Take the Guesswork out of Prospecting with Data Driven Customer Discovery

Take the Guesswork out of Prospecting with Data-Driven Customer Discovery

Sales organizations struggle to prospect effectively because they don’t know their target customer well enough. Customer discovery is the process of identifying and learning about potential buyers before starting to sell to them. But it can only work if you have a deep understanding of the accounts you’re targeting. For most sales organizations that means a data solution has to be in place to handle the volume of data and deal with it quickly enough to generate usable insights: data-driven customer discovery.

Why we need customer discovery

Traditional sales prospecting is basically hamstrung by access to technology. Tools that should make everything easier are actually making sales’ job harder.

Years ago, salespeople were assigned a product to sell and a territory in which to sell it. Then they would look through phone books and trade magazines for businesses that might be interested and call them.

Later, when businesses put their information online and more communication channels became available, sales people took advantage – and prospects became tougher to reach and less willing to open up.

Locale mattered less in nearly every business. You could buy from someone on the other side of the world. And boy, did they ever want to call you and tell you about it.

Thus began a modern business arms race.

Sales roles diverged and multiplied: development reps, inbound and outbound teams, sales operations, all dedicated themselves to increasing the number of leads the sales organization could generate and handle effectively.

On the flipside, professional flak-catchers and gatekeepers turned these newly-focused reps away more and more. The standard required of even initial messages rose. The number of contacts required to get someone on the phone skyrocketed.

Why?

One simple reason: too much sales prospecting is based on guesswork.

There is no other way to do it without using massive amounts of data and applying machine learning to it. And that’s only just become available.

Customer discovery based on guesswork is expensive

Most sales prospects do not become qualified leads. And most cold calls or cold emails do not result in progression to the next stage of the sales process.

I know we’re all used to this, telling each other it’s a numbers game, encouraging ‘productivity’ and ‘loving the grind.’

And I’m the first to admit it – sales will be hard work forever. If you don’t fancy working hard this is the wrong industry, no doubt.

But our hard work… should work.

There’s some loss, some mistakes made and mismatches, in every industry. In manufacturing, they refer to this as ‘swarf.’ Every production process generates it.

(I’m not comparing anyone who didn’t buy from me to a waste product, that’s not my point here at all. Instead, I’m saying most of those who said ‘no’, should never have been asked at all.)

No production process should generate this much. Sales is something like 90% wasted effort. Across industries, we lose between 60% and 90% of all our potential deals after first contact; cold emails and phone calls typically generate no useful response.

From there we lose another huge slew every time we progress to another stage. This is a dummy sales funnel analysis:

conversion rates for close opportunities

Source

But it makes the point.

Imagine if a restaurant threw away 80% of its food before it was served to customers. Imagine a car plant that threw away 80% of the steel it bought before it rolled cars off the line.

And the funnel above starts with qualified opportunities; prospects who become leads are actually the minority, and leads that are qualified into opportunities are a minority of those. At the earliest stage of the sale process, the vast majority of effort does not produce the desired outcome.

Is this ‘just the way sales is’?

Well to some extent, yes. I think so. Sales will always be human to human and that will always involve some uncertainty, which will be reflected in the stats. Some people will always have a greater facility than others to close deals and some buyers will always be tire-kickers or slow movers.

But 80% or more? Come on. It doesn’t have to be this way.

Outbound fails when it’s not data-driven

The reason sales looks like this is the same reason that the biggest drop-off in most sales funnels is at the prospecting stage. Most prospects – the vast majority of them – do not become leads. That’s because they were the wrong prospects to begin with.

Then qualification and solution presentation are bedeviled by tire-kickers, the merely curious and people who have been hustled past a milestone by KPI-conscious reps who feel little responsibility to the whole process.

Consider that most SDRs are in their role for one year and three months, and the average sales cycle length in enterprise SaaS is around 90-120 days. A fifth of SDRs won’t even be in role when the prospects they’re working on now buy. How can they be expected to care about an outcome that far ahead?

