A new service using data from Alexa, Competitious provides an interface for managing your information resources about the competition. It sounds like a pretty exciting possibility for search marketing: organize your resources and keep a close tab on the services, popularity, traffic rank, and buzz surrounding your fiercest competition.
Although it is, essentially, a fairly simple interface for tracking competitive information which is easily available, the ability to push all that information together is certainly a worthwhile service. From a consultant’s perspective, it’s worthwhile because you can create multiple projects to track: keep your eyes on each client’s area at a glance.
Naturally, my first thought is about adding more features: for example, tracking ranking reports using your selected keyword list and checking against your competitors’ ranking. I know, I know…ranking reports are practically worthless. However, knowing how your own search performance holds up against your competitors’ is still valuable: and this would be a relatively easily automated tracking tool.
It’s interesting: I know perfectly well that Alexa data is practically worthless by itself. However, between a set of sites all in the same market (say…competitors), the relative performance data may still convey some usable information. It’s not the numbers you need to look at: it’s the relationships.
Statistics are an unending struggle for internet marketing. Understandably, clients always want some hard facts to demonstrate that their money isn’t going to waste – but what numbers do you give them?
Web site traffic statistics are famously variable in interpretation. Since no traffic service has a handle on the actual statistics for all website traffic, the numbers are usually based on particularly selective data sets. Rand Fishkin gave his detailed report yesterday on Alexa and Hitwise data, comparing them to the data provided through Feedburner and Indextools. The numbers tell it all – different statistics services provide vastly different data.
Imagine that all the webdev category sites receive 9 million uniques per day.
Pimpyourpro.com – 9 mil x 2.07% = 186,300 visitors per day
SEOmoz.org – 9 mil x 0.01% = 900 visitors per day
In reality it should be something like:
SEOmoz.org – 9 mil x 0.085% = 7650 visitors per day
7650 visitors a day is based on SEOmoz’s real visitor traffic. The other data is based on percentage of traffic as reported by Hitwise – see a problem?
So, given that the easily available data is, for practical purposes, only usable as a very general guideline, what kind of data should you actually report to your clients?
Recently, this question was asked at Cre8asiteForums. The answers vary – but the essential focus is that the only meaningful statistics to report must be based on the site’s business goals.
- Focus reporting on specific metrics: income earned, referrals, registrations, whatever is considered a success for your site’s conversions.
- Establish objectives for the campaign. The more specific the goal the better – and be realistic.
- Don’t depend on any statistic that you need to convince your client is relevant. If they don’t understand that tracking campaign revenue is relevant, it may be that you don’t want to be working with them!
- Identify the characteristics of a successful conversion. Differentiate between visitors who purchase the product and those who don’t. Determine everything you can which is different about these two key groups of visitors.
Pure traffic has some place in reporting – even if the traffic isn’t converting, a lot of traffic means greater exposure for your web business. But pure traffic numbers shouldn’t be the highlight of your statistics analysis. It’s worth mentioning any increase, but the bottom line is not directly related to visitors – it’s all about sales.
You don’t get much better than Danny Sullivan’s article on July’s comScore figures. The article provides a fantastic overview of search engine statistics on the whole, comScore’s figures in specific, and makes a valuable point about the care it’s necessary to take when looking at any interpreted statistical data.
Read the report yourself, then decide whether Google has actually lost "market share" – and, more importantly, whether it even matters!
Do not put faith in what statistics say until you have carefully considered what they do not say.
William W. Watt quoted in "Medical Statistics: A Guide to Data Analysis and Critical Appraisal" by Jennifer Peat and Belinda Barton