Keyword Metrics and Accountability

What do you think of when someone drops “keyword” into the conversation? I immediately start seeing dogs going at each other demolition derby style – a real dog-eat-dog extravaganza! Keywords are critical, and they are big business.

But you only want to dig deep into your pockets for those keywords that are going to earn their keep. How do you identify the hard workers? John Marshall of ClickTracks contributed this information about the “average time on site” metric for our book, Call to Action.

John writes:

ROI is a blunt instrument

Is ROI is the only measurement (for keyword effectiveness) that really matters? While ROI is indeed important, it suffers from being too crude in many instances, especially for search engine marketing. This is a pretty contentious thing to say, so let's look at an example:

A store selling a range of gourmet foods is buying keywords for Olive Oil. When doing ROI analysis this keyword shows poor results; that is, lots of clicks but very few sales. A closer investigation however might reveal that the olive oil comes in ornate glass bottles - attractive, but heavy and expensive to ship. The lack of sales results from the high shipping costs, not the keyword itself.

Don't blame the messenger

Search keywords often correlate directly to a product. Lack of conversion may well be related to the product itself or an attribute of it, and not the keyword itself. Proof of this for the example above would exist by examining the abandonment rate from the shipping confirmation page of the cart. Unfortunately extracting this data for each keyword, for each product will most likely be very time consuming, even if the analytics app can support it.

A different metric - ATOS

What's needed is a metric that can quickly be applied to all keywords and that indicates propensity to buy instead of actual purchase being made. In our experience this best single value that indicates this is 'average time on site' or ATOS. As an absolute value it doesn't mean much, and even expressed as a trend it's not a lot better. It could be argued that a long time on site is a bad thing because it indicates the site has poor usability. However, when viewed across search engine keywords, or segments of visitors, or different campaigns, a longer time indicates greater interest in the product/website. Since we assume usability problems are the same across all keywords the metric is valid when used for comparisons like this. People do not waste time browsing for products they have no intention of buying. They may decide against purchasing because of cost, shipping, return policy or phase of the moon, but that's not the fault of the keyword. ATOS is the single most efficient way of proving this.

Making Sense of Complexity

What with search engine marketing, usability, information architecture, copywriting, web analytics, consumer psychology and all the other specialties that go into creating online efforts that maximize your conversion rate, it can be difficult to see the forest for the trees.

That’s where Call to Action comes to your rescue. It’s our way of helping you understand and synthesize exciting, relevant information so you can design persuasively. Are you ready to start making sense of the complexity?

Volume 111:5/15/05


Contact Us


Bi-Monthly Online Conversion Rate Marketing Newsletter:

We Value Your Privacy!

  • Cardscan
  • ProFlowers
  • GE
  • Leo Schachter Diamonds
  • Overstock.com
  • SAS
  • Disney
  • Gatelys.com
  • NBC Universal
  • PriceWaterhouseCoopers
  • Volvo International
  • Southern Company
  • BuyTelco
  • Allegis Group's Thingamajob
  • WebEx
  • CafePress
  • Elance