I have two very good friends who have been blogging about metrics recently. Jeff Young, over on zd.net/blogs and David Churbuck at Churbuck.com. It's been interesting to talk to both in real time about what they're writing. Dave Churbuck was the founding editor and publisher of Forbes.com and has great knowledge of Internet strategies.
My other friend, Jeff Young, has an iconoclastic view of Internet metrics and their importance to shaping sites. Jeff's argument against metrics and data collection appears to be based on fears about losing his privacy, something I'm sure many people have thought about over the last several years. his screed on meric and the Cookie Monster is rally worth reading. Go to:
http://blogs.zdnet.com/Young/?p=17
I've taken a path down the middle. From a business standpoint, I understand that it's important to know as much as you can about your customer. I also believe that the most important metrics, for commerce sites, is a concept of turning a site view into a sale. But for some companies metrics have gotten in the way of commerce. That's not good.
I suspect that when this happens, it's the fault of companies who are putting managers who lack sales experience directly in the way of transactions. It's this scenario though that overly complicates web commerce.
Realistically, customers should never be aware of data collection. And most of all, customers should never be made to feel that a transaction (including cookie collection or web histories) compromises their personal privacy. Jeff Young argues that data collection is intrusive and big brotherish.
Dave Churbuck takes a more reasoned approach and points out that commerce sites need this data to establish direction and monitor effectiveness. Dave brings up one of the most important points though: sites need to have baseline information on the lifetime value of a customer.Read it at:
http://www.churbuck.com/wordpress/?p=517
My point is this: anyone who doesn't understand this one metric has never spent time around serious brand management. For example, pharmaceuicals, automobiles, or even (gasp) the defense industry. As someone who has worked on the marketing side of those industries, I've seen firsthand how valuable a customer lifetime value metrics are. The best two examples I can think of are the auto business and the defense industry. Brands like Ford have an excellent trend base for how many like-branded cars a customer may buy over 40 years. On the defense side, the former Grumman organization knew very well how likely it was to win a bid for naval aviation aircraft, based on the past history of its customer and how buying matched governmental trends on a decade by decade basis for it's various classes of products.
Grumman never sold it's strike aircraft on the web, but it knew everything there was to know about its customer. The same is true with Ford, which indirectly sells cars on its website. And you can bet a new pair of Towncraft jockey shorts, that JC Penney's, Amazon and other successful web sites know how much you spent in the current and past quarters and whether those transactions were tied to customer loyalty programs unique to an individual's buying habits, site-wide sales, or other key attributes. It's good baseline information. Also once a marketing or sales organizations have a trend line, they readily can project potential lifetime or product sales with reasonable accuracy.
In the heady days of the Internet Bubble, i sat through numerous dog and pony shows where entrepreneurs would go off into marketing Fantasyland, trying to project customer life time values based on too few or unrelated transactions. One example still leaps to mind, a retail site aimed at a specific category of users that sold items the sites founders found useful for their pursuits and then found data that matched their buying behavior. Perhaps worse yet, they were parsing changes in unrelated retail sites like Amazon, which at the time was moving beyond books.
Most of the lessons I learned from this experience were straightforward. Know your customer but don't assume you're the top brick in a mighty pyramid shaped market. Successful brands like Amazon, Thinkpad, JC Penney's and others all know their customers and a handful (Amazon) are using blogs to link directl to their customers. Going back in history, HP and Tektronix also thoroughly understood their customers from the time when they had their first user experience in college, through professional employment and on to pre-retirement buys. Big ticket items were sold based on personal relationships and the mother lode of customer data resided in individual sales force members, not in distributed customer relationship management software.
To this day, the pharmaceutical industry still stands out for its use of face time in the sales process. Virtually every marketing decision made in pharmaceuticals is designed to help a detail (sales)person get in front of the purchase decision maker. Although drug companies use CRM software, their implementations give the salesperson the information they need to lock down sales. 20 years after CRM appeared pharma is still face on face, shoe leather worn thin by walking hallways, as an exercise in patience waiting to get in front of a doctor or group pharmacists. And it works
What organizations like pharmaceutical companies and auto makers have done is to make their customers part of the brand. Doctors may hate it (but the next time your waiting to see a doc, check out their pen collections, or look closely at their prescription pads. Better yet, look at big ticket items used on transportation or agriculture. The Caterpillar brand is so ingrained in my ethos (thanks, Dad!) that if they made a small wheeled tractor with a mowing deck, my green Bambi Deere would be gone the same day that Cat began shipping the product to my local dealer. Peterbilt is another example of a big ticket( $200,000 and up) brand with incredible loyalty.
While libertarians and iconoclasts may scream that collecting customer buying information represents some sort of invasion of their privacy, the most successful Internet brands going forward will belong to the companies that do the most effective job of data mining and which through trial and error develop the most accurate predictive buying algorithms.
It's those two pieces of technology, plus the sites that let people easily buy merchandise quickly and efficiently that's going to build brand loyalty. Conversely, interjecting technologies that disrupts an otherwise painless experience or detracts from a brand is a surefire way to sever the cord between a customer and a brand or site.
(obligatory disclosure statement: i wear Penney's Towncraft underwear, prefer to buy and use ThinkPad and HP Presario/Pavilion computers, and shoot rampaging gophers that invade my gardens with a trusty Beeman .177 pellet rifle and a Nikon scope. I have a Suzuki 4-stroke 40 hp outboard on my 18-foot panga and enjoy fishing. That's way too much information to give out, eh?)--Jim Forbes, writing outside on a glorious 75 degree morning, keeping an wary out for burrowing rodents in my garden here in rural northern San Diego County.
Brilliant post on life-time value of customers. And I respect your middle-ground approach to metrics. Metrics are nothing but measurement. Analytics are the insights you get from the ruler. But all of it is flapping in the wind without action.
And Jet-Skis do suck. There allegedly was a bar in the Florida Keys that was a haunt of the local fishing guides. On Friday nights they would set one of the infernal machines up outside and sell whacks with a sledgehammer to the pissed-off crackers.
Ya know this would be a fun thing for me and other members of Bloody decks.com to sponsor.
Posted by: David Churbuck | April 21, 2006 at 05:18 PM