The $6.5M Question Behind beehiiv Hex

A note on the numbers: beehiiv’s supplier pricing and Hex operating costs are private. Any costs or potential savings discussed below are illustrative estimates based on public pricing and stated assumptions, not verified figures from beehiiv. I use them to explain the economics behind the decision.

Many email platforms own the editor, analytics, billing and contact database.

Some spend years, millions of dollars and countless engineering hours building them. Others vibe-code an entire ESP to save $59 a month… but that’s a story for another issue.

When you press Send, though, the email often travels through infrastructure run by another company.

beehiiv says it has changed that.

Its new sending system, Hex, includes a proprietary sending engine, its own mail transfer agent (MTA) and IP addresses that beehiiv says it owns and operates.

Sounds like a technical backend update for newsletter publishers.

It’s actually a business decision that raises a question for any company that relies on email for revenue:

Do you know who controls the infrastructure your email travels through?

You don’t need to use beehiiv to care about that.

If you use HubSpot, Klaviyo, Braze, Mailchimp or any other platform, the same question applies to you.

DEEP DIVE

What has beehiiv built?

An MTA is the system that moves email from the sender toward Gmail, Yahoo, Microsoft and other receiving networks.

Before Hex, beehiiv depended more heavily on external delivery infrastructure.

Now it controls more of that process itself.

According to beehiiv, Hex currently handles more than 137 million emails per day.

At that daily rate, beehiiv would send roughly 4.1 billion emails per month and approximately 50 billion per year.

Separately, it reports peak sending capacity of up to 3.7 million emails per minute.

beehiiv also says Hex can send to one million recipients in about 16 seconds and deliver the median welcome email in 1.8 seconds.

Sounds positive right? But speed is only part of the story.

One of the hard parts of email infrastructure is knowing when to SLOW DOWN.

Gmail might accept one sending pattern while Microsoft starts temporarily rejecting it.

Yahoo might need a different retry strategy.

And one group of senders might have excellent engagement while another generates complaints.

Hex reportedly monitors these signals and changes its sending speed when a receiving provider starts pushing back. At least that is promised by beehiiv.

The $6.5 million benchmark

At beehiiv’s scale, the economics get interesting.

Amazon SES lists these Enterprise rates (note: Beehiiv is using Sendgrid as main infrastructure, but Sendgrid is less transparent about it’s prices for that level of sends):

  • First 10 million emails each month: $0.23 per 1,000

  • Next 90 million: $0.18 per 1,000

  • Everything above 100 million: $0.13 per 1,000

  • Monthly account fee: $500

At 137 million emails per day, those public rates would produce an illustrative Amazon SES Enterprise bill of approximately $6.57 million per year, before data transfer and additional services.

But this does not mean beehiiv was paying $6.57 million.

And it does not mean Hex now saves beehiiv that amount.

beehiiv’s previous supplier contract is private. Companies sending billions of emails can negotiate custom pricing.

Plus, Hex has significant operating costs of its own:

  • Engineers and deliverability specialists

  • Servers, bandwidth and data storage

  • IP addresses and reputation management

  • Monitoring and abuse prevention

  • Incident response

  • Work with mailbox providers

So here’s the honest answer: we don’t know exactly how much Hex costs or how much beehiiv saves.

The Amazon SES calculation is a benchmark, not a confirmed saving.

Still, at about 50 billion emails per year, even a small reduction in cost per thousand adds up.

If beehiiv reduced its effective infrastructure cost by just $0.02 per 1,000 emails, that would equal roughly $1 million annually at its current stated volume.

What does owning the infrastructure change?

Operating its own sending infrastructure gives beehiiv more direct control over traffic, troubleshooting and sender separation.

External providers can offer these capabilities too.

The trade-off is how much flexibility beehiiv gains against the cost and responsibility of operating the system itself.

Owning infrastructure also means owning outages, maintenance and the staffing required to keep it reliable.

Here’s what the extra control could look like in practice.

More direct control over traffic

beehiiv can decide how traffic is distributed between IP pools, regions and mailbox providers.

It can build different sending policies for established senders, new accounts, imported audiences and potentially risky senders.

