Plain-spoken Online Conversion Rate Newsletter - covering web design, sales, marketing, copywriting, usability, SEO, relationship marketing and consumer psychology.
Bottom Up, Top Down, Measuring All Around
Strike up the band! Dust off those New Year's noisemakers! Today's the day we celebrate two whole years of publishing GrokDotCom online. So, as my anniversary treat to you, I'm doing things a little differently. One, big, juicy article on a subject near and dear to all our hearts: what the heck do I do with all these numbers?
You may recall Jim Novo. Not so long ago - December 1st last year to be exact - I had the pleasure of sharing with you some of his wry thoughts on tracking customer-level numbers that could actually benefit you (Déjà vu: Engineers and Customer Metrics). It's time to share some more of Jim. He's recently been working with Bryan Eisenberg, CIO of Future Now, to create a brilliant document on understanding and managing your web analytics (so I'm a little lacking in humility!). Along with our second birthday, it's information truly worth getting excited about!
A Little Background
by Jeffrey Eisenberg, CEO, Future Now, Inc.
In May of 2000, Future Now, Inc. added a new feature to their Web site: a free, downloadable calculator suite. Presented as an Excel Spreadsheet, this unique application allowed users to enter 8 numbers related to their Web site’s commercial and visitor activity, and generate calculations for 10 conversion metrics.
Since then, thousands of individuals have downloaded the Future Now Conversion Metrics and Digital Sales Calculator, and the application has grown to include 22 metrics that can give both commerce and content companies the kind of hard information they need to make informed business decisions in an online environment.
Bottom Up, Top Down
by Jim Novo
The Future Now Calculator takes the raw data input by a user from the standard WebTrendsTM report and turns this data into actionable metrics you can track over time. What do I mean by actionable? All the metrics created by the calculator tell you specific things about the way your visitors are behaving, and you can literally "take action" based on the metrics and track the changes in visitor behavior your action caused. These are not "Gee, that's interesting" metrics. They are the ones that have dramatic impact on the profitability of your web site, and by tracking them and trying to affect them, you can get your web site a lot closer to the goals you are trying to achieve.
There are two general areas worth your time to track and analyze - Content and Commerce.
Many folks get hung up with looking at page views, browser types, IP addresses, and so forth. These numbers might be “nice to have,” but does knowing gross browser types used by version actually help you in any way? This is data, but not information. There is a tremendous amount of “stuff” generated by server logs and log analysis software. How do you screen out what is valuable and what is a waste of time?
With any data tracking or metrics project, always ask yourself this: What will I do with the information once I know it? What kind of action will I take based on knowing this information? If the answer is “none”, then the data is not worth tracking. Reports with data like this are the ones that pile up in the corner of the office - and are never looked at. What you want to do is create metrics and reports that clearly say, “Fix this and your visitors will view more pages” if you are a content site or “Fix this and you will sell more product” if you are a commerce site. In many cases these two ideas work together - a good content strategy results in more sales and a successful sales experience leads to visitors spending more time at the site.
Every web site has at least two big issues to deal with: generating targeted traffic and making sure this traffic gets "hooked" when it arrives at the site. Server log analysis can create the raw data you need to track important metrics surrounding these issues of targeted traffic and engaging the visitor, but the raw data isn't very useful in the log analyzer format for these tasks. The raw data has to be converted to useful metrics you can track over time. This is the purpose of the content area of the Calculator: it takes raw data from server log analysis, sifts it for the most valuable data, and creates actionable metrics for you. This part of the calculator is set up to follow the WebTrendsTM log analysis reporting format.
What kinds of issues are addressed with Content Metrics? Optimizing navigation, layout and design, and content to drive the most profitable activity from your visitors. Content Metrics are also be used to track the ROI of advertising programs and tweak their effectiveness - for example, to test the conversion rate of different landing pages in Pay-Per-Click campaigns.
