Email marketing has the highest ROI of any marketing tactic. $44 for every $1 spent, say Campaign Monitor: Email marketing is a tactic that’s hard to ignore. But how do you make sure you get good conversion rates? There’s plenty of blogs, articles, and user-guides you can download – but how do you separate the folk-law from the real laws? How do you separate the signal from the noise?

Here. Right here.

We’ve run gold standard Randomised Control Trials with robust test protocol written by psychologists and behavioural experts, and run through over 400,000 respondents; there’s no noise here.

1. Cold email: Getting it opened – subject line

On a study of business to business cold emails (approx. 45,000) we tried a variety of subject lines: Loss aversion (Kahneman & Tversky, 1979), Curiosity gap (Lowenstein, 1994; Wainer, Dabbish & Kraut, 2011), Social validation, ‘people like me’, Personalization (Sahni, Wheeler & Chintagunta, 2016) where ‘things’ associated with the self are persuasive, as they’re more liked and trusted (Kahneman et al., 1990; Nuttin, 1985; Reed, 2004).

  • The best and worst?
    The self is the best: e.g. The name of the business/their name
    The curiosity gap is the worst: e.g. ‘3 simple steps…’

Outcome: Use the ‘self’ subject lines.

2. Cold email: Getting a click to your website

On a study of business to business cold emails (approx. 9,000, error range way under 1% at 95% confidence) we tried a variety of content positions: Limited time offer, Personalisation, Social proof, Ease (of use), and all of these effects combined.

The best and worst?

  • Personalisation *or* ease-of-use techniques were the best
  • All effects combined were the worst

Outcome: Use personalisation *or* ease-of-use techniques, but watch out for that in the next section. This section is about clicks, not replies. Also, there’s another effect added to all of them – we’ll leave that for another post.

Overall we improved clicks from 2% of emails sent to 3.6% of emails sent (that includes all the emails unopened – the kind of responses you want from a cold campaign).

3. Cold email: Getting a reply

Everything but the kitchen sink; this is the answer to this challenge. With a study of 1,126 business owners (error range under 5% at 95% confidence) we included a range of behavioural techniques:

  • Reputation and reliability markers
  • A strong social proof message
  • Implicit endorsement (from the Financial Times)
  • Temporal narrowing (‘this week’) and a duration estimate to make the action more concrete/proximal
  • A schema
  • And a short, proximal, concrete outcome description (‘simply let me know when you have 10 mins and I’ll look forward to speaking with you’) 

Outcome: Use all of these. When compared with an email with fewer of these effects (and where they were written less well) we saw:

  • 33% increase in reply rates (from 7.71% to 10.29%)
  • As a rule-of-thumb we evoked 10 responses for every 8 in the pre-Hunting Dynasty control version


Beyond this work we delved into preview lines on emails, and mobile SMS text-back rates (~90,000 respondents and ~47,000 respectively for a global startup and a UK hospital Trust). We’ll come to that one day (or ask us!).

So, there’s some solid, concrete techniques to use today (and every day) to get your emails ‘sticky’. There’s more information – which is very useful – on ‘The ABC of RCTs’ here, on sample size for tests here (‘How many people do I need to test on? (Answer: It’s not as many as you think.)’), and all about pre-testing the value of understanding and motivation with our Comprehension-Motivation Test© here.

I’d email the links to you, but even though I know how to write a sticky email, I don’t have your email address… (how to get email addresses is for another time, and another post).

For more on this speak with us

Also, as co-founders and supporters of the London Behavioural Economics Network, join the Meetup group and Facebook group for more details and events