How Data Mining Can Help You Dig the Nuggets Out of Your Business
Imagine having a rare and valuable art collection and keeping it locked in the closet. Now think of data practices. Same situation. Marketers spend millions of dollars a year to gather and store large quantities of data that are never seen by the human eye.
Other companies spend a great deal of money to run reports, graphics and analysis to show to marketing and sales personnel, but are unable to transform these expenditures into day-to-day actions.
Data is merely words and numbers without meaningful actions attached. That is why data mining—extracting the knowledge from large chunks of raw data—is absolutely critical for any organization that collects data about its customers.
How Do I Get Started?
The pillars necessary for any company to transform data into actionable decisions are executive support and a data-driven culture, evident when, on a regular basis, business action is taken as a result of data (reports). Executive support is essential to attain the financial backing and build the necessary infrastructure.
However, it’s important to start small and be very specific in what you are trying to accomplish. Do not let technology drive your decisions. It is better to invest in a temporary or smaller solution and throw it away when you outgrow it, rather than invest big and get lost in the options and complexity before you even get started.
3 Keys for Building a Data-driven Culture
1. Business Objectives. Define the objectives you are trying to achieve, e.g., increased market share, reduced cost or increased salesforce effectiveness. Break each objective down and determine the elements that impact your ability to meet the business objective.
For example, suppose you wanted to increase sales-force effectiveness. Your impact elements might include:
• Field reps need more face-to-face time.
• Field reps have too many accounts.
• Field reps need more qualified leads.
2. Measurements: How will you know that you are successful? Set a baseline measurement for how you are doing now. Define what data you need to determine how to measure where you currently are, and then decide what will determine your success. If you do not have the data to determine where you are, get it!
When defining your measurements, make sure you decide who needs to get the information and in what form. That will help design your reports, and determine the frequency of distribution and how the data will be used.
Don’t collect data you won’t use. If it seems like too much information to capture, narrow it down to the critical success factors.
3. Discipline (Tracking). Garbage in is garbage out. The only way to make sure you have high-quality data is to use it. Make the measurements and reporting visible within the company. The CEO should be reviewing the measurements with his or her team at least monthly; managers should be using it to drive performance and make decisions on what to do differently. Set the standards for what data you need to collect and your expectations for compliance, and then use and post the data.
What’s in It for me? The Benefits of Data Mining
2003’s IDC study, “The Financial Impact of Business Analytics,” reinforced the fact that implemented analytic applications (data mining) have returns ranging from 17 percent to more than 2,000 percent. The median ROI, over a 5-year period, was 112 percent.
Many operational and decision-making components of the business can be vastly improved by applying the learnings gathered from data analysis. For example, business analytics can help:
• increase sales force effectiveness;
• build sustainable relationships with customers/channel partners;
• increase ROI for all marketing/channel partner campaigns (reduce marketing expense);
• reduce decision cycle time; and
• increase product penetration or market share.
How Will My Data Improve My Business?
Your data won’t actually “do” anything. But it definitely will provide you with answers to key questions, as long as you know what questions to ask. So what questions should you be asking to achieve benefits? It depends on your goal. Here are some questions to ask if you want to:
Increase sales force effectiveness:
• How many leads are in each step of the sales process?
• How long does it take to close a sale from inquiry to final disposition?
• How many face-to-face visits are my field reps making?
• What is the dollar volume of the lead in each step of the sales process?
Build sustainable relationships with the customer/channel partner:
• What customers have stopped purchasing from us?
• What customers have significantly decreased their purchases with us?
• What customers have significantly increased their purchases with us?
• Who are our top performing customers?
• What products are our top performing companies buying and what else should they be buying?
Increase ROI for all marketing/channel partner campaigns:
• Of the number of inquiries generated, what percent turn to closed sales?
• What is the cost per inquiry for each type of media used?
• What are the conversion ratios for each media used?
• Which marketing campaigns had the highest conversion ratios and ROI?
• What markets had the highest response and close rate?
Reduce decision cycle time:
• Within a given market what is the best customer profile?
• What markets are the most profitable?
• Which products are the most profitable?
• Which top customers are purchasing less than they did last month?
• What is the sales lag for introducing a new product?
Make the Data Actionable
Once you have the data, put what you’ve learned into action by adjusting processes and tools for implementing the initiative. You need to think through—as a result of the data—what you should be doing differently, and develop a plan to execute.
For example: Data shows that a company’s top 10 percent of accounts represented 90 percent of its revenue, and there was a significant gap in average order sizes between the top 10 percent and the next 10 percent of accounts. The data showed that the top accounts were purchasing, on average, 12 different products per year, while the next tier were purchasing 2.7 products per year.
Business Objective: Increase product penetration and increase revenue/profit.
Measurement: Increase the number of products tier-two customers purchased by 100 percent.
Solution: The company built a “Playbook” for the sales reps (see p. 54), delivered via e-mail, that showed what products customers were purchasing, and using a predictive model, brought up suggestions of other products they should be buying.
This initiative increased the average number of products purchased by 179 percent, and spurred 26 percent incremental revenue growth.
All implementations should be deployed in a phased approach; concentrate on impacting one business objective at a time. Standardize one program at a time until the team can measure results and effectively deploy the process.
You must be able to tie the actions to measurable desired results, and the desired results (measurements) must tie directly to the business objectives you are trying to achieve. Data can be used at every point of your business process to drive revenue, either through increasing revenue, reducing cost or increasing efficiency.
Cindy Stuckey is vice president, consulting, and Jeff Pelzel is senior program manager for Hunter Business Group LLC, a B-to-B marketing firm in Milwaukee. They can be reached at (414) 203-8060 or via e-mail at email@example.com or firstname.lastname@example.org.