elise.com: On the Job: Market Research with Alexa.com

Attention marketers and market researchers! Amazon.com's Alexa has teamed up with Google technology to deliver amazing free traffic and ranking statistics for all websites at www.alexa.com. By monitoring the site viewing activities of a million or so web surfers who have installed Alexa's free toolbar in their browser, Alexa tracks traffic, page views and visitors, ranked and over time for every site on the web. Alexa can tell which other sites are linking to a particular site. It also can tell you other similar sites that a particular website's visitors have viewed. This data is a treasure trove for anyone researching markets, companies, or industries. The traffic data is relative, not absolute, but still quite useful for evaluating a company vis-a-vis its competitors. One thing to note is that the Alexa toolbar only works with Microsoft Windows and Internet Explorer. So given that Mac users are left out of those whose web movements are being tracked, the Alexa data may not be as representative for Mac product companies.

Relationship of Inbound Links to Site Traffic

One of the outcomes of Google's rise to dominance as a search engine is that sites that have more inbound links — other sites linking to it — get ranked higher in search results and therefore get more traffic. (You can read more about this at: Search Engine Optimization). But what exactly is the relationship of inbound links to site traffic? I did some quick analysis on a random group of consumer-focused art sites and found the following results.

Note that both axes are logarithmic in scale. I've included a trend line so that it would be easier to see the slope. It looks like for every 10 fold increase in links, there is a 40 fold increase in site traffic rank.

Thinking that websites oriented towards technology might have a slightly different relationship of links to traffic, I also ran the analysis on a random group of technology sites.

Note that the spread of data was wider with this group, so the scales are not equivalent from the previous graph to this one. In this group, it looks like for every 10-fold increase in links there is a 10-fold increase in site traffic.

Superimposing the two graphs yields the following.

The slope is lower for the tech sites. What does this mean? One interpretation is that tech-oriented sites link to each other more than consumer sites. Also, blogging, which is still mostly confined to tech-savy folks, encourages a lot of outbound linking, resulting in overall more inbound links if these tech-savy folks are linking to tech sites. If we assert that inbound links have a causative impact on site traffic (due to higher Google rankings), then it will take more inbound links for a tech site to have an increase in traffic than it would take a consumer site.

The caveats are of course that I'm using a small data set - about 30 data points for each group. And my random picking of sites isn't scientific. I just picked sites in a certain category and didn't look at their data before deciding whether or not to include them in the data set.

Why oh why go to all this trouble? or What's the point?

Is it once a BCG consultant, always a BCG consultant? Perhaps. I just get curious about things and then want to go run some anaylsis. Here's the point. There is lots of interesting data at Alexa.com. Check it out.

This article is written by Elise Bauer and licensed under a Creative Commons License with some rights reserved. If reposting this article on a website, please host all graphics on your own site and link back to this article at http://www.elise.com/web/.

Posted by elise on January 30, 2004 11:52 PM
Copyright © 2004 Elise Bauer. All Rights Reserved (unless otherwise specified). http://www.elise.com/web/