Build or Buy an Email Editor in the Age of AI?

AI coding assistants such as GitHub Copilot, Cursor and Claude Code have made building an email editor for SaaS products tempting, because an assistant can write most of the code. They haven't made owning an editor cheaper. Testing, fixes, feature requests and security continue for as long as your users depend on the editor, and the cost of owning the editor is what decides whether a SaaS team should build its own editor in-house or buy one.

This article explains what building an email editor involves and which part of the work an AI coding assistant speeds up, what AI does not change about owning the editor, and what embedding a ready-made editor costs instead. It then covers what four SaaS teams that embedded an editor report, when building your own is still the right call, and the questions to answer before you decide.

An email editor your users depend on is several systems working together, and the drag-and-drop interface they see is only one of them. Each system has work that continues indefinitely after launch:

An AI coding assistant writes the code for each component of an email editor faster than a developer can by hand. It helps most with the drag-and-drop interface, because an interface like this is standard web code, and open-source email editor projects give the assistant code to start from.

Owning an email editor means keeping its output correct and its features current for as long as your users depend on it. Over the years, this means a long list of rendering fixes, security fixes and feature improvements. An AI coding assistant can help write each fix, but it cannot produce the list of things to be fixed, because this comes only from running an editor in production with real email clients and real users. We at Topol have ten years' experience of running an editor in production.

There are four types of challenges that maintaining an email editor involves.

Email clients do not display HTML the way web browsers do, and they do not display it the same way as each other. Classic Outlook for Windows displays email with Microsoft Word's HTML engine, and no Outlook for Windows version from 2007 through 2019 supports CSS flexbox, grid or background images. An editor's output therefore has to be nested HTML tables, with workarounds for individual email clients.

An AI coding assistant can write table-based email HTML, but the only way to confirm how an email client displays an email is to open the email in it. And every template your users build has to display correctly for every person they send it to, in whichever email client that person uses...Not easy.

Email clients update on their own schedules, and an update can change how an existing workaround behaves. When this happens, templates your users built months earlier can display incorrectly for the people they send them to. Sure, an AI coding assistant can write the fix faster, but someone on your team still has to learn about the break, reproduce it, review the fix and release it.

Every editor feature your users request after launch, such as a new content block or a dark mode preview, is work for your team.

Additionally, adding an AI writing assistant inside an email editor brings a running cost that other features do not, because it means paying a model provider for every request your users make. Providers such as Anthropic price per token, so this cost scales with usage.

Templates are HTML your users supply, which makes an email editor a stored cross-site scripting (XSS) surface. A script saved in one user's template runs in the session of whoever opens the template next, whether that is a teammate in the same account or someone on your own support team.

For editors in which users author styled content, OWASP's cross-site scripting guidance recommends HTML sanitization, because output encoding would break the formatting users expect to keep. Regardless of who writes the sanitization code, your team owns it, and every new editor feature that accepts content from your users has to be checked against it.

Embedding an editor means paying a vendor for an editor you add to your product, while the vendor owns the editor after launch. Topol Plugin, Topol's embeddable email editor, costs $70, $140 or $300 per month, depending on whether you prepay for 50, 100 or 500 users, with flat pricing that has no per-export or per-API-call fees. It comes as official typed packages for React, Vue, Svelte and vanilla TypeScript.

Every tier is white-labeled, so the editor shows no Topol branding. Every tier also includes an AI writing assistant that drafts and rewrites copy inside the editor, with no credit packs to buy and no API key of your own, so your team does not pay a model provider for your users' requests. From the $140 Expansion plan, AI Chat, a conversational assistant inside the editor, builds and edits the template as your users describe what they want. Topol Plugin renders designs through MJML, the open-source email framework, and its output works across 90+ email clients.

On the $70 Startup plan, Topol hosts your users' images, with 50GB of data traffic included. Serving images from your own AWS S3, Google Cloud Storage or Cloudflare R2 storage starts on the $140 Expansion plan.

Building your own editor is paid in developer time, from the same budget as your product roadmap. A flat monthly price is known on the day you decide, so it can be weighed directly against the developer time an editor of your own would need each month, while a price that grows with usage cannot be forecast until the usage exists.

The figures in this section come from the teams' own case studies.

