Want to send human sales messages? You need a robot.

Sending human-sounding emails and other sales messages aren’t best done by robots, obviously. But it’s actually not best done by humans either. Ideally, both are used.

There are two parts to winning sales messages: preparation and performance.

Muhammad Ali famously observed that he had never won a fight in the ring. Instead, he said, ‘the fight is won or lost far away from the witnesses, behind the lines, in the gym, and out there on the road; long before I dance under those lights.’

We agree: without preparation, performance will fail.

But a drawerful of notes isn’t a novel, a couple of demos isn’t an album, blueprints aren’t a house. In the same way, insights and ideas don’t come together and make fantastic sales messages with results double or triple industry norms. Preparation isn’t enough by itself.

So how do we ace both these?

Each is suited to a particular approach. The preparation requires us to identify the right prospects and learn a huge amount of information about them so that we can later use that information to inform and structure our sales messages.

It’s far more effective to handle it as a single large task. There’s a lot of data to be worked on, and spotting patterns in very large volumes of data from multiple sources is extremely difficult for the human eye, so we suggest using automation (you can read more about our recommendations here).

In fact, most lead providers offer a variety of tools to search for leads in their databases. 

But assuming that you have all the information you need at your fingertips, how should you then create effective, resonant sales messages?

Data is the key

The key is to personalize based on that data that automation delivered.

Typically, sales messages are personalized based on fairly superficial knowledge: name, company, industry, website. Even really good examples like this one:

(Source: http://blog.topohq.com/sales-prospecting-emails/)

This is a good sales email. But it relies on material taken from the target’s blog, as well as the author’s skills as a copywriter.

The personalization in this email – name, industry, website – doesn’t cover the recipient’s company in any detail, but otherwise, it’s pretty on the nose. Without strong support from a broad base of data, it’s tough to do better than this. Check out the advice Yesware gives their readers: It’s not enough, they say, to use custom company and name fields.


So what should salespeople do?

Find uncommon commonalities, highlight mutual connections, send at the right time, talk about something they did that you liked. Use their name more.

That’s all going to work better than nothing, but not as well as having deep knowledge about exactly what their specific pain points are, what their company needs right now, and their (publicly available, non-creepy!) personal information.

Who writes the email?

We’ve established that we think automation is best for gathering data and deriving insights from it. But who should write emails?

Right now, the typical method here is to have SDRs writing emails. This is sometimes a problem because many SDRs do not have copywriting expertise and the use of language can easily push a possible sale forward or cease discussions.

Templates are often worse because they become so common they are easily ignored or identified.

In reality, copywriters should write emails, while SDRs are get training.

How should they write those emails? By personalizing them based on deep understanding of several key areas.


Write for an industry or a type of company. Different industries use different languages. They talk about different things or they talk about the same things using different terms. We have to talk to them in their language. It’s crucial to get this right – and it’s not that easy.

Salespeople often default to talking about the product or the company they’re representing because that’s the language we speak. It’s tougher to see outside that to the way people naturally communicate with each other in another industry.


Score 1 for each acronym you understand without having to stop and think:

  • AE
  • SDR
  • CRM
  • BANT
  • ICP

If you work in sales, you should have a perfect 5.

Now try these:

  • ADC
  • ECS
  • LTP

Any luck?

Unless you work in the aerospace industry, probably not. Somebody else’s language.

Learning someone else’s language is a classic copywriting challenge. Take a look at this guy:

(Source: https://giphy.com/gifs/hello-steve-buscemi-30-rock-ifxLK48cnyDDi)

Yeah… let’s not.

Obviously, it’s not just a matter of particular words or phrases: a different industry or type of company will have its own colloquial styles and references, its own calendar, its own heroes. Being in the same industry isn’t enough – Tesla and General Motors both make cars, but they’re different types of company.


Each individual business is different. Writing for an account is more than remembering to paste something in place of {company} in the email. This is where things can get tricky, and the relationship between the copywriter and the data that’s used becomes paramount.

