Web analytics has become something of a buzz word. You overhear mention of it on street corners, at parties, in the bathrooms of fashionable restaurants. Trend setters refer to it in stage whispers.
Why? What does it matter?
If you’re paying for search engine ads, there are good reasons why it matters. You can’t do PPC (pay-per-click) advertising without analytics. At least, you can’t do it well.
To quote Ned Pepper in True Grit, “I call those bold words for a one-eyed fat man.” Admittedly, it's a bold statement. Many consider their PPC campaigns successful without the complication of web analytics. If they’re using analytics at all, it’s probably to track broken links, page views, and organic search engine referrals. There’s nothing wrong with that but it’s not nearly enough to remain competitive.
CTR (click-through-rate) is a measure of your cost, conversions a measure
of your profit, but web analytics measure how you
move from one to another.
PPC advertising is so successful because it’s relevant to the customer’s intent. Customers don’t browse search engine pages like they might stroll through the stacks of a library. Serendipity plays no part here. They’re pursuing a goal even if they don’t yet know how to articulate that goal. The closer you come to meeting that goal—aligning with their intent—the more successful your PPC campaigns become. Web analytics is about understanding your customer’s intent.
Optimizing PPC with Analytics
Let’s put it another way. Early in the game most PPC advertisers optimize for CTR. As they become more sophisticated, they optimize for conversions. CTR is a measure of your cost, conversions a measure of your profit, but web analytics measure how you move from one to another. You can’t get better at moving the needle between cost and profit without analyzing data, making assumptions based upon that analysis, and testing those assumptions. Increasing the efficiency of your search advertising campaigns isn't an option; it's the cost of remaining in the game.
You’re paying for every click but 20% of the people
clicking your ads may be leaving your landing page
seconds after arriving...a wasted expense.
Say you’re selling widgets. You’re paying for every click but 20% of the people clicking your ads may be leaving your landing page seconds after arriving. You never had a chance with those visitors. They weren’t your customers—weren’t interested in your product, your services, or your offer. They were a wasted expense, a mistargeted ad campaign. With analytics you’ll know 20% of your paid visitors are bouncing off your landing page like an India rubber ball and you’ll know which ones—which campaign they’re coming from, which ad group, which keyword, which ad. You can then begin refining your campaigns to reduce your bounce rate—rewriting your ad copy, adding negative keywords, targeting by age, gender, geography, time of day or day of week. You can test the changes to see what works. If the bounce rate decreases, permanently incorporate the change. If it doesn’t, learn from the experience and come up with a different hypothesis.
Inferring Intent
From the landing page, your visitors’ next click can tell you a great deal about who your PPC campaigns are appealing to and where they are in the buying cycle. Let’s diagram a typical search engine return page (SERP) below.
Are they in the early stages of researching widgets, further along and comparing features, or near the end of the cycle and shopping prices? Have they made their decision but want to be assured you’re trustworthy? Or do they simply want to know your dealer locations? Knowing more about our customers’ intent can help you target your campaigns, provide relevant content, and ultimately increase your conversions.
Web analytics is rather like reading animal tracks in the mud beside a stream. There’s a great deal to learn but it takes some analysis and some patience. First you’ll create a hypothesis to explain the data, test the accuracy of your hypotheses by making a change in your campaigns or on your site, then watch for a change in the data that confirms or disputes your hypothesis. It’s the commercial version of the scientific method. The advantage is that you’re dealing with data rather than opinion.
If you’re new to web analysis, I’d recommend Web Analytics, An Hour A Day by Avinash Kaushik. The author’s style is accessible, the material well organized, and the recommendations actionable whether you’re just getting into analytics or a gnarled old hand.