An algorithm for writing the perfect email
Wouldn’t it be amazing if one of these machine learning geniuses came up with a product that generated perfect email templates for you? I’m talking templates that get 95% open rates and 75% response rates. That’s the dream!
Let’s explore this idea…
The most annoying answer in the world
Before we get into algorithms, let’s first understand what makes an email good. There is no one magical factor that will turn any string of garbage into a successful email. A good email needs good copywriting and this copy needs to follow some fundamental rules.
What are these rules, you ask? Well, it depends…
Sorry, I know that the worst and most annoying answer in the world is it depends. It doesn’t shed any light on problems and mostly just serves to further confuse you. Depends on what?!
Imagine that you’re friends with a wise mentor — the buddha of email writing. When you ask him a question he’ll almost always answer with it depends because as annoying as the answer is, it’s usually the right one. If you have a specific problem in mind, email buddha will be able to offer some solutions for different “depends” scenarios.
That’s great. But what happens when email buddha goes on holiday and you’re left trying to figure out the next problem by yourself. You don’t have the godly wisdom to know what contingencies your problem depends on. That means you’re stuck!
It depends always raises more questions than when you started in the first place. It depends always leaves your brain thinking harder when all you wanted was a neat, actionable answer.
Let’s think of one of these rules
We have to start somewhere so let’s pick a rule of thumb for writing better emails and see where this takes us…
One of the most commonly cited tactics in persuasive communication is to allude to credible sources. When you meet a stranger for the first time, neither of you will really trust each other. If it feels like this new acquaintance is trying to pitch or sell you something, your skepticism will increase and you’ll keep your guard up.
But let’s imagine that five minutes into the conversation the stranger mentions that they went to school with your boss/partner/peer. It turns out that both of them started and ran a profitable side-business and still vacation to the coast together every year. This new information won’t make you get your wallet out but your skepticism will drop. A mutual and credible connection boosts the rapport and trust you have with this acquaintance (let’s assume they’re not a con artist).
Mutual connections work really well in elevating trust and rapport. It’s no surprise that sales people have been leveraging this cognitive hack for decades. It applies equally as well in email communication so let’s use it as one of our email rules.
You’ve discovered a rule, now what?
If email buddha were your friend he’d tell you to store all of your rules somewhere easily accessible, in some kind of library of email rules (or guidelines or heuristics or whatever you want to call them). For the sake of narrative, let’s assume you’re part of the sales/marketing/growth team for some hot startup. Instead of relying on your own intuition and imperfect memory, you decide to use your rule list as a screening process to refine your business emails. In other words, you’ve got a checklist and you’ll be using it to optimise your emails.
You head into the office, ignore the foosball table and instead start pulling in some leads. You’re going to email them all and use your shiny checklist to perfect the email template you’ll use. Your list only has one rule in it at the moment — that you make some kind of reference to a credible person (or company) that your prospects will know of. Maybe you’ve worked directly with this credible person or maybe they’ve once made a convincing testimonial about your product.
You send your first batch of emails to 1,000 prospects. For the sake of this contrived example, let’s assume they all see the email in their inboxes within 5 minutes. 150 of them open the email and 15 respond.
Of the 15 people that replied, 5 decline your offer and 10 keep the hope of a sale alive. It’s time to follow up and close 10 sales! You open up your editor and start drafting an email. Before you hit send, you go through your checklist and screen your mail before hitting send.
Oh snap! It looks like you’ve got a bit of a problem. Your single item checklist wants you to reference someone credible. But you’ve already done that in your first email. What now?
Let’s break things down a bit
We can assume in this contrived example that 3 of the 10 leads are kind of skeptical and the other 7 are almost ready to buy. You could give the 7 a slight nudge with a final sales push and close those sales. If they’ve signalled that they want to buy but your follow-up response cites another credible reference, you might just scare them away. What kind of psycho constantly keeps name dropping references in every. single. email.
The story is a little bit different for the skeptics. Here it makes sense to extend the conversation by first telling them about some additional case studies. If your case studies (and existing customers) include names like Google, Disney and the BBC then that’s pretty compelling. Technically by mentioning these brands you’re adhering to the rule in your checklist.
