Sales can be a tough and unpredictable place to work; can artificial intelligence give sales reps a better way of managing their leads and give their bosses a better way of forecasting?
Clari's predictive sales management system aims to help companies identify the right deals and risks in their sales pipeline and is used by companies including GE and Audi. ZDNet spoke to the company's CEO Andy Byrne to find out more.
ZDNet: Tell me a little bit about yourself and your company?
Byrne: Clari was born in 2013. Before we started Clari I had been involved in starting a company called Clearwell Systems back in 2005. That was doing machine learning on large volumes of enterprise data to allow you to respond to a securities investigation or a litigation issue.
And we grew that company until it was doing $100m in revenue and Symantec acquired us. That was in June of 2011.
I wanted to bring that up because we have been doing machine learning since '05. We called it predictive coding back then.
My co-founder, Venkat Rangan, came to me and said let's start another one but with the thesis that machine learning was going to be really, really big in Enterprise Software.
It became very clear that the place that we would start was in sales, because sales was really broken. They had good accounting systems to pull records into a database but that was really all it was. It wasn't helping rep sales managers to drive more revenue and executives to forecast accurately.
Here was a $50bn market where all of the customers that we talked to were super disappointed. Poor data quality was problem Number One. Number Two was blind spots where managers didn't know what was going on. And Number Three was the forecasting process.
It was all fundamentally broken. They were pushing stuff out using Excel spreadsheets. It was just brutal. They were really only looking at one signal and that was CRM data. And that data was manually entered by sales people and they hated doing that.
That was the pain that we saw and we realised that we could build a new build management platform that could use machine learning at its core.
That was what we set out to build.
And now we are starting to see the huge shift that is happening in sales with machine learning.
In terms of spend, we can say that two years ago AI was an interesting experiment for sales that CIOs were sort of considering. Now it's one of the top areas, if not the top area of spend.
How's it been going since then?
If you look at the math behind our customer base then we've nearly tripled it year over year. If you look at the spend in artificial intelligence in the enterprise -- JP Morgan pegged it today at $12bn -- you can see that it's predicted to grow in the next three years to just under $60bn in spend.
We predicted back in 2013 that machine learning was going to be big. I'm not saying we're clairvoyant, we just got lucky. And here we are, a small company that suddenly becomes a power company. You're always looking for a trend that you can ride and there is no question that that is there with artificial intelligence.
What sort of applications do you see driving this?
Our contracts go back and forth. There's a bunch of signals in there. If you add that signal to the signals that you find in account records and CRM and you marry additional marketing signals like Marketo -- we're analysing all the Marketo data -- and add to that support for DocuSign and these are super-interesting features that you can use in machine learning.
You're identifying where you have risk across tens of thousands of deals. How does that manifest itself into an application that's in the sales management platform? I'll give you three examples.
Take the sales rep. Instead of logging into the CRM system that's a poorly designed system that is not really helping them sell, they're looking into Clari and Clari is reminding them where they have risk, where they see risk and where they haven't spent enough time. We prioritise the actions that they should take to close their deals faster.
Now look at the sales manager. He's got ten reps and they've each got ten deals. That's 100 deals. Before Clari, you were logging into Salesforce and you had no idea what the reps were doing. They are not going to manually enter all of the things that they do. They're guessing.
Now, with Clari, Clari immediately shows you where you have risk. You can see that you have some deals with some risk and they look super-risky. It could be that the customers are not engaged in the way they should be but Clari helps them and then they can drive more revenue.
In the third area, think about the executive and the nightmare of forecasting in Excel-hell. They're rolling up all these Excel spreadsheets or Google sheets into a poorly designed tab in their CRM and it doesn't scale.
They log into Clari and Clari is showing them things like their goal is 100 million this quarter but they're actually doing 120 million. Or their team is needing to do 160 and Clari is calling 125. We help them to see that.
And then we suggest things like, here are the areas where we see risk and you can take action to mitigate that and you will then make your number.
You've got a wide range of customers covering a wide number of business areas. How did it start and in which areas did it build up from?
What you are seeing is a movement that is happening among chief revenue officers, and among VPs of sales in that they need to graduate from a CRM system and Excel spreadsheets and BI reports to a new way of running their sales teams with a predictive, sales management platform.
It depends on your time horizon. For us, we're laser-focused on predictive applications for sales teams. The amount of opportunity that's there is vast. All of that pain within sales is what we want to deal with.
Going forward what you will notice about us is that not only are we looking at opportunity management, pipeline inspection, forecasting, that's our core area but there's other areas where we are starting to get traction. As we extend our platform and our reach and the signal that we support, we're analysing Marketo data, it's super-interesting for us to combine both sales and marketing signals into one platform. That's never been done. They've always been separate systems.
Ultimately, it's not about top-of-funnel marketing sales data and middle- or bottom-of-funnel sales data. It's all just about revenue.
I'll give you an example. Take Okta: we were looking at the user data and we saw a bunch of users that were getting into the system and we thought who are these people? And then we realised they were not even in sales. They're in finance, they're in marketing, they're in HR, they're in operations.
So we're looking at their sales hires and we can predict, within six months, who's going to make their number and who's not. So all of a sudden, human resources and recruiting has a good sense of how good they are at hiring. Or where they're not, we can show them where they are going to have to make changes to improve things.
How better software could improve your sales operation