The Argument for Automation-at-Every-Stage-of-the-Sales-Process

The Argument for Automation at Every Stage of the Sales Process

Some sales activities can’t or shouldn’t be automated. A lot of what salespeople do depends on their human professional judgment, and besides, clients can’t foster relationships with AI enabled bots. Meetings, calls and conversations are person-to-person, whatever the industry, product and account size, and they likely always will be.

But other sales activities fit the bill perfectly.

At each stage of the sales process there are actions that are repetitive. They represent low added value in themselves, but high aggregate value. And they have to be done efficiently, in high numbers, and to a plan.

In other words, they’re perfect for automation so human sales staff can turn their attention to higher-value tasks. From the start of the sales process through lead prioritization and prospecting to setting meetings and scheduling, here are 4 examples of what we’re talking about.

1: Constructing ICPs

Before anybody picks up the phone or even starts work on a list, nailing your “ideal customer profile” (ICP) is one of the most productive things you can do that will pay off down the line. Yet, many sales organizations neglect it or don’t do it well.

There are several reasons for this, but it usually boils down to automation.

Companies look to their best customers to base their ICPs off. But they usually use far too few data points when they’re identifying them, so they get poor matches. Most companies will use just two or three data types and look for similarities based on those. But the fewer data types you use, the grainier the picture you’ll build up of your accounts, and the more prone to error it will be. You’re also more likely to generate ICPs based on ‘false positive’ matches – i.e., they look like they match, based on the small number of data types used, but they don’t really. A clearer picture would have excluded them and resulted in a more accurate ICP.

The solution to this one is simply to use more data types, but the amount of information you have to sift through gets exponentially larger as you do this. Jump from three or four data types to six or eight and it looks like you’re doubling your workload; in fact, it’s more like you’re multiplying it by a hundred because of all the possible combinations.

You can’t do this as a human being – it just gets to be an impossibly large pile of data to crunch through looking for patterns. Until automation came along, we just had to tolerate a lot of mismatches and a fuzzily-defined ICP, but now we don’t. Automating the analysis of existing data to construct a really accurate ideal customer profile based on solid correlations across a high number of data types is dry, wonky and dull. It’s as far away from making it rain as you can imagine. But it could literally double the effectiveness of every other part of your sales process.

2: Prioritizing Leads

Having a great ICP is a starting point. Once you have one, you can use it to prioritize leads. But here, you’re up against the same problem; wading through a huge stack of data, looking for the best leads according to that ICP.

dont guess how many accounts to target in account based prospecting

Image Source

These numbers are obviously fillers. But the point remains: if one SDR can prospect 100 accounts monthly, they really need to be the right accounts. This is all about laying systemic bases for success later.

Let’s say you’re getting results like these:

number of accounts to number of deals to close

Image Source

The first and most obvious way to increase revenue is simply to up the number of accounts that SDRs prospect monthly.

But that can only bring incremental improvements; maybe SDRs can prospect 130 accounts, instead of 100. At best that will yield a 30% increase in account-SQL conversion. Your 2.5% will go to 3.25%, you get an extra 10 SQLs and an extra 3 deals a year. Out of running everyone so hot that burnout is inevitable – and the quality of real sales development work might fall as more and more SDR time goes to hitting numbers on the phones.

What can we do that bumps numbers all along the pipeline by drastic amounts, without sacrificing real sales development to numbers-oriented prospecting?

Switch from ‘more accounts’ to ‘the right accounts.’

That sounds glib, sure. But it really does work.

We’ve seen open rates around 86%. Compare that to industry averages:

sales email open rate by industry

It’s nearly double the very best industry for open rates, nearly three times the computer software industry average.

Assuming that effect continues to all your other numbers, we’re talking twice or three times as much revenue. That’s the kind of increase we should be looking for: exponential, not incremental.

