Demystifying Business Intelligence
By: Kim Fernandez
True or false? Business intelligence systems are costly and confusing. False! As some associations have discovered, the return on such programs can make for better decisions, extra dollars, and more members.
The Texas Medical Association (TMA) had always done a great job collecting data. Staff members knew who their members were by age, area of residence, education, and a number of other demographics, and they also had information on people who were not members. They had information on income and reasons people did or didn't join, and they could—theoretically—develop marketing campaigns to very specific groups of people.
The problem was that the group had become so proficient at collecting information that its sheer volume was overwhelming. Sorting through all of it was nothing short of a daunting task, and trying to compile a list of members with any specific demographic trait could take days or longer. It had also become almost impossible to keep track of specific bits of information that affected the association's bottom line: For example, nonmembers were not supposed to get insurance through the group, but many were because there was no realistic way to track that as fast as applications came in.
"The hurdle we faced was that we had all the data in the world, but we didn't have a way to easily get to it without finding someone with a specialized IT background to come in and slice and dice things," says association COO John Dorman.
"We had all the data in the world, but we didn't have a way to easily get to it." —John Dorman
"What we came up with was a decision cube," says Dorman. "Think of it like a Rubik's Cube: You can take that data and turn it and dice it up to see it in a different way."
From there came a membership cube, which put individual bits of information (a member's age, county of residence, sex, or other specific tidbit) onto the virtual cubes, so they could be turned and manipulated. "We now have access to this cube, and we can break our members or nonmembers down by member type, then find out who's valid to recruit based on education or residence or whether they're deceased," he says.
What TMA developed has become known as "business intelligence," or BI, and it's something of a raging trend in the for-profit world.
"Business intelligence is a software product that gives a group the ability to mine for data," says Wes Trochlil, president, Effective Database Management, who says cost was typically the biggest reason nonprofit groups didn't follow the trend when it began several years ago. But that has changed, and BI systems have fallen significantly in price.
"It's much more affordable to the midmarket and the nonprofit area," says Trochlil. "The cost was primarily driven by software providers, and they maxed out the market at the top end. Now it's being pushed down."
David Gammel, CAE, president, High Context Consulting, agrees. "It's not free," he says. "But there are a lot of tools out there that aren't expensive. Some of it has been built for Microsoft applications. It doesn't have to be a bank breaker."
How It Works
BI systems work by categorizing data into tiny chunks, which can then be lined up in thousands of ways to give associations the exact information they're looking for.
TMA Business Intelligence Developer Cathy Faulkner says she can use her system to come up with lists of members or prospects with very specific characteristics. These can then be exported to Excel for mail merges, meeting applications, or dozens of other uses.
And, she says, the system is simple.
A TMA search starts with a list of five or six characteristics of members—sort by age, by location, by education, or by several other factors. Faulkner selects the characteristic she wants—doctors age 30 to 39, for example—and that list comes up. She can then select other characteristics, one at a time, to further narrow down her list; she can sort by sex, by specialty, by number of years practicing, by membership status, and even by which meetings they've attended in the past or what association "extras" they've purchased in the last year (or any other timeframe).
The system appears on a computer in a grid form, in plain English. Users simply mouse down to the characteristics they're looking for—there is no coding to learn and nothing special to remember.
"It is hugely valuable," says Faulkner. "The combinations you can get out of this are absolutely endless."
Dorman agrees. "We know that females are entering medical school right now at a higher rate than in the past," he says. "We know that from picking the gender option when we're sorting; we see more females than males. So we can make sure our marketing pieces are starting to reflect that fact. We can target to females age 30 to 39, which is different than in the past."
He says one of the greatest values they've seen from the system is the ability to know exactly who nonmembers are. By figuring out those people's demographics, the association has figured out new products to offer people to entice them to join and new ways to market to people who haven't yet joined. After all, they can get an exact snapshot of those people in a matter of seconds.
"We want to know who our members are and what they do, but we're more concerned about the people who aren't yet our members," he says. "We want to go after and market to those people, and we need to understand why they're not members."
He opens the system and selects physicians in Texas who are not members; the search comes up with 51,000 names. Dorman then drills down to people who are eligible for membership—people who haven't let certifications lapse or retired or passed away or moved out of the area. The list pops up; it lists 33,000 names, addresses, and other demographic data. He instantly pulls that into Excel, and that list can then be passed along to membership staffers for their use.
"In the past, we might have sent letters to all 51,000 people we saw at first," he says. "We were just able to immediately reduce that to 33,000 who are eligible for membership. Just like that, we got a more marketable target."
The results, he says, speak for themselves.
"Before 2003, our annual net change in membership was an increase of 350," he says. "We started the system and in each of the next four years after that, we averaged an increase of 1,200 new members."
