Hey Numbers Cruncher: Generating Analytic Value | SalesAndMarketing.com
LinkedIn  Twitter  YouTube  Facebook
Share |

Hey Numbers Cruncher: Generating Analytic Value

"Tell me something I don't know"

Even though everyone's had to work harder in this economy, those who work in business analytics can be forgiven for thinking they're being cut more slack. After all, they're part of the solution, not the problem: they produce analytics to help their companies save and make money.

And yet, that hasn't always been enough. Some companies have laid off entire analytics business teams in the recent downturn. So, just like everyone else, analytics teams now must dig deeper into their bag of tricks to find more valuable nuggets in their data troves.

From Rear-view Mirror Reporting to Forward Looking Guidance

In order to contribute to the business as directly as possible, data analysts must generate value that goes beyond reporting on rearview mirror trends of sales and marketing activities. Yes, the typical cycle of producing knowledge with analytics, making gradual adjustments, and then running more analytics, has been very lucrative when done right. But there is a need to shortcut and generate value more quickly these days. As such, it is time for predictive and event-based customer analytics that can guide upcoming sales and marketing activities more directly.

Both predictive and event-based customer analytics enable companies to discover opportunities for growing business with individual customers—opportunities neither slicing nor dicing could tickle out of the data warehouse.

Predictive Analytics

Predictive customer analytics come in many flavors. Probably the most "salesy" application is to use predictive modeling solutions for scoring how likely each prospect is to make a purchase in a particular product category. The models can also predict how much the prospect will likely be spending.

Direct marketers employ these predictive scores to decide who they are going to prioritize for contacting and what they will pitch when doing so. This kind of predictive prioritization is a necessity for direct marketers. If marketers simply contacted everyone on their list, their campaign would be so costly it could never break even.

Besides focusing sales and marketing attention on the most lucrative prospects, predictive analytics guide finer decisions, too. For example, a prospect who seems valuable but is still on the fence may be prioritized for incentives to move them over the edge. Yet, a better prospect that is already very likely to make a purchase may not be offered any incentives.

Event-based Analytics

Event-based customer analytics, on the other hand, could be called a special kind of predictive scoring. Event-based analytics solutions sift through detailed transaction and interaction records on a daily basis to flag prospects and customers that exhibit significant behavioral events or changes to their normal behavioral patterns. The idea is to enable the sales team to get in front of prospects as soon as possible when such events or changes occur, because they suggest that the individual may be in an active buying mode or at heightened risk for attrition.

You could say that event-based analytics predict that these customers are high-opportunity targets for sales and marketing efforts, right at the moment. Two or three days later it may already be too late, as the individual may have made up their mind otherwise.

For example, SunTrust Banks has used event-based analytics with tremendous return on investment in order to prioritize customers for a helpful customer services call, when warranted. Cell phone carriers employ event-based analytics to trigger retention marketing offers when a customer's use of their cell phone is trending down in concurrence with other signs of attrition risk. An online stock brokerage may trigger cross-sales efforts to customers who view the broker's Web pages on options trading multiple times in a week but aren't yet authorized for trading options.

How To Go About It?

Here are the top five recommendations to sales and marketing leaders to help them guide their predictive and event-based analytics efforts in the right direction:

1. Ask for behavioral data to be prioritized over demographics data. Behavioral data is needed as input into your event-based and predictive analytics. If you are a retail banker, you can analyze the frequency and amount of bank account transactions. At a telco, you can analyze call center records. As a retailer, you can analyze the shopping behavior of your top customers.

And yet, the richest source of behavioral insights is often forgotten: your customers' interactions with your Website. Short of a live conversation, few things give greater insight into a customer's current interests than their interaction with your site as captured by your Web analytics solution at the individual visitor level. For example, when customers type in search query keywords on your site they even tell you in their own words what it is that they are looking for.

2. Don't wait for a 360-degree warehouse to be built. Seek value that can be generated quickly with the behavioral data sources that you already have. Don't think that you must first create a complete data warehouse. Event-based marketing software solutions work with your raw data streams. And Web analytics data along with registered online customer information is a far narrower slice than the full 360-degree pie. Yet, it is very predictive of customers' current propensities.

3. The data won't always reveal what offer or product interests a customer. As the bankers at SunTrust know, for example, the behavioral events or changes often don't reveal what the customer's intentions are. But when getting in front of customers with a helpful services oriented call, at the right time, they will invariably explain the underlying causes.

4. Pair up business minds with data geeks. The wonderful thing about analytics geeks is they're able to find all sorts of unexpected insights in data. But that is also the problem. Most nitty-gritty details are simply too much for the business to translate into action. There is a perfect symbiosis for business minds to guide the data expedition into a direction that makes business sense.

5. Think twice about how you will automate predictive and event-based analytics. While manual or homegrown solutions for predictive and event-based analytics are conceivable, most mid-sized and larger companies have found dedicated software applications saved them a lot of effort. Even more importantly, they've generated positive ROI within weeks instead of months or years.

Akin Arkin is director of Internet marketing at Unica Corporation (www.unica.com).