The way you get there is by building lists that are highly accurate with no dead phone numbers or undeliverable emails, and that tightly match your ICP. And the only way to achieve that is with automation.

It’s a question of data volume again. To achieve the fit we get, we source data from 20 commercial data sources and a proprietary database of 50 million business decision makers. It’s not just a question of access to the information; any sales organization can buy data, and many do.

But cross-referencing one giant database against another, going through a huge list and making sure that multiple data types match up to give a really close fit? That’s a job for a building full of people, or a computer.

(We use a computer.)

Without automation, this can’t realistically be done. But if you don’t do this, you’re leaving it to your SDRs to do it on the phone. They get a pile of leads that aren’t the best strategic choice for the business. And they miss out on a pile that would be.

This problem runs wide and deep in sales. When we asked a selection of sales leaders, mostly in SaaS and tech, if they were targeting their most strategic accounts, we found that only a few were.

how confident are you that your team is engaged with strategic prospects chart

Over 80% were somewhere between ‘selling completely opportunistically’ and ‘we’re covering the basics, but…’

This area represents a huge opportunity to increase the quality of the leads that are coming into the pipeline before anyone picks up a phone or dials a number. You’re empowering your SDRs and giving yourself the chance to multiply revenue, not just inch it up. But you have to automate the lead identification and prioritization process in order to do it.

3: Messaging

Few sales organizations turn their SDRs loose on their target accounts with no instructions or equipment. And since few SDRs are high-flying sales copywriters, most are equipped with templates for emails, scripts for calls and other tools to make sure they’re saying the right stuff to the right people.

The difference this makes is huge.

Personalized email templates were credited by Woodpecker.co with taking reply rates to cold email from 7% to 17%. That’s 2.4 times the baseline.

So we should definitely be using personalization tools.

Once again, though, automating messaging can deliver exponential increases in message success. What if we could see even bigger increases here?

We typically get 32% reply rates on cold outreach, against an industry standard of between 17% and 7%. That’s twice as much again – five times more effective than totally unpersonalized messaging.

There are two ways to look at automating messaging. One is to look at achieving incremental improvement by automating message delivery.

call feature automation

Image Source

Auto dialers, automated call logging, automated voicemail, and automated call scheduling are increasingly the norm in larger sales orgs and SaaS will deliver them to three-guys-in-a-room operations soon enough. So first, they’re practically table stakes, and second, they remove obstacles but they don’t create opportunities.

The other approach to automating messaging is to look at the targeting and production of the messaging itself.

We use a three-tier approach to personalizing messaging that takes into account:

  • Industry
  • Company
  • Role and individual

Here’s a good email template:

a good sales email template

Image Source

We can see how this one is personalized to the business it’s addressed to. Someone working from a list and doing some of their own research before sending the email can fill in the pain points, industry and the most important benefit they’re able to deliver.

Sure beats a lot of the cold emails I’ve gotten.

It’s the same on the phone. Here’s a good example for voicemail:

cold calling script for voicemail bob bentz

Image Source

Again, you can see the personalization based on minimal but effective research. It’s quick and you can craft it to its target.

This level of personalization is pretty basic. As we’ve seen, it delivers. But deeper personalization can be more effective.

We recommend starting with the industry, looking for pain points and common threads that everyone in that industry will care about. Then moving on to the company, using firmographic and technographic data. Again, there is a lot of this to wade through, and we don’t think a person could do it quickly and well enough to be economical.

Finally, take all that to the personal. This individual’s job role will affect how they feel about the benefits you’re pitching, so look for things that directly affect them every day or that their job success is judged on. Add in actual personal (non-creepy) information gleaned from social and other sources and you have a recipe for a truly personal message that matches what you have to offer to what this individual wants.

All of this has to come together. Miss one component, and the whole thing fails.

For instance, back when we first got funding, I got about a trillion emails and LinkedIn messages from recruiters trying to hook me up with software engineers.

