For Julian Culhane, it must have felt like he’d climbed aboard a rocket just as it was taking off. When he joined Lastminute.com as its CFO in November 1999, the online travel and gift retailer had revenue of only £195,000, or $283,000. One year later, that figure had soared 1,800 percent to reach $5.4 million.
And just as rockets burn through fuel at a frightening pace when they take off, so too was Lastminute.com blasting its way through the $224 million of cash that the company had raised when it floated in March 2000. What’s more, the territory into which the company was racing was the uncharted hinterland of the new economy. It all left Culhane with an enormous challenge: How best to steer a business moving at breakneck speed through an unfamiliar landscape?
If he was to have any chance of controlling the company, Culhane reasoned, he would need a decent set of performance metrics. So he set about building a dashboard for his rocket. Financial indicators were first: sales, profit and loss, cash position, and controls on two traditionally troublesome areas for dotcoms — spending on marketing and technology.
But, says Culhane, such measures only went part of the way. As the CFO of an E-business, he wanted to complement traditional numbers with indicators that gave a deeper insight into how Lastminute.com was doing. To that end, he added a raft of “Web metrics” to the dashboard: measures such as how many shoppers have visited the firm’s sites across Europe, how many of them made purchases, how many are repeat customers, and how many items are bought on average with each visit.
Armed with such data, Culhane is now able to calculate metrics such as how much it costs to draw a visitor to Lastminute’s site and what its customer payback levels are. And by getting data in real-time — Culhane downloads a one-page flash report every day — he is able to change the course of his company at a moment’s notice if signs of risk or opportunity appear.
“Having good metrics in real-time makes us much more responsive than an offline company could be,” enthuses Culhane. “For example, we change the special offers on our Web site several times a day based on click-through rates. It’s our shop window, so we need to get the biggest bang for our buck from it.”
Sadly, for many of Europe’s firms, Culhane’s fondness for Web metrics is something of a rarity, as a study released last month illustrates. Carried out by four business schools on behalf of SAS Institute, a software firm that specializes in performance management, the survey found that, of 145 companies in Germany, France, Italy and the UK, more than 81 percent had an E-business strategy in place. And yet, of those companies, just 41 percent were measuring E-business performance.
On top of that, when CFOs do monitor the progress of Web ventures, say experts, they often use the wrong types of metric. In particular, says Richard Roth, managing director of Hackett Benchmarking & Research (now part of Answerthink, a best-practices research firm), CFOs tend to rely on the same performance measures for their E-business as they do for their offline business.
Of course, traditional financial metrics will always have a place of honor — Web ventures need to earn a return just like any other project. But, argues Roth, using only old-economy metrics can be misleading: “The Web is different to the offline world; it has different drivers. Take retail. A lot of companies regard their Web site as just another store and use the same bricks-and-mortar measures to analyze it. In reality, it’s much more than that.”
For Peter Bull, manager in charge of consolidation, reporting and analysis software at SAS Institute, CFOs who don’t monitor their E-businesses are missing a trick. “One of the great things about the new economy is that it gives you so much more data than in the old economy,” he asserts.
One important source is log file data, generated by the Internet servers that host Web sites. This data shows, for example, exactly who has visited a site and how often, which pages they looked at, what they bought, where they came from and where they went. “Trying to manage and make sense of the massive quantities of Web data is a big challenge for CFOs,” says Bull. “But those who achieve it can reap rich rewards.”
So how should CFOs go about installing metrics for E-business operations? According to Paul Pilkington, a senior manager at PricewaterhouseCoopers, the first step is to recognize that Web metrics will vary from one firm to another. “Whereas accounting measures are meaningful across all businesses, Internet measures aren’t,” he explains. “Companies need to work out what the drivers of their E-business are and set their metrics accordingly.” Thus, for example, portals that rely on advertising need to attract viewers for as long as possible, while auction sites need to increase the number of transactions that take place.
An important part of this initial process is to agree on definitions, notes Anne Engel, senior analytics consultant at NetGenesis, which helps firms gather and interpret Internet data. “You can’t manage what you don’t measure, and you can’t measure what you don’t define.”