YourPCM Limited runs four SaaS products on a server-side stack backed by SQL Server, and its developers embedded Topol Plugin, a browser-side JavaScript component, in all four without a single support request to Topol.

• 2x faster email creation for YourPCM's own team

• Satisfied customers, who build their own newsletters and outreach emails in the editor

Message34, a platform for bulk, transactional and multi-channel messaging, swapped its built-in editor, which had no drag-and-drop interface, for Topol Plugin and has used it for more than five years.

• 3x faster template creation, from 30 to 40 minutes down to 10 to 15 per template

• 30 labor-hours saved per customer each month, on average

• 30% fewer support requests related to email design

Affinect, a customer data, loyalty and lifecycle marketing platform for hospitality businesses, embedded Topol Plugin fully white-labeled and built custom blocks for coupons, loyalty, rewards and platform data.

• An estimated 5+ developer-months saved on building an editor, plus ongoing maintenance costs

• Consistently happy customers, many of whom prefer the editor in Affinect to Mailchimp's

• Developers free to work on what makes Affinect's platform different

Customerscore.io, whose platform runs automated email workflows that help brands re-engage their users, chose Topol Plugin after the other editors it tested produced emails that looked different in inboxes than they did in the editor.

• An estimated 3+ months of development work saved

• 2x faster email template creation

• 100% positive feedback from Customerscore.io's customers

Embedding buys a general-purpose editor that a vendor maintains. Building your own is the better choice when a general-purpose editor cannot do what your product needs, because owning the editor then becomes part of your core product rather than a cost on top of it. There are two such situations:

• The editor is the product you sell. If customers pay you for the email editor itself, a general-purpose editor limits how far your product can differ from its competitors.

• Your product needs output other than standard email HTML. If no vendor's rendering produces the format you need, embedding cannot deliver it.

Some products let users change the words and images inside a few fixed templates, and nothing else. These products need a template form, a fixed email template with editable text and image fields, rather than an email editor. An AI coding assistant can build a template form quickly, and owning one costs far less than owning an editor, because your team tests a few fixed layouts instead of every layout your users could build.

Answer these questions before you commit to the build or start a vendor trial:

• Is the email editor what your customers pay you for, or a feature that supports what they pay for?

• Does your product need email HTML that works in every major email client, or a format of its own?

• Who on your team will test the editor's output when an email client changes, and which roadmap work will wait while they do?

• What will an editor of your own cost each month in developer time, and how does this compare with a flat subscription?

If your answers favor embedding, our embeddable email editor SDK evaluation guide lists the questions to ask each vendor, and our comparison of the best embeddable email editors describes the vendors themselves.

A team that already runs its own email editor faces this build-or-buy decision each time the editor needs work: keep paying the ownership cost, or replace the editor. Message34, whose users regularly needed support with its built-in editor, chose to replace it with Topol Plugin, and its design-related support requests fell as a result.

Yes, an AI coding assistant can write the code for an email editor, including its drag-and-drop interface. The larger cost is owning the editor after launch: testing its output in each email client, fixing it when email clients change, adding the features your users request, and securing the content they supply.

Buying is cheaper whenever the developer time your own editor would need each month costs more than a flat subscription. An AI coding assistant lowers the cost of writing an editor, but testing, fixes, feature requests and security continue for as long as your users depend on it. Topol Plugin's flat tiers cost $70, $140 or $300 per month, and Topol does this ownership work.

Yes, open-source email editors give your team working code for the drag-and-drop interface, and an AI coding assistant can adapt it to your product. Your team still owns everything around it: the rendering tests, the fixes when email clients change, image hosting, and the security of the content your users supply.

An AI coding assistant writes an email editor's code faster than hand-coding does, and the drag-and-drop interface fastest of all. The code was never the expensive part of an editor: testing, fixes, feature requests and security continue for as long as your users depend on it, and they stay with whoever owns the editor.

Embedding turns the ownership of an editor into a flat monthly cost you can forecast. The teams that embedded Topol Plugin spent their developers' time on their own products instead, and one of them used its AI coding assistant for the integration.

Build your own when the editor is the product you sell, or when your product needs output other than standard email HTML. If your users only change text and images in fixed templates, a template form may be all you need.

If the email editor is a feature of your product rather than the product itself, start a 14-day trial of Topol Plugin and test it inside your own app.