For eighteen months, we used a well-known email marketing platform for client campaigns. I'm not going to name it because the problems weren't unique to that platform. They're systemic across the entire category.
The problems were straightforward. Deliverability was inconsistent. We'd send a campaign on Tuesday and see 94% delivery. Send an identical campaign to a similar list the following Tuesday and see 87%. No explanation, no transparency into what changed. The platform's shared sending infrastructure meant our client's emails were affected by every other user on the same IP range, and we had no control over that.
Warm-up was a black box. The platform claimed to handle domain warming automatically. What that actually meant was unclear, and when we pressed their support team, the answer amounted to "trust the system." That's not an answer when a client's sender reputation is on the line.
Analytics were superficial. Opens, clicks, bounces, unsubscribes. The basics. But the granular data we needed, per-recipient engagement patterns, send-time analysis, inbox placement rates, were either locked behind an enterprise tier or simply unavailable.
So we switched. Not to another email marketing platform. To dedicated sending infrastructure.
Instead of using an all-in-one platform that handles list management, campaign design, sending, and analytics in one bundle, we separated the functions.
Sending goes through infrastructure we control. Dedicated IP addresses assigned to specific clients or campaign types. No shared reputation risk. If a campaign underperforms, we know exactly why because nothing else is running on that infrastructure.
Domain warming is a process we manage deliberately. We control the volume ramp, the engagement patterns, and the timeline. Instead of trusting a black-box algorithm, we follow a structured warm-up protocol: 20 emails per day to engaged contacts for the first week, increasing gradually over three to four weeks. Boring, manual, and it works.
Analytics come from the sending layer itself, combined with tracking we set up. Reply detection, open tracking at the recipient level, link engagement patterns, bounce categorisation (hard vs soft, and why), and inbox placement testing through seed addresses at major providers.
Deliverability improved by roughly 8% on average across campaigns in the first three months. That sounds modest until you do the maths. On a campaign of 5,000 emails, 8% better deliverability means 400 more emails reaching the inbox. If 20% of those get opened and 5% reply, that's 4 additional conversations from the same list, same content, just better infrastructure.
Warm-up became predictable. Instead of hoping the platform would handle it, we could see the exact sending volume each day, monitor reputation scores in real time, and adjust if anything looked off. One client's domain had a minor reputation dip in week two of warming (an overly aggressive spam report from a single recipient). We spotted it within hours and paused the domain for 48 hours before resuming at a lower volume. On a traditional platform, that dip would have been invisible until it became a serious problem.
Analytics became actionable. When we could see that prospects in the manufacturing segment were opening emails at 7:15am but prospects in professional services were opening at 9:45am, we adjusted send times by segment. Open rates improved by 6% just from that one change.
I'm not going to pretend this switch was painless.
Setup is more complex. A traditional platform takes an afternoon to configure. Dedicated infrastructure takes a week or two to set up properly, including DNS configuration, IP warming, and integration with whatever CRM and list management tools you're using.
There's no drag-and-drop email builder. Campaign design happens in a separate tool or in HTML directly. For someone used to clicking templates in Mailchimp, this is an adjustment. For us, it was an acceptable trade-off because we were already designing emails in code anyway.
The cost structure is different. Traditional platforms charge per subscriber. Dedicated infrastructure charges per email sent. For businesses with large lists and low send frequency, the traditional model can be cheaper. For businesses running regular campaigns to segmented lists, dedicated sending is usually more cost-effective.
And you need someone who understands email infrastructure. SPF records, DKIM signing, DMARC policies, IP reputation monitoring. These aren't skills every marketing team has. If you're a business owner running your own campaigns, this approach might be more technical than you want to manage.
If you're sending fewer than 1,000 emails per month, a traditional platform is probably fine. The deliverability issues exist but at low volumes they're manageable, and the convenience is worth the trade-off.
If you're sending 5,000 to 50,000 emails per month, particularly cold outreach where deliverability directly affects pipeline, dedicated infrastructure starts making a compelling case. The improved delivery rates, the warm-up control, and the granular analytics justify the additional setup complexity.
If you're running campaigns for multiple clients or across multiple brands, dedicated infrastructure is close to essential. Shared platforms create cross-contamination risk: one client's poorly targeted campaign affects another client's deliverability. Separate sending infrastructure for each client eliminates that entirely.
The switch was one of the better decisions we made last year. Not because the technology is exciting (it isn't), but because it gave us control over the one thing that determines whether a campaign works: whether the email actually arrives.
Martin Dugan, AA2
