The Future of MTAs: Why Amazon SES Won the Sending Layer and Where Retailers Should Compete, The Zembula View

When Amazon announced it would retire Pinpoint on October 30, 2026, it didn’t just sunset a product, it made the most honest statement any vendor has ever issued about the email stack. The message? There’s no future in competing at the sending layer. And if Amazon won’t fight there, neither should you. At zembula, we see this moment not as a disruption, but as confirmation: email’s real value isn’t in how you send, but in what you send and when you decide it.

The MTA market has already spoken. Amazon SES charges $0.10 per 1,000 emails. SendGrid killed its free plan in 2025. SparkPost sold to MessageBird for $600 million and became Bird Email. These aren’t random events, they’re a coordinated retreat from a race to the bottom. Sending email is now a utility, like bandwidth or electricity. The only real question left is: if the pipe is commoditized, where should retailers actually invest?

Pinpoint’s Retirement Is a Market Signal, Not a Migration Crisis

Let’s be clear: Amazon isn’t just sunsetting Pinpoint. It’s directing customers to bare SES, no journeys, no segmentation, no campaign analytics. Just sending. AWS’s own documentation states: “If you are using Amazon Pinpoint to send email, we recommend you migrate to Amazon Simple Email Service (SES).” That’s not a product transition. It’s an admission that the value above the pipe isn’t worth building.

And they’re not alone. Twilio ended SendGrid’s free plan in March 2025, replacing it with a 60-day trial. That move wasn’t about UX, it was a margin play. Why give away sending capacity when no one competes on price anymore? Even third-party analyses now frame SendGrid as a mid-tier option between budget SES and full ESPs, with deliverability tools as the differentiator, not content or orchestration.

Then there’s SparkPost, once the largest independent MTA, selling for $600 million in 2021. The acquirer didn’t keep the brand. It rebranded it Bird Email and folded it into an omnichannel stack. The takeaway? Even the biggest pure-play sender knew raw sending couldn’t scale as a standalone business.

MTA vs ESP: What Each Layer Actually Does (And Doesn’t Do)

Let’s clarify the roles. A Mail Transfer Agent (MTA) is the engine that delivers email. Amazon SES, SendGrid, and Bird Email live here. Their job is deliverability, throughput, and cost efficiency. They win on uptime and inbox placement, not engagement.

An Email Service Provider (ESP) adds orchestration: journeys, segmentation, templates, compliance. But most ESPs are built on batch architecture. That means your content is locked in at send time, even though 10% or more of email revenue comes from opens on emails sent over a week ago.

The problem? ESPs weren’t built for real-time decisioning. They batch data, schedule sends, and render content once. That worked in 2015. It fails today, when customer behavior shifts between send and open.

So when teams debate “Should we move to SES or SendGrid?” they’re asking the wrong question. The real issue isn’t the sender, it’s the frozen content inside the email.

The Experience Layer: The Missing Piece Between Data and Delivery

Between your customer data and your sender sits a layer nobody owns: the experience layer. This is where content decisions happen, not at send time, but at open time. What does a subscriber see when they open your email a week later? What if their cart dropped? What if they earned loyalty points? What if pricing changed?

Most email workflows can’t answer that. But the opportunity is massive. According to our 2025 email performance benchmark report, static batch content earns roughly $130 per 1,000 sends. But multi-signal personalized content blocks, like Smart Banners and Smart Kickers, benchmark at $469.65 per 1,000 sends when combining cart, loyalty, and price-drop signals. That’s not optimization. That’s a new revenue tier.

zembula operates in this layer. We don’t replace your MTA or ESP. We sit between your data and your sender, rendering dynamic content at open time as a single, pixel-perfect image, compatible with any ESP output, including SES-delivered email.

No template rework. No 102KB clipping risk. Just smarter content that earns more, regardless of how you send.

The Retailer Math: Sending Costs a Dime, Revenue Lives Elsewhere

Let’s run the numbers. Sending 1,000 emails via Amazon SES costs $0.10. That’s not a decision point, it’s a rounding error.

What matters is what those emails earn. The retail industry baseline for click-to-conversion (CTC) is 2.5%. But personalized content blocks, using real-time behavioral, transactional, and loyalty data, average 13.6% CTC. That’s not a minor lift. That’s a 444% improvement in conversion efficiency.

And because zembula attributes revenue to the block and variant level, with 7-day click windows, we can prove it. Our Q4 2025 benchmarks show Cart-only content earns $135.30 per 1,000 sends. But Cart + Loyalty + Price Drop together? $469.65. That’s a $334.35 incremental gain, for a feature that costs $0.035 per 1,000 impressions.

The spread between sending cost and revenue potential is where your email strategy should live. Not in migrating MTAs. Not in re-platforming ESPs. In winning at the experience layer.

How Zembula Fits: Complementary, Not Competitive

We don’t sell sending. We don’t sell journeys. We help you monetize the gap between send and open.

Our architecture is simple: a single dynamic image URL embedded in your email. When a subscriber opens, we decide in real time, based on their behavior, preferences, and context, what they see. No batch delays. No template lock-in. Just revenue-driving content, rendered perfectly.

And because we charge per personalized impression, $0.035 per 1,000, our incentives align with yours. We only get paid when your content performs.

This is why we’re platform-agnostic. Whether you use SES, SendGrid, or Bird, zembula works. We integrate once with your data, and every template has access. No per-template wiring. No per-app sync. Just scale.

It’s also why we offer a written 10x ROAS floor on Smart Banners. If we don’t deliver at least 10x return, you don’t pay. That’s confidence in architecture, not just features.

Key takeaways

• Amazon SES, SendGrid, and Bird Email have turned sending into a utility. Competing here is a losing strategy.
• Your ESP handles orchestration, but its batch architecture caps revenue potential, especially for delayed opens.
• The real opportunity is the experience layer: deciding content at open time, not send time.
• Personalized content blocks like Smart Banners and Smart Kickers can earn up to 3.5x more than static batch content.
• zembula operates in the experience layer, integrating with any MTA or ESP to deliver real-time, revenue-driving content.
• The cost to send is negligible, $0.10 per 1,000 emails. The cost to earn more? A few cents per thousand personalized impressions.
• Focus your budget where RPM and CTC actually move: the content inside the email, not the pipe that delivers it.

For more on how real-time architecture changes email economics, read Moment-of-Send vs. Moment-of-Open: What Real-Time Email Really Means. To understand the financial impact of batch vs dynamic content, see Your ESP Batches Behavioral Data. Smart Banners Don’t. Here’s What That Costs You. And for a deeper dive into email as a performance channel, check out Email Is a Performance Marketing Channel, and the Math Proves It.

To explore how Smart Banners can increase your email ROI, start with The Ultimate Guide to Smart Banners™, or learn how The Real-Time Revenue Gap limits what your current ESP can earn.