Pipemaker Learns a Hard Lesson in Customer ProfitabilityBy Michael Fitzgerald | Posted 02-06-2006
Pipemaker Learns a Hard Lesson in Customer Profitability
It was late fall of 2002 in Portland, Ore., and Dave Ramsey, owner of United Pipe & Supply Co., a $174 million supplier of plumbing and irrigation systems throughout the Pacific Northwest, couldn't believe his eyes. "You've got to be kidding me," he said, staring dumbly at cio Mike Green's computer screen. "That's our worst customer?"
It's not unusual for brand-new owners of business-intelligence software to experience an epiphany upon installing it. When the numbers start feeding into the system, companies learn things about their businesses they never dreamed.
We spent how much on travel? Our inventory is how many months behind? Can we switch to a cheaper brand of coffee? But when Ramsey and Green first saw United Pipe's numbers on customer profitability, they shared the shocking revelation that the company's biggest and most prized customer was also its least profitable. In fact, their so-called best customer was costing them $50,000 a year.
"Dave turned white and got real quiet after that," recalls Green.
Thus began the process of analyzing how, exactly, such a critical customer could have become so unprofitable over time. And in an effort to maximize profitability across the company, United Pipe launched a companywide examination of each of its customers. The problems the company uncovered ranged from unpredictable pickup and delivery schedules, to revenue-motivated salespeople who overserviced their biggest clients.
After United Pipe addressed some of the issues, profits before taxes increased from 1.5 percent in 2003, to 4.7 percent in 2005.
"Salespeople are hunter-gatherers, out there to bag the next customer. But volume doesn't necessarily yield bottom line," Green says. The difficult part is convincing them not to say yes, automatically, to every single customer demand. Green believes many salespeople are willing to agree to terms that make the customer unprofitable from the get-go, and any increase in service just makes things worse.
Since the revelations, United Pipe salespeople have begun to understand that there are times when, if a potential customer says it can get better terms from a competitor, it might make more business sense to wish them well. "Sometimes, with information in hand, we have more of a spine," Green says.
United Pipe is a family business that was founded by Taylor Ramsey, Dave Ramsey's father, in Portland in 1953. By 2000, it had 24 branches throughout Oregon, Washington and Idaho. When Green was hired, that same year, the IT systems were overwhelmed by the company's growth, almost to the point of breaking down. Customers often had to stand in line to get their supplies.
Overhauling those IT systems was Green's first order of business. But even in Green's job interview, he and Ramsey had talked about how to get a better sense of the firm's costs. Green's knowledge of business-intelligence tools was one of the things that got him hired, even though he had no previous experience as a CIO. (Prior to his employment at United Pipe, Green was vice president of technical services at Pacific Information Systems, an IT consultancy.)
Ramsey wanted better insight into how his branches were performing on things such as cost management, and he also wanted to get a sense of how individual branches compared to each other on a per-transaction basis.
The tricky bit was that, like many companies, United Pipe organizes its data not by branch but by geography and business unit. The company has six regions, and within each of those regions the business is split into five parts: waterworks, irrigation, pumps, HVAC and a general supplies category. But because those units share warehouse facilities, administrative personnel and delivery people, it's difficult to break out costs transaction by transaction.
Asking IT people to run repeated reports is no one's idea of a fun time, and punching all the numbers into a spreadsheet or database is an arduous task rife with potential for errors.
First, Get the IT
Up To Speed">
In 2002, Green bought a package of activity-based costing tools from ABC Technologies (a company that was later purchased by SAS), and the game was on. The software set up individual data models for each branch's inventory and transaction records. That allowed the company to determine its baseline costs and look for anomalies. It found some, such as a branch manager in the Portland area who had average per-delivery costs of $200, versus $40 for the typical branch.
That turned out to be, in part, because the manager was dedicating a delivery truck and crew to a customer doing work 70 miles away, on the Pacific Coast (a customer that was closer to another United Pipe branch, in fact).