However, all measures should come with a warning label: What you see isn’t necessarily what you get. Consider the most basic building block of Internet metrics: the “hit,” a meaningless measure on its own. Every time a customer accesses a Web page, it counts as one page view. However, contained within that page are lots of different elements, such as pictures and panels of text, and each of those different element counts as one hit.
Put another way: just viewing a page automatically creates a hit for every element on that page. So, while a Web site with 100 million hits a month sounds impressive, it might mean that there were just 10 million page views, resulting from 1 million visits, made by just 100,000 different users, of whom only 10,000 actually made transactions.
“A lot of companies just focus on top-line numbers like hits and page views when they monitor their E-business, trying to build up the biggest numbers as if they’re in a popularity contest,” says Engel. Instead, she asserts, companies “need to take the covers off their Web site and look a bit deeper.” In particular, she advocates customer-centric metrics, such as customer conversion rates — the number of visitors to a site who become customers. But again, definitions are important. While an online newspaper might consider somebody who has signed up for a free E-mail newsletter as a conversion, a retailer might require something more, such as an actual purchase.
As Engel observes: “The Web lets you gather data on your customers in a way that isn’t possible for offline companies. Because of that, E-business is able to become customer-centric rather than process- or transaction-centric.”
PwC’s Pilkington concurs, and recommends that CFOs use a technique known as a “customer lifecycle funnel.” “At the top [of the funnel] is the number of visitors to your Web site. At the bottom is the number of completed transactions,” he explains. “It’s essential to measure and visualize the shape of the funnel.”
One company that has developed a variation of the lifecycle funnel is ProXchange, a U.K.-based Internet hub where secondhand goods such as machine tools and IT equipment are bought and sold. Less than a year old, ProXchange has set up shop in five European countries already, and has notched up “several hundred” transactions with an average value of about $14,000 each. A large part of the reason for its success is its close attention to customers through a clever use of metrics.
Of greatest importance is what the firm calls its “seven-level waterfall”, explains Michael Ogrinz, ProXchange’s CFO. Each level corresponds to a certain part of the Web site’s transaction process. At the top, level one is a first-time visitor, while level two is a repeat visitor and level three is a registered user. Moving down the waterfall, level four marks the moment when a registered user interacts with the site by asking a question about a piece of equipment for sale. Level five is where a potential buyer places a bid, while level six means the bid has been accepted. The final stage, level seven, is completed when ProXchange receives its commission from the trade.
“The idea is to get visitors to our site to move down the waterfall because that’s what drives our bottom line,” explains Ogrinz. “By monitoring each level, we can see where our customers get stuck in the system and take action accordingly.” For example, towards the end of last year, Ogrinz and his colleagues noticed that visitors to the firm’s French site weren’t progressing beyond level two. To address this, they devised a direct marketing campaign, encouraging repeat visitors to register. “It was highly successful,” boasts Ogrinz. “There was a lot of trickle-down.”
Along with the waterfall, Ogrinz also watches the performance of every one of the 21 business categories on ProXchange’s Web sites — be it heavy cranes or industrial cleaners — on a country-by-country basis. “Only those categories that earn a sufficient return will survive,” he says. And like Culhane of Lastminute.com, Ogrinz feels his Web metrics let him make those sorts of decision at lightning speed: “I can see trends as they’re unfolding and react immediately.”
Calling All Information
However, when it comes to gathering and analyzing metrics in the new economy, there’s more to it than harvesting Web-site data and splicing that with financial figures. For a complete picture, data from other sources should also be used. For example, information from fulfillment and call centers can shed light on how an E-business is performing. In the same vein, adds SAS Institute’s Bull, CFOs should take a closer look at working capital performance. In a Web environment, he suggests, supply-chain relationships become much tighter, enabling companies to reduce inventory “to a bare minimum.”
