Don’t underestimate the little things in a business.
Randy Mattran, vice president of professional services at Burnsville-based Lancet Software, consults with a lot of retail and hospitality companies. Retailers are working on increasingly slim margins these days. It used to be that the difference between profitability and ruin was a few percentage points, but now everything hinges on mere basis points. So Mattran builds business intelligence (BI) solutions for his clients so they can keep an eye on every penny.
“There’s a fast-food chain where we built a mobile application so that the franchisees could analyze things like food cost,” he says. “One of the operators realized that a couple of their stores were building their sandwiches to the recipe, but with one exception: They weren’t snapping the bacon in half and putting it in an X across the sandwich. They were using two full pieces of bacon instead of one snapped in half. That was increasing their food costs, and it actually had a pretty high impact on the margin. It was BI that made them aware of that training issue.”
In another client’s retail stores, Mattran put BI tools on iPads so managers could have access to live data while walking the floor. The managers can tell, on the fly, whether the store is busier or slower than usual, which helps them decide whether to call in additional personnel or send someone home. They can also track inventory, spot performance issues, and tell whether current sales promotions are working effectively.
“If they don’t have good and accurate information to go on, they lose their competitive edge,” says Mattran. “So BI is critical. That’s their dashboard. That’s their instrumentation on the business.”
Put simply, a business intelligence solution is any system that helps a company transform its raw data into useful information that can help improve its business outcomes. But Neelesh Raheja, vice president of consulting services at GNet Group in Roseville, says BI is just as much about leadership and communication as it is about number-crunching tools.
“How does a leader make an organization more data-driven and influence behavior and cultures to actually impact outcomes?” he asks. “At the end of the day, the outcomes are not going to change if there is no change in behavior. How do you change behavior? You use data. And how do you use data? Through dashboards and BI tools. So BI tools are more and more used as a communication tool rather than an actual intelligence tool.”
All kinds and sizes of businesses benefit from knowing more about their data. TrueCar Inc., a 300-person Los Angeles company whose auto-buying programs currently account for 2 percent of U.S. new vehicle sales, has been running MicroStrategy BI software since 2008 and maintains its own six-person team of business analysts who use a variety of solutions.
The company signed a contract with Lancet a year ago to give its system a health check and leverage Lancet’s resources to boost output. TrueCar was able to up its game, taking what used to be plain spreadsheets and transforming them into graphical representations that make the data come alive.
“We’ve done numerous new analyses to help our dealer development team identify the highest-priority dealers or highest-priority areas and brands for us to try to get onto the network,” says Jim Berger, TrueCar’s vice president of business intelligence and operations.
“We were able to create visual reports to show them where the demand was for a given vehicle brand and where our network was located versus dealers that were not on our network, to determine which dealers should we target to improve our coverage in a given area for that brand.”
BI is different for every business, since every business has its own key indicators and strategies. However, Mattran says, there are a few basics that almost everyone benefits from.
“All the retailers have some way of doing category management—what are the top sellers,” he says. “Everybody’s doing a little bit of market baskets—what’s being sold with what. They can tune their assortment and they can micro-market a little bit. If one particular store regionally needs snow shovels and another doesn’t, that becomes really apparent and can be centrally managed more easily with BI.”
Those kinds of capabilities have been standard-issue for about 15 years, he says. What’s newer is a greater focus on customer data. Now that more and more retailers can identify who is making a transaction through loyalty programs and private-label credit cards, they’re making decisions based on customer lifetime value and household trends.
They’re able to see whether the value proposition in their marketing programs is bearing fruit or not.
Increasingly, it’s hard to imagine a serious business that doesn’t use BI to some extent. Marketers in particular, with their immersion in CRM (customer relationship management) and social media, are likely to expect it as a matter of course.
“I really think it’s been redefined,” posits Karl Populorum, vice president of sales and marketing at RBA Inc. in Wayzata. “Many years ago, it was just about enterprise data warehouses and multimillion-dollar deals. But now, with the advent of social [media] and the Microsoft technology stack, we all have business intelligence. There are simple front-end tools that allow you to do analysis. There are web analytics, social data and analytic tools, all kinds of embedded BI, in Facebook and Twitter. There are dashboards built into a lot of web tools that are free to use. So I think it changes users’ expectations of what internal BI solutions should be providing.”
How, then, to begin? Populorum says the first step is to look around at what you already have in-house.
“People are sometimes surprised, or even unaware, that under their enterprise agreement, they have access to extensions to [Microsoft] SQL that they may not even know about,” he says. “IT might have bought an enterprise agreement but the business either isn’t aware of it or doesn’t have the technical skills or the approach to be able to leverage it. Or they’re eligible for funding to do a proof-of-concept to do something. There are a lot of both investment programs and partner incentives to help jump-start solutions or get customers thinking about what they can do.”
Raheja recommends against buying a BI solution just because it’s offered by, say, your ERP (enterprise resource planning) vendor.
“That’s probably the worst decision to make,” he says. “The reason people do BI is because they want to analyze data from a variety of sources and not just that one ERP solution. What that means is you don’t need to go monolithic with your approach. You have to use best-of-breed. Even small businesses have at least two or three applications that run their core business functions. So buying a BI solution that fits only one of the solutions is a lost battle on day one.”