As they dug deeper into the numbers, it started to become clear why the company's largest customer was also its worst. The customer had squeezed United Pipe to the bone on pricing, so it was barely making any money on its sales. And the customer sent crews to United Pipe branches to pick up supplies several times a week. Each time a United Pipe employee had to drop what they were doing to take an order and punch it in. Then the order had to be pulled.
That tied up United Pipe workers both at the front of a branch and in the supply area, and it did so in unpredictable ways. Then there was the kickerthe sales representative on the account had become a highly paid gopher, frequently running things the customer had forgotten over to its job sites.
Next came the hard part. United Pipe had to take steps to rein in its costs in servicing this customer, which Green declined to name for obvious reasons. The company knew this customer kept inventory in its own facility, so Green wondered why the customer's crews were coming in so frequently, sometimes several crews in a day, to pick up more supplies, in addition to the deliveries that United Pipe was making directly to the customer's work site. "It was wildly expensive, running out there so often," Green said. "We obviously had an inventory management problem."
United Pipe sales execs sat down and figured out how to consolidate those deliveries into one big weekly deliverypulling a large set of supplies once a week was only marginally more expensive for United Pipe than the cost of pulling individual orders. It also began the delicate task of negotiating changes in delivery frequencies without making the customer feel like it was getting bad service, and in particular without harping on topics that might offendlike asking why a United Pipe salesperson was acting as a gopher.
United Pipe executives decided that the best way to handle the negotiation was not to discuss its own costs, but to focus on the number of times the customer sent crews to pick up supplies. "We went to them and said, 'Your guys are coming to our branch multiple times a day, with an entire crew in a truck each time. That's gotta be costing you a fortune,' " Green says. The logic worked.
Just as important was banning the salesperson from delivering items. "Having a sales guy do delivery is crazy," Green says. In fact, it turned out there were other salespeople who also doubled as delivery people. United Pipe has since mandated that they no longer deliver to customers.
Auditing Customers Gets More
United Pipe is far from alone in using business-intelligence tools to shine the light on its costs and expose which of its customers are not profitable. The market for business analytics, broadly defined as both data-warehousing tools and the software used to ferret information out of such warehouses, made up $15 billion in sales in 2004, according to IDC. Dan Vesset, an IDC analyst, says he estimates that the market hit $16.5 billion in sales in 2005, an 11 percent jump. He expects 10 percent growth on average over the next five years, making business-analytics software one of the fastest-growing large software categories that IDC tracks.
For United Pipe, making its biggest customer profitable only whet the company's appetite for further improvements. Green set up a list of customers who, while profitable, made frequent trips to United Pipe branches for supplies, instead of keeping inventory.
One such customer was Don Farrelly, a partner in Portland's LBD Landscaping LLC, and a customer of United Pipe for more than 20 years.
Farrelly has always picked up his orders, even though United Pipe will deliver for free. In 2002, Farrelly's United Pipe sales representative started telling him in advance what supplies he was going to need, based on past buying behavior and open orders, and the representative also began walking Farrelly through the numbers of what it was costing him to send his crews out to pick up materials job by job.
The discussions convinced Farrelly to invest in storage space and inventory, and to let United Pipe deliver on a weekly basis. Though keeping inventory was an additional expense, it meant his crews could spend more time working.
"The biggest waste in a landscape company is time," Farrelly says. His firm employs between 25 and 35 people, depending on business demand, with two or three landscape construction crews typically out every day. If crews aren't working because they need parts, that costs him time. "Having stock and inventory, and having them deliver, that saves me time. It's been a huge factor in helping us get bigger," he says.
For United Pipe, of course, delivering on a regular schedule means lower costs. There's a corollary benefit as well: Customers who keep inventory are less likely to defect to competitors. In the last year, United Pipe has begun offering a "mobile warehouse" to customers such as municipal waterworks contractors. The trailer, branded with "This project supplied by United Pipe & Supply," is refilled at job sites. It also gives United Pipe more of a role in managing such large jobs, and makes it a full partner with its customers.