To write for a company, you’ve got to be able to address their culture, history, the business pain points that they have right now. This isn’t a question of ‘more persuasive copy,’ in the sense of tricks or gimmicks in the copy itself.

It’s about putting a solid, skilfull copywriter in the same place as extensive research and letting them turn that knowledge into a message that’s genuinely tailored – not off-the-peg with adjustments.

To achieve that, you’ll need:

  • Technographics: What does their tech stack look like? When did they change it, do we know why?
  • Hiring info: Are they hiring? Are they growing or is their company smaller than last quarter? Again, do we know why?
  • Revenue info: What are they bringing in, per customer and in total?
  • Firmographics: company history, location, development.

And you’ll need a copywriting team who can sit down with that information and use the prospect account’s own language to talk to them about their business.


The final level of personalization. Sometimes it feels like people try to jump here with guesses about sports affiliation or quips about the weather in Seattle without covering the business ground of the first two layers of personalization. The result can be a little light on genuine reasons to respond or even keep reading, as well as feeling desperate.

There’s a lot more judgment call involved when you’re personalizing right down to the level of the individual than when you’re dealing with a company or industry data. You’ll need:

  • Title: What’s their role called? This changes from company to company, and sometimes you’ll find someone with a title you never heard before. That’s often a clue to company culture. A business with a ‘Head of Customer Insights’ role is paying attention in that area.
  • Responsibility: What does this person own? Who are they responsible or accountable to?
  • Headcount: How many people do they work with, how many report to them? Title isn’t always a good indication of this. But it’s crucial for understanding their pain points and getting insights into their needs and likely attitudes.
  • Interests: Many people are OK publicizing their off-hours interests. If they’re made public, we can find out about them and add that information to the mix.Doesn’t have to be done in a clumsily overt way, but if Sandra is a massive ice hockey fan she’s going to respond a lot better to that metaphor about skating to where the puck will be.
  • Challenges: This is getting into the business nitty-gritty. You have industry issues and issues the company faces – but what challenges and pain points do these individuals face in their work?From a sales perspective, look at how AEs and sales managers view CRM: AEs hate adding data to it, managers hate that there’s never accurate, up-to-date information there.Both struggle with it but for different reasons depending on the role. If we’re to communicate with them we need to address the pain points which they personally encounter.
  • Personality: While there’s discussion about how accurate ‘personality type’ measurements are (aren’t there really 6 billion plus personality types, anyway?), we can use insights into how outgoing, open, curious and so on a person is as ways of understanding how best to approach that specific person. Highly extroverted people and highly conscientious, introverted people will have different preferred ways of communicating, and we need to be as in tune with that as possible.

The Knowledge Gap

Ever see the trailer for a movie and just think, ‘I just saved nine bucks?’ In the rush to show you how great the movie is they showed you… the movie. (It’s worse when you pay the nine bucks to find out that what they actually did was show you all the good bits.)

We don’t want to leave prospects feeling like they know everything there is to know about us and what we can do, and neither should you. You want to leave enough of a knowledge gap that they feel like they have enough information to decide to take the next step – to respond, to pick up when you call – but not that the story is over, and they know how it ends.


Generating the data and insights required to do this is a task best given to AI, but no machine can take that material and make effective sales messages out of it.

When we give SDRs templates to work from we’re combining the worst of both worlds – a rigid, formulaic approach and a person who isn’t a copywriter and may not be the best at expressing themselves in emails.

So at OutboundWorks, we take the insights our machines hand to us and let a great copywriting team loose on them. Machines can’t do it, and people – even great people who specialize in it – can’t deliver really effective sales messages without machines. It takes both.

How do you craft your sales emails – and how do you think the ideal sales email gets written? We’d love to hear your views, so leave us a comment. We read them all (non-automated, promise)!

Let us help you meet sales quotas by putting our data and copywriting skills for your team. Request a demo today.