Additionally, there are Content Metrics associated with each level of your sales funnel. At each “micro-action” level of the sales funnel, you can use log analysis to determine your conversion rate. Improving the conversion at each level of the funnel is cumulative - an improvement in one level dumps more customers into the next level - and all of these conversions build to higher levels of productivity and profitability. If you run a commerce site, they also feed directly into improving your Commerce Metrics - the bottom line.
Commerce metrics are about taking all this micro activity on the site and putting sales dollars to it. If you are selling products, it’s a way of determining how efficient and effective you are at converting visitors into sales dollars, providing clues for specific improvement in your bottom line.
Why do you need Commerce Metrics? Because it’s possible you could improve the number of pages viewed by each visitor using your Content Metrics, but this improvement would not result in increased sales. Perhaps there was a disconnect somewhere. The Commerce Metrics allow you to make sure any work you are doing on the content side is turning into profits on the commerce side.
If you are paying for advertising, Commerce Metrics are extremely important. You want to make sure the money you are spending is generating maximum profitability. If you pay $30 for a sale you make only $20 on, you will be in trouble pretty quick. Commerce metrics keep you on track and provide benchmarks you can follow to evaluate the profitability of your advertising campaigns.
Macro vs. Micro
Each of the metrics on both the content side and the commerce side can be further subdivided, and this is almost always instructive. There are no “average visitors” - there are specific visitors from specific places and you should subdivide your macro metrics into more micro metrics.
For example, on the content side, let’s say your “average visitor” looks at 3 pages in a visit. But what if you knew that for a Google visitor it was 8 pages and for MSN it was 2 pages? This information is tremendously important to understand because it can directly impact your advertising strategy and profitability. It is always easier to go with the traffic flow - go where you get the best visitors, don’t try to turn bad visitors into good ones. In this example, I would seriously consider cutting back on any bCentral network banner ads and using the money I saved to buy Google AdWords. This is optimization of an ad campaign - putting the money where it works the hardest.
Of course, on the commerce side, I would want to know if these Google visitors were generating sales - and if the sales per visitor was higher for Google than it is for MSN. Just because they view more pages, doesn’t mean they buy more products. So before rearranging my ad campaign, I would need to confirm using Commerce Metrics what I suspected from looking at Content Metrics - that indeed, a Google visitor not only looks at more pages per visit but buys more products.
Top-Down vs. Bottom-Up
Generally, the higher you move up in an organization, the more direct financial responsibility you have. For this reason, “C Level” execs and VP’s are probably going to be most interested in Commerce Metrics (if you are a commerce site) - a “top-down” view of the enterprise. They don’t necessarily care about the 8 pages per visit from a Google visitor - that’s too much detail for them to get into.
But hear this - the bottoms-up (micro) metrics all roll up into the tops-down (macro) metrics, so even though the top brass might not want to get reports on micro metrics, people lower in the organization should. Why? Because what will happen when a C-level person wants to know "How come a certain macro level metric is falling?"
Using the Google/MSN example above, if the average views per visit on the site fell during a certain month, one explanation could be visitors from Google dropped while visitors from MSN increased. If you know a Google visitor both views more pages (Content Metrics) and buys more products (Commerce Metrics) than an MSN visitor, you can tell this C-Level exec what happened and why. Then she’s off to the ad guys to find out why they’re driving more traffic from MSN than Google rather than breathing down your neck. Just because C-level people may not want micro-level reports, doesn’t mean it’s not extremely valuable for somebody to get reports on the micro level.
See what I mean? Totally cool stuff! Okay, it's a bit of a shameless plug, but indulge me. You know I don't do this often … and only when I think it will make you really happy! If the stuff in this article makes sense to you and you'd like to learn more, toddle off to HiqHq.com and get yourself a copy of The Guide to Web Analytics: How to Understand and Use Your Web Trends to Maximize Results. For a limited time, we're making it available to you this way!
Now excuuuse me please … there're candles to blow out and a cake to cut. Belgian dark chocolate, no less! (chortle)
We will never give, lease or sell your personal information. Period!
We strive to only send e-mail to those who want to receive it.