Another important source of data comes from other competitor Web sites. One company that can help to compile data such as how many visitors does a Web site get is NetValue, an Internet-monitoring firm. It does so by setting up panels of between 3,000 and 5,000 people in different countries and then observing how they surf the Web. With panels currently in seven European countries, NetValue believes it can give a fairly accurate indication of traffic flows across the Internet.
Bernard Ochs, vice president of business development at NetValue, says that the service is in big demand, and not just as a way to compare one firm’s site with that of a competitor. “Increasingly, people want to know what the effect of a merger or an acquisition might be. If this company merges with that company, how much customer overlap is there?”
Partner firms, too, can provide a wealth of information. As Willem Ackermans, CFO of KPNQwest, the E462 million ($431 million) data communications company based in the Netherlands, notes: “E-commerce is all about being open, building closer partnerships, and integrating with your customers. When that happens, you have to share information.”
As such, KPNQwest is currently preparing a system so that its customers are privy to the same performance metrics that the company uses. “We’ll let our clients see exactly what sort of service we’re providing,” declares Ackermans. “They’ll know the speeds of our network and the efficiency levels, just as I will.”
Of course, coming up with useful metrics isn’t a one-off exercise. Just ask Massimo Cristofori, CFO of Tiscali, the Italian Internet service provider that recently bought Dutch rival World Online.
As he sees it, one of the biggest challenges for CFOs trying to monitor an E-business is the rapid pace of change on the Internet. “Trying to keep your finger on what the drivers of the business are is tricky because they change all the time. Six months ago, I was using different metrics to those I use today, and I’ll be using different ones again in six months from now,” he sighs.
This Month’s Metrics
Currently, Cristofori focuses on four key Web metrics. Three of them — the number of active subscribers, the average number of minutes each subscriber spends online, and the number of pages they view — drive revenue, given that Tiscali makes its money by charging for connection time to the Web and by selling advertising on its portals. The fourth metric is the cost to acquire a new customer. Before the merger, World Online had an average customer acquisition cost of E70 ($65), while the figure for Tiscali was just E18 ($17). Now that the two companies have joined, Cristofori has set a target of E40 ($37).
As the CFO of a new-economy company, Cristofori probably has a head-start on many fellow finance chiefs. But, as the Web continues to evolve, that’s likely to change. In the future, even CFOs of traditional businesses will have a toolbox of Web metrics by their side.
Justin Wood is the managing editor of CFO Europe.
A Matter of Meaning
Web measures should come with a warning label: What you see isn’t necessarily what you get. Consider these definitions.
Abandonment rate. The number of people who start but don’t complete the buying process.
Acquisition costs. Advertising and promotion costs divided by the number of click-throughs.
Attrition rate. The percentage of existing, converted customers who have stopped buying from a site and have gone elsewhere during a specific period of time. The awkward-sounding verb form is attrite.
Churn ratio. How much a firm’s customer base “rolls over” during a given period of time. To calculate churn, divide the number of customers who attrite during the given period by the total number of customers at the end of the period. Take, for example, an Internet service provider that has 2,000 customers at the start of the month, gains 200 new subscribers, but loses 50. The churn ratio is 50 divided by 2,150, or 2.3 percent.
Cost per conversion. Advertising and promotion costs divided by the number of sales.
Duration. How long a customer spends on the site. Short visits may be better than long visits for certain companies.
Frequency. How often a customer visits the site. For newspapers it could be daily, while for florists it might be quarterly.
Recency. How long it’s been since a customer last visited the site. As recency diminishes, the potential for future purchases decreases. After a previously specified period of time elapses, the user is considered to have attrited.
Stickiness. The total time spent viewing all pages divided by the total number of unique users. (More exactly, it equals the total frequency (the number of visits in a certain period divided by the number of unique visitors in that period) multiplied by total duration (the time spent viewing all pages in a certain period divided by the number of visits during that period) multiplied by total site reach (the number of unique users who visited during a certain period divided by the total number of unique users).)
Stickiness is a composite measure of how well and how consistently a site’s content holds users’ attention and allows them to complete their transactions. Some parts of a site should be slippery rather than sticky.
Velocity. How quickly a user moves from visiting a site to completing a transaction on a site.