Experts say BI systems are absolutely the wave of the future. But that doesn't mean they'll be simple for all associations to implement, even with falling cost and ease of use.
"This is a new way of looking at things that people have been doing for a long time," says Gammel. "We're moving from looking at data as a simple side effect of a transaction to looking at it as one of your key assets. That's an important distinction. That data tells you so much about your members, your customers, your exhibitors—whomever touches your organization. You can learn a lot from it, but you need to capture that data and create internal efficiency."
Trochlil says associations that are considering a BI system need to have five crucial things in place:
- The group must establish clear objectives or outcomes. They have to know exactly what they want to accomplish through BI implementation, whether that's to increase overall membership, increase a certain class of membership, or bump up sales of products or services;
- The group must have clearly established data-management processes. "You have to know how you're managing data in the organization," he says. "Are you using a single association management system, or are you using multiple systems? How is the data coming in, and who's managing it? And how do you get it?"
- The data to be used has to be collected. "You have to have the data to answer questions," he says. That means asking members for their birthdates, addresses, educational backgrounds, income, and other things that could separate out exact demographics.
- Data has to be accessible, in a CSV file or other readable file. "If you're using multiple sources, if you outsource event registration, for example, you need to be able to use that data," says Trochlil. "Make sure you get that data from the registration company into your system."
- The association must be willing to use data to make behavioral changes. "If you find out a membership class should be created and you're not marketing to those people, you have to be willing to change that behavior," says Trochlil. "That can be pretty difficult. A lot of us like to treat our members exactly the same, and that doesn't always work with this."
Organizations that can work to achieve those five ways of doing business are prime candidates for BI system implementation, Trochlil says.
TMA's Dorman says he was pleasantly surprised when the association kicked off its BI system in 2002. "The original cube that presents the majority of what you see took a week of two staff members' time working with a specialized consultant," he says. "Without that consultant, we could have spent 20 weeks to learn all of this from the ground up."
That said, staff members had to work closely with the consultant to get what they wanted out of the system.
"He was used to working with, like, 7-Eleven," he says. "He was building these systems to track how much bread was on the shelves. We had to break his mindset away from selling widgets and to our members and nonmembers and what they look like."
So, he says, finding a consultant who understands associations is key. Also key is ensuring that everyone in the group is on board to build and use the system effectively.
"You need to have a commitment from upper management that they understand why we need access to this," he says. "And you need to have the information in your system."
That, says Trochlil, can be a challenge for associations that haven't traditionally collected detailed information from their members.
"My definition of BI is that it's a management process for using data to make decisions," he says. "Rather than operating from the gut or using anecdotal information, you're making decisions based on real data. So the question becomes one of what are you trying to accomplish, and do you have the data to make that happen?"
Some groups, Trochlil says, have luck with a shotgun approach, where they plug every bit of data into the system and let it go to see what they find out. "You create a warehouse with that data and then you use software to analyze that data for any trends you might be missing," he says.
Others choose a focused approach, where the system is developed to answer very specific questions. Says Trochlil, "You establish a clear outcome of what you're trying to achieve, and you focus your data analysis on that, whether that's reducing marketing expenses or getting better at targeting prospects or creating new products in certain markets."
He says it doesn't have to be expensive; while some large companies have spent upwards of six figures on a massive BI system, associations like TMA can build a great system for their needs for just a few thousand dollars. Trochlil says he knows of one association that built its own BI system for nothing more than the time of a staff member who had the knowledge to do it.
Even Dorman says there are good reasons to strongly consider a group's need for such a system.
"If an association has a membership that's pretty static and they're not going to be able to make a significant change of who joins or who doesn't, there may not be a need to be able to drill down to this kind of information. But if you have data you can take and slice down to achieve an objective, then this is an appropriate vehicle to use."
His system provided an immediate return when it was implemented. As noted earlier, TMA offers insurance to member physicians, but there wasn't effective recordkeeping between the association and the insurance provider, so many nonmembers were using the insurance product.
"We got the system going and sliced our information down by nonmembers who had our group insurance," says Dorman. "We instantly found 1,100 physicians who had our liability coverage and were not members. We then sent a letter to those people telling them to join or lose their coverage. And boom—1,100 members, just like that."
Kim Fernandez is contributing editor to Associations Now. Email: firstname.lastname@example.org
Mark Wisby , March 02, 2010
What a nice article that clearly shows that Business is taking a new approach to it's company data. They have been collecting lots of data for years but very few have ever put that data into the hands of the right individuals at their company to make decisions on the data. Business Reporting software is the only way to harness the data and present it in fashion that can be used by non-IT professionals.