Managing by Metrics: Web Metrics
June 19th, 2008 by JoelForrester is projecting that internet sales will reach $204 Billion this year. That’s a big number and growing in large part because the web has taken accountable, direct-to-consumer advertising to a new level. While we previously used TV, radio, and print advertising to target general demographics (say 25-45 year old females) we are now able to use the web to target extremely specific demographics (someone doing a key word search for “weight loss supplement”). Practically that means you may pay more money per impression BUT due to the fact that the advertising is specific and targeted (and efficient), you theoretically have the potential for higher ROI on your marketing dollars.
Sounds like the holy grail of advertising right? Well, not exactly. Just as we struggle to accurately measure and account for TV response calls (and consider unique callers, CS calls, junk calls, etc.) the web has also had its own difficulties. You’d think that the powerful internet would at least be able to tell us how many people visited a site but as reported in the New York Times that is not always the case:
How many people visited Style.com, the online home of Vogue and W magazines, last month? Was it 421,000, or, more optimistically, 497,000? Or was the real number more than three times higher, perhaps 1.8 million?
The answer — which may be any, or none, of the above — is a critical one for Condé Nast, which owns the site, and for companies like Ralph Lauren, which pay to advertise there. Condé Nast’s internal count (1.8 million) was much higher than the tally by ComScore (421,000) or Nielsen/NetRatings (497,000), whose numbers are used to help set advertising rates, and the discrepancies have created a good deal of friction.
As a marketer, this may leave you befuddled but take heart. As we’ve discussed on this blog before, the key to metrics based management is to define the metrics and then consistently track and manage to those metrics. Here’s what that means for the web:
Get consistent and get accountable. As highlighted by the NYT you could run metrics reports for a website from three different platforms and get three different numbers for unique visitors. How do you deal with this? Get consistent by having a single tracking platform and get accountable by requiring your web providers to operate on your standards. There are some excellent free options for this (Google Analytics, Index Tools) and even better paid alternatives (Omniture). Figure out what meets your needs and then deploy it across your websites. The beauty of these metrics platforms is that they can be deployed on both INTERNAL and EXTERNAL sites. This is the only way to create a consistent, level, and accountable playing field so you really know what’s occurring on your websites.
Know what to track. The first time you log into a platform like Google Analytics you might feel overwhelmed. Visits, sources, keywords, adwords, pageviews…the list goes on. Now if the goal for your site is to simply build brand awareness, then you’ll be looking at an entirely different set of criteria. If your goal is commerce (which we assume it is) then figure out both your core metrics and supporting metrics. One example of a good core metric is Revenue/Visit. This takes both conversion (orders/visits) and AOV (average order value) into account and allows you to quickly see your overall trend for the web.
Tracking by media acquisition channel. In the early days of the web there was a lot of easy money for DR Marketers. Paid search and banner advertising were the hot new mediums and TV driven campaigns enjoyed exceptional performance as they already had great brand awareness. Many marketers (including myself), didn’t know or didn’t care to really understand the full intricacies of their web business, so they’d rely on a service provider who would charge very low, or free, setup fees (DR marketers, often times to their demise, are really attracted to FREE), in exchange for a percentage (sometimes as high as 30-50%) of every order. This model works great if you have insanely high margins but it fundamentally doesn’t consider the internet as both an order taking medium and a customer acquisition channel. The customer who comes to your website (or finds it via google) because they saw your product on TV should be completely separated from the customer who found your product advertised on a weight loss forum and clicked through to your site. While paying 30% for the customer coming from the weight loss forum, it does NOT make sense to pay 30% for the customer who came to your site because they saw your TV ad which you already spent a lot of money on. This may seem like a trivial point but it is absolutely critical that you distinguish your customers by media acquisition channel and not only by ordering channel. Pay close attention to your metrics but be sure to do so in a way that allows you to make this important distinction.
Test, Test, and then Test Again. The internet provides quite possibly one of the inexpensive and accurate ways to conduct true A/B tests. When you test an A/B script at a call center you’re confronted with many potential variables (agent mix, training, compliance) but on the web you have the opportunity to normalize many of these variables and use consistent metrics to make comparisons. Take advantage of it! Don’t be afraid to test new offers, new creative, A/B site splits or e-mail splits and in the end roll those learning into your overall marketing strategy.