Faster troubleshooting

When delivery problems appear, beehiiv no longer needs to depend entirely on another vendor to investigate them.

Its team can see more of what happened, which providers were affected and how the system responded.

Clearer sender separation

A platform like beehiiv has to stop one bad sender from damaging everyone else.

Owning the infrastructure lets it write its own rules for screening lists, detecting complaint spikes and moving risky senders into separate pools.

This doesn’t eliminate shared-reputation risk. But it gives beehiiv more control over managing it.

There may be a commercial upside too.

Large customers care about predictable send times and visibility, and an MTA that reliably processes billions of emails a month may be harder for competitors to copy than another analytics screen.

🪄 What should other companies learn from this?

Few companies will need to build their own MTA.

At smaller volumes, an external provider is often cheaper, faster and safer than hiring an infrastructure team.

But as email volume grows, the calculation changes.

If email drives a meaningful part of your revenue (and I am talking about billions of emails per year), ask:

  1. Do we know our effective cost per thousand sends?

  2. Who actually operates the infrastructure underneath our platform?

  3. What happens when Gmail or Microsoft starts rejecting our traffic?

  4. Can our provider isolate our reputation from risky senders?

  5. Do we receive performance data by mailbox provider?

  6. How dependent are we on one infrastructure supplier?

  7. Would changing providers interrupt critical customer emails?

You probably don’t need to build your own sending infrastructure.

But you should understand who controls it, how much it costs and what happens when it fails.

How to find out who runs your sending infrastructure

Here’s a quick way to start.

Step 1: Look for clues in the headers.

Send a test email from your platform to a Gmail address and open the original message. Look at the sending IP addresses in the Received lines, and check who they belong to.

Step 2: Check the SPF and DKIM domains.

They provide clues about the domain used for SPF authentication and the domain signing the message.

Step 3: Ask your platform directly.

Ask whether it operates its own MTA or uses a third party, whether you’re on shared or dedicated IPs, and what mailbox-provider reporting you can see.

One important limit: headers provide clues, but they do not prove who owns the sending software.

An Amazon IP could host a company’s own MTA. A branded DKIM domain identifies the signing domain, not necessarily the delivery provider.

Confirm your findings with the platform.

What do the launch numbers actually prove?

beehiiv reports that 99.1% of emails sent through Hex are accepted by inbox providers, and that the average open rate is 43.3%.

Those figures sound strong.

But they need the right interpretation.

Accepted does not mean inboxed

An accepted email has been received by Gmail, Yahoo or another destination.

It can still land in:

  • The primary inbox

  • The Promotions tab, which is also part of the inbox

  • Spam

  • Another filtered location

A 99.1% acceptance rate tells us receiving servers aren’t rejecting most Hex traffic.

It does not prove that 99.1% of messages reach the inbox.

Open rates don’t prove infrastructure quality

Opens are influenced by:

  • Audience quality

  • Subject lines

  • Sending frequency

  • Newsletter category

  • Apple Mail Privacy Protection

  • How recently the subscriber joined

That average does not tell us how much engagement comes from readers, automated opens or changes in the mix of publishers.

It cannot establish that Hex improved performance.

To evaluate Hex properly, here’s what I’d want to see:

  • Temporary and permanent rejection rates by mailbox provider

  • Complaint rates across different sender groups

  • How quickly risky senders are identified and isolated

  • Results for established senders before and after migration

  • Whether senders see measurable changes at Gmail and Microsoft

That data would tell us far more than one platform-wide open rate.

Better infrastructure can’t fix a bad email program

A better sending engine can’t make up for:

  • Purchased or scraped lists

  • Bot signups

  • Subscribers who didn’t expect the email

  • Long-inactive audiences

  • High complaint rates

Even with better infrastructure, your signup practices, audience expectations and complaint rates still affect whether your emails reach the inbox.

That’s it

At this scale, the cost and control of delivery become consequential business decisions.

Hex suggests beehiiv decided the benefits of operating that layer justified the investment.

But owning it does not make deliverability automatic.

beehiiv now controls more of the delivery chain.

Senders still control the quality of the audience they send to.

What email platform does your company use, and do you know who operates the sending infrastructure underneath it?

Reply in comments.