanted messages a day, which is why its tolerance for partial setups and "good enough" configurations has dropped. Even two years into enforcement, roughly a third of bulk senders remain partially non-compliant on at least one requirement, most often the one-click unsubscribe header, which means being in the compliant majority is now a quiet competitive advantage rather than a baseline everyone clears. A DMARC record that exists but is misaligned, or an unsubscribe link that is visible but missing the required header, both count as failures even though they look done.

The detail that catches people out is what happens on failure. This is no longer a filtering question. Non-compliant bulk mail is rejected at the SMTP level, with Google returning a 550 error and Microsoft a 550 5.7.515, and the message never reaches any folder. The providers also differ in the small print, Yahoo calculates its complaint rate against inbox-delivered mail only, which makes it effectively stricter than Gmail's, so a sender who is compliant at one can be over the line at another. We manage to the strictest interpretation rather than the most convenient one.

These are operational defaults for us rather than a project. The compliance gate below runs before a campaign goes out, so a misconfiguration is caught on our side rather than discovered through a wave of rejections.

# authentication — applies to all mail
$ argus mail preflight --campaign spring-newsletter
  [ok] SPF pass   ·   DKIM pass   ·   DMARC aligned
  [ok] PTR (reverse DNS) set   ·   TLS available

# marketing requirements + reputation
  [ok] one-click unsubscribe: List-Unsubscribe + -Post (RFC 8058)
  [ok] complaint rate (all platforms): 0.04%  < 0.1% target
  [ok] list hygiene: invalid + unsubscribed removed
  cleared to send  ·  est. inbox placement: high

Passing the bulk-sender rules gets you considered; it does not get you the inbox. Even fully authenticated mail still lands in spam a meaningful share of the time, because mailbox providers weigh how real people engage with your messages, opens that lead somewhere, clicks, replies and the absence of complaints, far more heavily than a clean authentication check. Authentication is the condition for being assessed at all; engagement and reputation decide where the mail actually goes.

This is where most marketing operations stop short, because the levers that move engagement and reputation live below the platform. We run our own sending infrastructure, PowerMTA with dedicated IP ranges and authentication we control, so we can shape the things a hosted platform hides: which IP a stream sends from, how a new domain or address is warmed, how reputation is protected when volume changes. Industry data consistently traces the large majority of deliverability failures back to list quality rather than content, which is why an operation that owns the sending layer can fix problems a creative-led agency cannot even see.

The complaint rate is the number that decides whether a programme survives, and it is unforgiving: a sustained rate at the 0.3 percent enforcement line will get a domain blocked regardless of how clean the authentication is. Keeping well under it, ideally below 0.1, is operational discipline rather than luck. It comes from segmentation so the right message reaches the right people, list hygiene that removes invalid and dormant addresses before they drag engagement down, suppression of unsubscribes and complainers inside the required window, and re-permission of contacts who have gone quiet rather than continuing to mail them.

Two operational facts shape how we manage it. The rate is aggregated across every platform that sends from your domain, so a clean marketing stream can still be sunk by a noisy transactional one, and all of it has to stay under the threshold together. And the providers measure differently, with Yahoo's inbox-only denominator producing a higher rate from the same complaints than Gmail's, so we manage to the strictest of them. Cold outreach is its own category of risk here, routinely producing complaint rates several times the threshold without careful targeting and hygiene, which is why it needs a separate, disciplined operation rather than being folded into the marketing stream.

The same discipline that keeps mail in the inbox keeps it lawful, and the two are not in tension. We operate on a proper lawful basis for the data you hold, consent or legitimate interest as appropriate, honour unsubscribes and data-subject requests promptly, maintain suppression lists so a removed contact stays removed, and keep the personal data inside the EU. A list that is consented, current and well-suppressed is both compliant and good for deliverability, because the people on it actually want the mail and are unlikely to complain.

The failure modes are shared too. Mailing people who never opted in, ignoring unsubscribe requests, and holding stale data are simultaneously a data-protection problem and a deliverability problem, and they tend to arrive together. Treating marketing operations as data-protection operations, rather than two separate concerns bolted together at audit time, is what keeps a programme out of trouble on both fronts at once.

An operation is only as good as what it can prove, and the metrics that used to anchor email reporting have decayed. Privacy features that pre-load images on the provider's side inflate open counts, so an open is no longer reliable evidence that a person read anything. Reporting that still leans on open rate is measuring a number that looks healthy while telling you very little, which is how a quietly failing programme can appear fine on a dashboard for months.

We weight reporting toward what still carries signal: inbox placement measured against seed and provider data, clicks, replies, conversions and revenue, and the deliverability indicators that predict trouble before it shows up in results, complaint rate, authentication status, and the reputation signals in Google Postmaster Tools and Yahoo's feedback loop, which we monitor as separate dashboards because they disagree by design. The result is reporting that connects a send to what it actually did, so a decision to change a segment or a cadence rests on outcomes rather than on a vanity metric that survived the privacy changes by accident.

We run your email and lifecycle programme on infrastructure we operate ourselves, keep it compliant with the bulk-sender rules by default, hold the list and the data to GDPR, and report on outcomes rather than vanity metrics. You keep strategy and brand; we keep the machine running and the mail landing. Where it makes sense we send from our own PowerMTA stack for the deliverability advantage, and where you have an existing platform we work with it, while being clear about when that platform is the thing holding your results back rather than your content.

Our remit is operations and deliverability, and we name its edge rather than overselling. We can produce and adapt campaign content, but where you have a strong creative team we run the engine room underneath them rather than competing for the brand work. If the limiting factor on your results is the list, the platform or the compliance setup rather than the copy, we will say so and fix the actual problem, because an operation that flatters the creative while the deliverability quietly fails is the exact failure we exist to prevent.