So at the end of the day you’ve worked your charm and closed 10 sales. But only 30% of your email exchanges followed your email playbook every time. What gives? A machine learning algorithm won’t be able to discern nuanced situations like these as well as you can. So when do you stick to the playbook?
The answer is that it depends
Depends on what?!
This is the multi-billion dollar question. The truth is that nobody actually knows. Ever. If someone tells you they know and they sound very convincing, they’re just confidently guessing. Sure some people are super successful and they might be right most of the time, but that’s just them being skilled at this particular type of guesswork.
The reason why it seems that experienced communicators know the answer to the question of it depends is because their experience has been shaped into their corpus of intuition. Their brains know when and how to use certain rules of thumb, but they won’t know why they should use them — at least not definitively and not for more complex rules.
The simplified answer to this depends on what question is: context. The context and circumstances of your situation will dictate when you should apply specific rules and logic. In the contrived example above, the context was defined by the initial response from each of the prospects. But that’s not the entire story.
Life is a lot more complicated than a silly example. In reality any bilateral exchange (for instance an email thread between 2 people) has waaaaay more “rules” than just the inclusion of a credible reference. And each of these rules have different weightings that when summed up will result in a probabilistic outcome.
Confusing statistics jargon aside, even if you came up with the perfect algorithm of which rules you should follow and when to follow them, you have to remember that there might be contextual factors you haven’t considered. Also, humans aren’t exactly rational creatures. That’s why mere context isn’t the full picture of why and when you should follow certain rules or guidelines. But nevertheless, using the contexts that you can think of is a good place to start.
That’s so pessimistic. Give me hope!
Don’t lose hope just yet. There’s a ginormous difference between “you can never have the perfect algorithm” and “there’s no point in having any algorithm since none of them can ever be perfect”.
We’ve all received really, really, really bad emails before. Sometimes the sender is fresh out of high school and hasn’t sent a single sales email in their life. Sometimes the email comes from a startup founder that’s a genius coder but has never spent a day doing marketing work.
When we read these shockingly bad emails we judge the senders harshly. Idiots! But the truth is that they simply don’t know what rules to follow when crafting emails, or they don’t know when to follow these rules. Realistically it’s a combination of both.
What’s the difference between you and one of these technical founders with imperfect social skills? Experience and intuition. If you write good emails it’s because you’ve been writing them for a long time. You’ve simply got more training. That’s all.
Why bother with checklists when I’ve got intuition?
Because you don’t have a photographic memory…
The best teams in the world all use some form of checklist. UX teams use them when designing software, product teams use them when planning features, SpaceX use them when launching rockets into SPACE. You should use checklists because clarity in interpersonal communication is one of the trickiest things to master. And your mind has a fallible memory so it would be madness to rely on it exclusively and expect the best results.
I understand that some people reading this may disagree because you feel that your email performance is already fantastic. When you write, your emails seem amazing and often work. Maybe so, but try codifying your intuition and then screening your emails a few times. I guarantee you don’t always follow all the best rules you can come up with. If you try it but don’t like it, worst case scenario is that at least you’ll have some email performance documentation that you can share with the less experienced members of your team. Your best case scenario is that your overall conversion rate actually increases.
By documenting these guidelines and by using them to screen your emails, you’ll also start to understand more readily when to apply them. That’s the magic solution you’ve been searching for. Only by actively evaluating when certain emails need to adhere to certain rules will your conscious brain actively understand which contexts apply and which don’t. It’s kind of like rubber duck debugging but for email optimization.
That’s how to answer the question of depends on what?
Conclusion
The truth is that you won’t be able to create an algorithm that will generate the perfect email. I doubt even super intelligent AI would be able to do that. But with enough intelligent rules you’ll at least be able to minimise your mistakes, assumptions and miscommunication flaws. You’ll tend toward clarity and empathetic writing.
By cutting out mistakes and being able to refine your email exchanges based on contextual rules (even if those contexts are imperfect) you’ll definitely do better than you’re currently doing. By documenting your email critiquing process, both you and your entire team can grow to become smarter email communicators. Perhaps even email buddha himself!
Want to experiment with guidelines?
By the way, if you’re interested in trying this approach out with your team, HoneGrow is an email tool that was designed specifically for helping your team use and refine email performance rules/guidelines.
Originally published at www.honegrow.com.