I bet some of those engineers would have been a great fit. I’d have loved working with them, we’d have done great things together, it could have been wonderful.

Just one problem: I’m not the guy who hires engineers. That’s not my job.

So the timing of these messages was awesome: we had funding, we were hiring, there was never a better time to hit us up with a message about someone who could help us build our product.

And to be honest, the personalization was pretty good too. Many of these recruiters would do good research and send me a message that I’d have loved to reply to.

But all I could give them would be, ‘sorry, wrong guy.’

Everything else was well above average, but the relevancy was awful.

We also recommend getting a copywriter to write it. There’s a full rundown on our perspective on this issue here.

The point is that by using automation to crunch the big numbers, you can take this stage in the pipeline and make 100% or 200% or 500% increases in effectiveness.

We’re working on a project right now to make this more effective, through an AI tool we recently acquired, Hexa.ai.



Hexa lets us build enormous, accurate social graphs and use those to enhance our personalization. With it, we can bridge all the common connections between two email addresses, then prioritize the closest connections and base initial messaging off those.

Then we’ll drop those names and that connection automatically into the email’s introductory sentence. It’s another layer of precision to let us target individuals as accurately as possible, leveraging the power of machine learning and automation instead of trying to meet the numbers by hand.

4: Appointment Setting

You’ve managed to overcome the hurdles at the very beginning of the sales process and you’re in contact with someone at a target account who actually wants to take a meeting, or at least another call. Congratulations, you’re ready for the most frustrating waste of time in sales: finding a time that works for everyone. Also known as ‘never-ending phone tag.’

Your sales meeting is more important to your company than to your prospect. They will often change dates and times on you even after they’re set if a higher priority activity intervenes. And getting a solid time in the first place can be difficult.

By now, you’re probably sick of my answer to everything being ‘wheel out a supercomputer and science the heck out of the problem’. So you’ll be pleased to hear the answer to this one is simple, obvious, and you can just go out and buy it.

We’ll look at two choices, and talk briefly about them.

First is Chilipiper, the tool we use to set appointments ourselves. It’s extremely simple to use, and takes a lot of the back-and-forth schedule-juggling out of your hands.

intelligent calendar for teams

chilipiper.com

We chose this tool for several reasons. It comes with the same functionality as dozens of others, including the ability to have contacts grab some of your calendar without you having a whole conversation over email about it.

That on its own saves hours of non-work. But Chili Piper goes further. It lets us set appointments directly from forms, and it routes any inbound leads to the right rep based on account-based ownership – so it facilitates the relationships we’re trying to build instead of short-circuiting them.

But it also lets us schedule cadences with a click, arranges out emails for the best times for each recipient, and a whole lot more.

This isn’t futurology AI, something that might be a game-changer down the line. On the other hand, we’re not talking chatbots either. Instead, Chili Piper is a perfect fit for our philosophy of actionable, business-oriented AI that proves its usefulness. That’s how it changes the way we do business.

Take things a step further and you get Clara Labs, a virtual assistant that handles appointment setting and other tasks like a real admin assistant.

You just CC ‘Clara’ into communications and she takes over the business of setting up your appointments.

coordination email thread

Image Source

I’ve never used anything like this myself, but it’s obvious that a lot of workaday communication is going to be automated going forward. And every second that SDRs don’t have to spend chasing someone to set up a meeting is time they can spend on real sales development work.

Conclusion

Nailing automation as far into the sales process as possible isn’t about presenting prospects and leads with a bland, robotic experience.

That’s guaranteed to turn them off.

Instead, it’s about seeking incremental improvements in client experience and sales productivity by automating as many basic, repetitive tasks as possible. At the same time, businesses should be chasing big, times-two-or-more improvements in outcomes by automating data analysis and applying the results to how they prospect and sell.

If you have thoughts on other stages in the process that could be automated, please tell us – especially, let us know if you can see a way to power jumps in closed deals with machine intelligence!