The result of all this is that each stage of the sales process involves working with people who shouldn’t really be there, while people who should are never contacted at all. And the root cause is that sales prospecting is based in large part on guesswork. It’s educated guesswork, but it’s still guesswork.

The solution to this is to put customer discovery on a data-driven footing.

Customer discovery comes before prospecting – it’s the work that has to get done before anyone picks up the phone, to identify ideal customers and make sure the list, the messaging and the product all fit.

Customer discovery only works when it’s data-driven

There are already some great customer discovery processes out there, and we’re not here to reinvent the wheel.

So I’ll just point you to Eric Ries and Steve Blank.

Chances are, you’re already doing this – that’s why this post isn’t called ‘hey, ever think of trying customer discovery?’

The problem with the processes that Blank and Reis identified is that they work best when they’re fed good data, and good data is tough to come by.

Consider this: Steve Blank used to teach his ‘4 steps to epiphany’ customer discovery system at business school. Every year, he’d describe a startup that, after agonizing over whether to charge for its beta software, settled on a figure: $50,000.

He’d ask the class their opinions, then do the big reveal: the VP in question, Mike Maples, then of Motive, was sitting in the back of the room.

Turns out Motive only managed to find five customers willing to pony up to that $50,000 price tag.

The good news is, those customers were Netscape, Disney, CompuCom, JD Edwards and Microsoft. Motive pulled in $60 million in sales bookings in the 18 months after launch.

It’s possible to do excellent customer discovery on a small number of high-value targets. It takes skill, no doubt – but people can do it.

When you’re chasing a small number of high-value targets, finding out what you need to know about them to line up messaging perfectly is worth doing the old-fashioned way.

But the average business isn’t like that. They’re in a middle zone where they’re chasing accounts, but maybe a thousand, two thousand, five thousand of them. Or their deal size is too low for that kind of effort: if they worked that hard on all their prospects they’d have no revenue.

In that situation, it’s tough to make customer discovery work effectively without an assist from AI, because there’s too much data to get through. Many businesses wind up doing a surface sweep, and the result is the chart above: a catastrophic drop-off in the early stages of the sales process.

More than preparation: optimizing from the data your sales efforts produce

Some of that could be fixed by nailing customer discovery before the sales process gets rolling. And some of it could be fixed by feeding response data back into the sales process, improving customer discovery on the fly.

Marketers have been doing this for years, often with huge numbers of accounts, but their needs are not ours; they handle higher volume, for sure, but the personalization requirements of sales messaging far exceed anything marketers need to do.

They use software that runs on a series of ‘if this then that’ gates, so customers who didn’t open one email get forked over to another type. That makes sense if you’re delivering marketing drip emails to consumers, or running targeted content marketing campaigns aimed at businesses.

But most sales organizations aren’t set up for that.

The solution to actually making customer discovery work for the average sales org, especially in SaaS or midsize deal scenarios, can’t be the if-this-then-that tools marketers are accustomed to either.

It has to be AI, to handle the volume of data and do it at the speed that’s required. And it has to be committed to collecting that data, at the start and all the way through the process.

Let’s consider sales email subject lines. (This must be one of the most written-about subjects in the world: a search returns around 3.5 million results.)

Top tips aren’t what moves the needle, though.

Instead, email subject lines need to be carefully matched to the recipient’s needs and the way they naturally communicate.

Once those emails start actually being sent, though, they’re generating data. Use that to inform iterative redrafting of those email subject lines and you can take a campaign that’s already successful, and target it ever more closely to both the real needs of your recipients, and the way they prefer to be communicated with.

We know this approach works. That’s why we A/B test all messaging – email subject lines, body text, length, messaging, and everything else we can think of.

While we’re testing messaging, we’re validating our customer discovery work, ensuring that the ICPs that we identified are accurate and always working to improve them.

Conclusion

Customer discovery is necessary, but it doesn’t work without data. To be effective, customer discovery needs to be driven by data both before the sales process begins and during it. The data that the sales process generates needs to be fed back into planning and used to identify potential improvements. There’s no way to acquire and manage this kind of data volume without using AI unless you’re working with a very small number of accounts.

What are your biggest customer discovery challenges – and how have you defeated them?