In general, BI isn’t plug-and-play, and there aren’t a lot of off-the-shelf software programs that spew out ready-made business intelligence results. That’s because BI is at its most useful when it’s customized—and it’s not terribly difficult to build template data models and dashboards in platform tools such as Informatica, MicroStrategy, and the Microsoft stack.
“The tools are so good and so productive that you can build what you want yourself pretty quickly,” says Mattran. “The construction time is not the issue, so the advantage of having a canned application isn’t great. You’re going to change it anyway. Your end users are going to build their own reports based on the metrics that you give them. You’re very quickly going to deviate from what any one product release would provide.”
That ease of construction lets GNet put generic BI starter kits or “accelerators” (tailored for various industries and for departments such as marketing, finance, pricing, and manufacturing) into the hands of its clients with breathtaking speed.
“Excel has over a billion users,” points out Raheja. “So we are able to deliver solutions in a matter of weeks that would traditionally take months.”
There’s no point in spending six months to create the perfect solution before letting the client play with it, says GNet’s director of business development Greg Fair.
Better to put some simple tools in their hands and watch the sparks fly. Each question begets another: Why are sales low in this area? Why isn’t that product doing well? Is it a salesperson or a customer-demand issue?
“What we’re doing is we’re putting it in their hands quickly and using the feedback that we’re getting from them to actually build the foundation [of the BI solution],” he explains. Because you still need that foundation: the data warehouse, the data mart, the OLAP [online analytical processing] technologies. But by having them see and play around with the data quicker, we’re using that feedback to build that foundation and we don’t have to go back and rework anything later.”
When an organization implements BI, its staff will soon come to depend on the information it generates. That can present a problem if there are only a few people who know how to make queries. So most organizations either partner with a consulting firm that provides additional manpower or train people across the company to use the tools, or both.
“If we could change one thing, we’d have more staff!” says TrueCar’s Jim Berger with a laugh. “So we try to distribute and do self-service. We train up analysts in some of the other departments and make sure they have access to the tools, so it increases the value and distribution of business intelligence across the company. Of the 300 people across the company, we have about 50 licensed users in MicroStrategy, and I’d say half of those are very active daily users. We have probably 300-plus, maybe more, reports going out monthly to about 300 to 500 end recipients, both internally and externally.”
In general, says Tony Morse, marketing director at Plymouth-based firm Digineer, the company’s COO, CFO, or chief marketing officer should lead a coalition of both IT and non-IT staff who work together to translate business goals and strategies into business intelligence questions. Even though those executives aren’t computer specialists, they shouldn’t have too hard a time with the tools, he says.
“The truth is, people in each of those positions are used to viewing and managing and making decisions based on data in some fashion as it is,” he points out. “I wouldn’t look to any of those three to implement and get data into the systems. But they can certainly use them. Their roles are focused on identifying the types of indicators that they need to manage their business.”
Training should happen over the long term, says Raheja. It’s not a one-time event or something that can be considered “done.” Users will need to become comfortable with the tools, but leaders will also have to become more adept at change management—translating the information garnered from the BI solution into new policies and directions within the organization.
“But we always help out,” says Raheja. “Whenever we do a project, we do it in partnership with IT resources from the customer. We make them go through the process, and then we hand over the project in such a way that they are well-equipped to run it independently. We teach them how to fish rather than just catch the fish and throw it over the fence.”
Business intelligence consultants see it all: Companies that worry more about buzzwords than about strategy, customers who go all-in on an expensive BI solution and then somehow forget to train their staff to use it. It’s too bad, because the pitfalls are easy to avoid. Here are a few of their hints for a successful BI implementation.
Assign the BI program a strong, committed executive sponsor.
Someone has to own the BI effort, says Bob Thomas, vice president of consulting services at Digineer in Plymouth. “Generally speaking, the key roles within the organization that really need to stand up and sponsor this would be your COO or someone in operations leadership, your CFO or your CMO.”
Create a cross-functional team of “purple people.”
GNet Group’s Neelesh Raheja says that if you think of IT as red and business as blue (or vice versa), the purple people are the go-betweens. The best practice is to create a cross-functional team of these people to support the executive sponsor.
“Business intelligence is not owned by IT,” he says. “Neither is it owned by business. It’s a partnership. Every organization will have a different degree of ownership between IT and business, but what’s important is they come together at the table.”
Crawl first, then walk and run.
Work with your consultants or business analysts to create a few simple tools and show all your purple people how to use them. Then, once you’ve mastered the easy stuff, start asking harder questions and building more ambitious tools.
Align your BI strategy to your business strategy.
Thomas says many companies seem to gravitate to BI because they think their competitors are doing it. Their real motivation should be figuring out how to turn their business strategy into a reality. “Ask yourself, ‘What are the things I need to measure to know that my strategy indeed is taking hold and being successful?’ And then create the projects that are going to drive the organization to achieve those metrics.”
Align your BI strategy with industry benchmarks.
If you want to improve a particular metric, a good starting point is to see where your competitors are and use BI to measure your own company against that standard.
Budget, plan and hire for the long term.
“BI is a pursuit, not a destination,” says Raheja. “There is no end to BI projects. If you’re not doing any more investments in business intelligence, you’re basically saying you’re going to go out of business, because you’re not going to improve anymore.”
Jamie Swedberg is a freelance writer and a frequent contributor to Twin Cities Business.