Green says that this approach of helping customers streamline operations in ways that save them money has helped turn a number of unprofitable customers into profitable ones. In his experience as a consultant, Green learned that most companies will find that a chart of customer profitability will look like a sperm whale (see "Whale Tale", page 52)a high flat head to represent the most profitable 10 or 20 percent of their customers, followed by a long gradual slope representing the 50 percent or so that are marginally profitable or unprofitable, and then a falloff to the bottom 10 or 20 percent that are significantly unprofitable.
"What we've done is elongate that whale," Green says. There have been some customers that United Pipe simply cannot service profitably. It evaluates them on a case-by-case basis, and it might, if it has to, "fire" a customerthough it wouldn't do it outright. Instead United Pipe begins to negotiate costs, maybe asking an unprofitable customer to start paying for delivery. Green says what's more important has been getting regional sales managers to coach their salespeople to think more about making sure their new customers start out as profitable ones.
A Great System That
Despite the obvious benefits United Pipe has realized from its customer analysis, there were problems. Besides costing the company about $250,000 up front, the ongoing maintenance costs were considerable. At the time of implementation, "it was not well integrated" with United Pipe's CRM systems, Green says. That made it difficult to get data from the branches into the BI tools. United Pipe had to employ a business analyst just to keep data flowing to it.
When United Pipe's owner and president, Dave Ramsey, died unexpectedly, in May 2003, the project was almost lost forever. The CEO who replaced him was less fond of technology than Ramsey in general, and in particular he did not like the SAS tool. In fact, he looked at the cost of keeping the data warehouse up-to-date and told Green to mothball it.
Green stopped updating the data, but he quietly kept the tools around. The models he'd built were still useful for analysis, even with stale data. He caught a break, too, when that CEO left the firm after ten months, before Green had to decide whether to pay his annual maintenance fees to SAS. A new, interim CEO was more numbers-oriented and told Green to maintain the system, updating it with new versions, but not keeping the data fresh. In late 2004, United Pipe hired Ed Kolasinski as its president and CFO. And after he got his feet under him, Kolasinski told Green the SAS system would be important.
In the early part of 2005, Kolasinski and Green set new priorities. United Pipe was expanding againit now had 28 branchesand needed to improve its inventory control. The SAS tools would be key to thatand the data started flowing again from the branches.
It helped that while the system lay dormant, SAS had developed an Exact-Transform-Load (ETL) utility that makes it easier to get data into the SAS tool. It also helped that other managers, such as Dan Kraxberger, United Pipe's vice president of purchasing, were getting interested in analytics tools. Kraxberger wants to use the activity-based costing tool to do things such as examine the company's competitors to make sure that United Pipe is being aggressive enough on pricing, and use the data to get recalcitrant suppliers to lower prices. He also would like to look at his product lines to see whether some aren't generating the revenue they should.
Kolasinski also charged Green with bringing business-intelligence tools into new parts of the business, such as budgeting, financial reporting and general analytics. For this the company is employing less sophisticated softwarespreadsheets on steroidsinstalled on workers' desktops. It's typical for firms that use business-intelligence tools to adopt several types of them and drive them down throughout the organization, says Kathy Quirk, a research manager at Nucleus Research Inc. in Wellesley, Mass. "They're addictive," she says.
For the new tools, which come from OutlookSoft Corp., company controller Cheryl Summers has become the project champion. Green says the impetus behind this investment, which will also work out to be about $250,000, is less about direct costs, and more about the quality of work being done. He thinks, for instance, that Summers will be able to reduce the time she and the company's assistant controller spend closing the books each quarter, from ten days to three. The budgeting process should also go faster. Will that show up as clearly in the bottom line as finding unprofitable customers? Probably not, Green says. But Green expects the time savings will prove meaningful.
The addition of the general analytics tools will spread their use into a mainstream group of users at the company. Green thinks that can only help reinforce good business practices, like making sure customers start off on a profitable note, and keeping them there.
Michael Fitzgerald is a freelance business journalist based in Massachusetts.