Plain text emails have higher open rates than HTML emails.
I heard this recently from someone who’d consulted ChatGPT like it was the Oracle at Delphi and wanted to overhaul an email strategy based on this single pronouncement, without consideration of their company’s industry, products, audience, or business goals.
These days, marketing “best practices” are spread through the digital equivalent of a game of telephone, with LLMs serving as the always-enthusiastic amplifiers of oversimplification. You don’t even have to read skim a marketing book or blog post anymore!
Part of the problem is that LLM users over-rely on their own personal oracles because they lack either the knowledge (maybe) or the confidence (definitely) to speak intelligently on a given subject. Also, LLMs are trained to be agreeable, which is why they sound so chipper every time you call out their absolute bullshit hallucinations. Only when you follow-up with more specifics and push ChatGPT, Claude, or others LLMs into being more careful can you get something like worthwhile advice. But that takes time…maybe as much time as reading a few blog posts on the topic.
As with most matters in life, some good old-fashioned discernment can go a long way when it comes to LLMs. So let’s apply some to this “Plain text emails have higher open rates than HTML emails” assertion.
First Ask: Does This Make Sense?
Email recipients decide whether to open an email based on what they see in their inbox—the subject line, sender name, and maybe a preview snippet. The email’s internal formatting is usually only visible after an email is opened, so how exactly would an email’s format affect open rates?
That was my first thought when this particular “best practice” was shared with me, and only later did I learn that open rates are tracked through tiny pixels embedded in HTML emails. When you open an email, that pixel loads from a server, registering the open. How cool is that?
Truly plain text emails don’t include images, so they can’t be tracked for opens in that way.
What’s Actually Going On
When plain text emails do perform better, the real factors at play are far more interesting than format preference.
Deliverability matters. Plain text emails have fewer elements that spam filters flag as suspicious. HTML emails, with their complex coding and embedded images, are more likely to land in spam folders. An email that never reaches the inboxes of your potential customers has a 0% open rate, regardless of format. Plain text emails are more likely to get delivered and be read as intended, but that’s not usually a concern for most hyper-senders.
Context is everything. In B2B sales, plain text emails can feel personal rather than promotional, which is why many B2B emails might look plainer—or simpler, let’s say—even though they are not “plain text” emails. B2C brands and more visually-oriented industries often see the exact opposite results, however. Fashion retailers don’t succeed with plain text emails. The B2C clients we help don’t generally send plain text emails.
Measurement has changed. Apple’s Mail Privacy Protection and Gmail’s image caching have made HTML tracking less reliable, potentially skewing recent data toward plain text emails that rely on click-tracking instead. If a link is clicked in a plain text email, that email was obviously opened…but to make the case that the email was opened and the link was clicked because it was a plain text email is a reach.
The Real Problem: Outsourcing Your Brain
This plain text email confusion is really about how we’ve started treating Claude, ChatGPT, or other LLMs as equal partners in some kind of shared duty of thinking.
The trouble, of course, is that LLMs don’t think. They are trained on Everest-sized mountains of marketing content, with much of it repeating the same oversimplified “best practices.” Ask an LLM about email marketing, and it will synthesize this conventional wisdom into confident-sounding advice. It probably won’t ask about your industry or consider the nuances of your specific audience. It won’t account for the logical contradictions in the claim itself, as I’ve demonstrated above. If you do point out a rhetorical mishap, you’ll surely be praised (“You are right to question that…”).
In short, it will just tell you what answers get repeated most often.
This is how the cycle perpetuates: A limited study shows interesting results, context gets lost, a simplified version spreads across various websites, repetition creates false consensus, and then a manufactured “best practice” becomes gospel. LLMs accelerate this process by making it easier than ever to find support for any claim without actually thinking about whether it makes sense.
As research tools, LLMs are remarkably powerful. But they’re tools for augmenting your own thinking, not replacing it. A one question/one answer interaction is unlikely to provide everything you need to make an informed decision about your business.
The person who consulted ChatGPT about plain text emails wasn’t wrong to seek information. Everyone has to start somewhere. But they were wrong to accept that information without engaging their own critical faculties, digging deeper, and allowing ChatGPT to be the final authority on that matter.
The Case for Actually Thinking
The wrong data can be as harmful, or more harmful, than having no data. Clear-eyed marketers know not to focus on vanity metrics or put too much trust in those who do. As for changing a strategy based on a single test, or preferring quick answers over rigorous testing, well…I guess some people like things to be easy, even when it’s wrong.
Actual thinking means questioning claims that sound too universal to be true, understanding the mechanism behind any supposed effect, and testing within your actual context rather than copying someone else’s method and hoping for similar results.
When I first came to Rare Bird three years ago, a colleague mentioned that our fearless leader was not especially fond of “best practices.” Now it’s easy for me to understand why.
Marketing success doesn’t come from using the latest “proven” tactic. It can be found by understanding your audience, questioning oversimplified advice, maintaining healthy skepticism toward anything that claims to work universally, and being willing to buck common trends if that’s what your business needs in the moment.
There’s a great line about portfolio diversification on the website of Woodley Farra Manion, one of our clients, that perfectly captures the problem with marketing best practices: “If you own the entire market, it’s impossible for your portfolio to outperform it.”
Translation: If you’re doing what everyone else is doing, you’ve guaranteed mediocrity.
You have a brain. If you launched a business before the end of 2021, when we first learned of ChatGPT, and have been even moderately successful, it’s clearly a capable brain.
Trust it.
Ready to build a marketing strategy that works for your actual business, not everyone else’s? Let’s talk about your audience.
You Might Also Like:
- The Authenticity Paradox
- Leading Your Brand Through Tariff Turbulence
- Does Your Digital Presence Serve Everyone?
- The Myth of the Rational Buyer
- Building Customer Loyalty on a Budget
Sign up for Bird Feed, our monthly newsletter, to receive articles like this in your inbox.
