Getting beyond the basics of using CRM— tracking the evolution of the customer relationship—means using the system to improve time management close rates and customer satisfaction. In the real world, that means playing the probabilities and using the system to prioritize your team’s efforts.
A wide range of business decisions in marketing, sales, services and even engineering should be guided with inferences from CRM data:
- Which social network or advertising medium is going to have the best yield?
- Which leads should be sent directly to reps, and which should be held back for nurturing?
- Which prospect deserves an on-site sales call or a pre-sales engineer?
- Which sales territory should get that loaner equipment for a demo?
- Which deal is worth the investment of a proof of concept?
- Which deals will close this quarter, and which will get pushed out?
- Which customer is most at risk of cancelling?
- Which features and bugs will have the biggest payoff?
- Which customers are most likely to buy again…and when?
We all have to make these decisions with incomplete information. Even the Harvard Business School characterizes management success as the ability to make better bets.
You don’t have to have a fancy MBA to increase your odds—but you do need to have a better way of knowing the score. (Before we start, though, it’s important to understand that there’s no single score: There’s a different score for each CRM use-case, and the scores should be changing on a monthly, if not weekly or even daily, basis.)
First Quarter: Using CRM for Pre-Pipeline Scoring
This is the noisiest area of innovation at the moment, as everyone in the social CRM and marketing automation markets races toward El Dorado. Most of the scoring work focuses on four things:
- Lead prioritization (which leads need to be worked right now)
- Routing (whether the lead should be handled by the inside development team, inside sales or the field)
- Nurturing (which vertical marketing campaigns it should be part of)
- Qualification (which leads have matured to MQL status).
How-to: Use Social CRM for Revenue Production and Support and Services
To that end, there are four main scoring mechanisms for scoring and managing leads:
- Profile (explicit criteria). Who is this person? In particular, does he work for a company that’s already a customer? This score changes only when you are using progressive registration and external data enrichment, both of which are best practices.
- Behavior (implicit criteria). What has this person done? What has she responded to? In particular, has she personally purchased from us before? This score changes almost any time the lead takes an action. Having a carefully documented and measured scoring algorithm is a best practice here.
- Temporal (decay criteria). How long has it been since the last action? In many U.S. market sectors, this is an amazingly important factor, as lead interest is highly perishable. In markets suffering less from ADHD, the time factors can be calibrated in weeks rather than hours.
- Social criteria. Who does this person influence, and who influences her? If you’re lucky enough to have access to detailed communication network patterns, there is significant predictive value here. This is an area of truly Black Magic, so watch out for overselling from vendors and consultants alike.
Second Quarter: Using CRM for Pipeline Scoring
Once a prospect has engaged in a sales cycle, the most visible scoring in the CRM world starts to happen. It’s called the forecast, and it has a raft of its own political and organizational issues on top of accuracy.
The scoring criteria for individuals can be extended to deals:
- Profile criteria apply to both the company (knowing it’s in your target market and has the budget) and the individuals (namely, an executive champion).
- Behavioral criteria should focus on the actions of the prospect, following models you have developed for the type of transaction—new vs. upsell vs. renewal or repeat—as well as the type of customer, distinguished by segment such as vertical industry or company size. To be relevant, behavioral scores much be based on explicit prospect actions such as recorded actions or survey responses, not the self-interested suppositions of the sales team.
Commentary: When Your Customers Vote, Does Your CRM Listen?
- Temporal criteria are hugely important to the forecast; a deal delayed is typically a deal denied. The opportunity score should be decremented if there is no activity or update in the last two weeks, if the close date moves out or if the deal stage goes backwards. Client actions that almost certainly mean a delay, such as changes in the deal team on its side, should pull the score down, too.
- Social criteria have the benefit of being observable during the sales cycle. It’s difficult to get credible numerical scoring on the quality of the client’s deal team, but the really sharp sales rep can provide a realistic estimate of the win. Getting that information, however, requires carefully designed incentives for and measurements of forecast accuracy.
Third Quarter: Using CRM for Customer Scoring
This is the most often ignored area of scoring, but it’s also the area that has the best chance of being accurate and the biggest payoff in profitability. There are three reasons why.
First, in almost any industry, the marketing and sales cost involved with a new customer acquisition is between five and 10 times the cost of getting an existing customer to buy more from you. While the deal sizes of re-up business may be smaller, the really profitable sales happen over the life of the customer relationship, not during the initial sale.
Commentary: Top 10 CRM Tricks Guaranteed to Lose Customers
How-to: Use Big Data to Stop Customer Churn
In addition, with an existing customer, you have the most credible sources of data coming naturally out of your service, support, consulting and training organizations. If you harness the data properly—and use automated Web-based surveys as often as you can—you have an almost continuous stream of signals from the customer.
Finally, you can construct models of customer behavior that involve fewer assumptions, shorter extrapolations and more interpolations. This means you can tune the scoring algorithms better, as you have tighter feedback loops.
Fourth Quarter: Using CRM to Set Expectations Around Scoring
Scoring systems are relatively easy to set up, but it’s difficult to get really good results from them. Technology is not the problem here. Rather, scoring must be based on a model and a set of assumptions, and nobody has a solid model for how individuals make purchasing decisions. If somebody did, then you’d never see mass-market advertising, spam email, or pot-shot marketing campaigns.
The modeling errors from a single bad assumption or weak relationship easily overwhelm the predictive power of a dozen good elements. Expect to tune your scoring models for many months, if not quarters, before the results are credible. Because of this break-in period, it is critical to set expectations low, particularly with executives and sales people who will really only give you one chance.
Post-Game Analysis: Things That Shouldn’t be Scored
The whole point of scoring is predicting prospect behaviors or customer decisions before the outcome can be measured. The contrasting approach is testing, or measuring what actually has occurred and drawing conclusions from correlations of historical data. There are several ways to do this:
- Evaluate an A/B test to see which Web page design has a better yield
- Compare click-through or conversion rates to see which message is more effective
- Analyze the pipeline over the last 12 months to find out which market segment has the shortest sales cycle
- Summarize pipeline-creation rates from last year’s shows to learn which tradeshow generates more pipeline
- Report on closed deals, by product, to determine which product version sells better
How-to: 10 CRM Mistakes and How You Can Avoid Making Them
While these items could conceivably be modeled and scored in advance, there is no point in using scores—which are, at the end of the day, a “best guess”—when the CRM system has enough data for a solid extrapolation.
As always, there’s a key caveat: Analytics will inherently magnify any data quality problems you have. Be sure about your semantics and data filtering before you make critical decisions based on either scoring or analytics.
David Taber is the author of the new Prentice Hall book, “Salesforce.com Secrets of Success” and is the CEO of SalesLogistix, a certified Salesforce.com consultancy focused on business process improvement through use of CRM systems. SalesLogistix clients are in North America, Europe, Israel and India. Taber has more than 25 years of experience in high tech, including 10 years at the VP level or above.
Follow everything from CIO.com on Twitter @CIOonline, on Facebook, and on Google +.