Q&A: Analytics Ushers in New Era of Data-Driven Sales Management
Combine data from a compensation management system with the power of analytics, mix in the right algorithms, and you have a prescription for far more powerful sales performance management.
- By Linda L. Briggs
- August 4, 2015
Compensation management systems typically store vast amounts of data about sales performance, particularly about compensation. Through analytics and good algorithms, companies can do a far better job of managing their sales teams, especially top performers.
In this interview, we talk with Giles House, executive VP with CallidusCloud. The company has integrated its cloud-based sales performance management platform with its analytics engine. By analyzing historical sales performance data, House says, you can train sales representatives to adopt more effective selling behaviors.
"Better yet," he adds, "you can give them incentives to encourage those behaviors. However, before you can do that, you need to understand how skilled each sales rep is and how behavioral changes will affect their performance. That is something a predictive platform can help with."
BI This Week: What is sales performance management (SPM) and how does it tie into business intelligence?
Giles House: SPM is a blanket term for a whole set of software tools aimed at improving the performance of sellers -- not just in-house sales staff but also channel partners. Those tools start with hiring and onboarding sales staff and extend through coaching, training, territory and quota management, gamification, and incentive compensation. Essentially, they deal with all areas involving hiring, training, and incentivizing your sales staff.
From a business intelligence perspective, compensation is a vital and often-overlooked tool in managing your sales and sales staff. Your compensation system contains a wealth of information about the performance of sales staff and can be examined with various variables in mind -- for example, examining the performance of different sales representatives when they are assigned the same territory. Are poor results the fault of the reps or is the territory drawn wrong or is it simply a tough territory to sell to? The answers are captured in compensation.
We talk a lot about analytics at TDWI. How can past data be used to predict future trends?
The past performance of your sales staff can be understood from the data captured in your sales compensation system. Teasing this data out, you can see macro trends -- the effects of new products, competitors, and even economic trends -- that affect the entire sales staff to a similar degree. Then you can zoom in on individual performance and see how your individual performers have fared. With enough data, you can develop a variable for each sales rep base on past performance.
Now, apply this variable to expectations for the near future -- the expected change in performance caused by a new process, new technology, fresh training, or whatever changes you plan to make. Those sets of algorithms will provide you with a window into what future trends should look like.
So the behavior of individual sales people can be used to predict future sales performance?
Sales reps do what they do because of their motivations -- which are primarily incentives offered in the form of commissions, bonuses, and other financial mechanisms. A hierarchy exists in almost every selling organization, and the behaviors they display in reaction to existing incentives tend to remain constant. There are generally A-, B- and C-quality reps; the A reps already operate in an optimal fashion, and the B reps can be brought up to A level with the right tools and training. The C-quality reps are often targeted for replacement.
How can sales performance be improved using that information?
First, it's important to recognize what you're trying to improve is not the number of deals but rather sales rep performance. Those are two different things. Performance is a set of behaviors you hope to instill that will lead to more sales, yes, but conducted in a manner that sets in motion a longer-lasting relationship that ultimately has far greater lifetime value.
Getting a deal at all costs can cost you because in the subscription era it takes roughly three years for a customer to become profitable, so with historical data on sales performance, you can coach reps to adopt more effective selling behaviors. Better yet, you can give them incentives to encourage those behaviors. Before you can do that, however, you need to understand how skilled each sales rep is and how behavioral changes will affect their performance. That is something a predictive platform can help with.
What are the challenges of producing, combining, and analyzing all of that sales data? Why hasn't this been done sooner?
The principal challenge has been that the data had to be collected in the compensation management system while the analytics were done in another application. There was a layer of data migration and conversion between the two that made it very cumbersome, and the time delay between the creation of the data and the analysis of that data mitigated its usefulness.
Also, it isn't easy to develop algorithms to predict outcomes around sales; we worked at the problem for years before we were comfortable that we were on the right track and felt we could start to talk about a Predictive Sales Performance Platform.
What sorts of analytics can be used to produce good results?
It depends on what you're looking for. There are metrics about lead conversion rates that indicate how well the sales team is working leads and how good the leads from marketing are. There are metrics around selling price and margins that can tell you how effectively the sales team is using discounts. Also, there are metrics for customer satisfaction and customer experience that can help you predict customer lifespans and prevent churn.
Actually, the question is not what sorts of analytics can produce the best results. The question is: what do you want to change in your sales organization? The analytics exists and is readily available, but first you need to know what you what to change.
What does the advent of big data bring to the sales performance and analytics picture?
Big data simply means we have more to examine, more to work with, and more insight locked into our various sales management systems. What we really should be excited about is the advent of the "big analysis" era. We've had the data for a long time, in many cases, but we've lacked the analytical tools to derive insight from it. Sales has been one of the last areas of business to automate, but the convergence of big data and a desire to understand sales data in a more meaningful way has been nearly simultaneous. It promises great things to the businesses that invest in it.
How does what we've discussed here tie back to CallidusCloud?
In April, we announced the Predictive Sales Performance Platform, which combines our analytics engine Thunderbridge with access to the data stored in our CallidusCloud Commissions product.
We see this as a massive help to businesses in several areas, starting with forecasting. Most forecasting today is very much a guess, and 70 percent accuracy is viewed as successful. With the PSPP, that number goes up dramatically, allowing businesses to plan far more efficiently. We also see it as crucial in helping our customers change sales behavior -- switching from behaviors that simply close deals to behaviors that earn the customer's business over the long term.
The PSPP can forecast the outcomes of the behavioral changes and reveal those changes down to the level of the individual sales person. That provides opportunities for planning, coaching, and other investments. This isn't just speculation. We've used real data, usually 18 months' worth, then backed out the last several months and run it through the PSPP. The resulting "predictions" mirrored reality extremely closely. The PSPP will be available toward the end of this year, and we think it's going to usher in a new era of data-driven sales management.