Best Way to Separate Email Myths From What Works

There is no shortage of email marketing advice. Subject line formulas. Send time research. Personalisation studies. Format experiments. The problem is not the volume of advice — it is that most of it is contradictory, sourced from different audience types, and applied without testing whether it actually works for your specific programme.

The businesses that consistently improve their email performance are not the ones following the most advice. They are the ones with a clear method for separating what works for their audience from what sounds like it should work based on general principles.

Here is that method.

Why Most Email Marketing Advice Does Not Transfer Directly

The majority of published email marketing research is based on consumer campaigns. Retail newsletters. Subscription services. E-commerce flows. The audiences, buying cycles, decision-making structures, and content expectations of those programmes are fundamentally different from B2B.

A consumer email programme is optimising for impulse decisions made by individuals. A B2B email programme is often influencing a multi-month decision made by a buying committee. The tactics that work for one rarely transfer cleanly to the other.

When you read that the best subject lines use numbers and create urgency, that finding comes from consumer retail data. When you read that Tuesday at 10am is the optimal send time, the underlying research is from consumer newsletter programmes.

Neither of those findings is necessarily wrong. They are just wrong as universal rules. Whether they apply to your B2B programme depends entirely on your audience — and the only way to know is to test.

The Test Framework That Actually Separates Myth From Evidence

Test One Variable at a Time

The most common testing mistake is changing multiple things at once and attributing the result to the most recent change. Subject line, send time, CTA copy, and email length all changed between last week's send and this week's — and this week's performed better. What caused it? No one knows.

Test one variable. Keep everything else identical. Run the test on a segment large enough to produce statistically meaningful results. For most B2B lists, that means at least 500 contacts per variant.

Use a Matched Control

Every test needs a control — a version of the email that does not change. The control lets you measure the effect of the variable you changed rather than the effect of random variation between sends.

For B2B email testing, the control should come from the same source list as the test variant and be sent at the same time to avoid time-based confounds. If you send the control on Tuesday and the test variant on Thursday, you are not testing your variable — you are testing Tuesday versus Thursday.

Measure the Right Outcomes

The outcome you measure determines what you learn. If you test two subject lines and measure open rate, you learn which subject line gets more people to open the email. You do not learn which drives more replies, more bookings, or more revenue.

For B2B email, the hierarchy of meaningful outcomes is: reply rate first, booking or conversion rate second, click rate third, open rate last. A test that improves open rate by 8 percent but reduces reply rate by 3 percent is not a winning test. Optimising for opens while ignoring downstream metrics is how you build a programme that looks healthy on vanity metrics and consistently underperforms on revenue.

Specific Myths Worth Testing Against Your Own Data

"Shorter Subject Lines Always Outperform Longer Ones"

This advice circulates constantly. The evidence behind it is mixed and audience-specific. For some B2B audiences — particularly senior decision-makers who scan email headers quickly — shorter subject lines do outperform. For others, a longer subject line that communicates specific value outperforms a short one that is vague.

Test both. Send a 5-word subject line to one segment and a 12-word descriptive subject line to a matched segment. Measure reply rate, not just open rate. The answer you get will be specific to your audience.

"Plain Text Emails Outperform HTML Emails for B2B"

There is good evidence that plain text outperforms HTML for cold outreach to decision-makers, because plain text signals a personal, one-to-one communication rather than a marketing broadcast. But this is not universally true for all B2B email types.

For nurture newsletters, HTML with clear structure often outperforms plain text because readers expect a format that is easy to scan. For direct outreach sequences, plain text usually wins. The principle to test is whether the format matches the reader's expectation for the type of communication they are receiving.

"The More Personalisation the Better"

Personalisation improves performance up to a point. Beyond that point, over-personalised emails feel intrusive and damage trust.

An email that references the recipient's name, their company, their industry, and a specific piece of news about their company can feel like surveillance rather than relevance. The sweet spot for most B2B cold outreach is one or two contextual personalisation signals — industry-specific problem framing and role-specific language — rather than every personalisation field you have access to.

How to Build a Test-and-Measure Habit

The teams with the best email programmes are not the ones with the most sophisticated tools or the most experience. They are the ones who run tests consistently and apply what they learn.

A practical cadence is one new test per month on your most active campaign. Variable options: subject line format, email length, CTA language, send timing, level of personalisation, content type at a given funnel stage.

Over twelve months of consistent testing, you will have replaced general advice with twelve specific findings about what works for your audience. That knowledge compounds. A programme built on tested evidence outperforms one built on borrowed advice every time.

When Benchmarks Are Useful and When They Are Not

Industry benchmarks for email performance — average open rates, average click rates, average reply rates by industry — are useful for one thing: identifying whether your programme has a significant problem that warrants immediate attention.

If your open rate is 8 percent and the industry benchmark for B2B email is 22 percent, something is fundamentally wrong with your data, your domain reputation, or your subject lines. That benchmark is useful.

If your open rate is 19 percent and the benchmark is 22 percent, the benchmark tells you very little. It does not tell you whether 19 percent is good or bad for your specific audience, your specific programme, or your specific campaign type. Only your own historical data can tell you that.

Use benchmarks as a threshold test — are you dramatically below what is typical? If yes, investigate. If not, focus on your own trend data rather than comparing against industry averages that may not reflect your audience at all.