Although this site has been produced for specific courses and groups of students it is designed as a public resource. If you find it useful then please let me know.

If you want to comment feel free to do so and if you find something wrong get in touch.

hide alert

Can we predict the revenues of an ecommerce site?

Written by: Jonathan Briggs

August 29, 2005 [3078 views]

I am currently engaged in a research project at Kingston University to look at whether we can estimate the success of an ecommerce venture by looking at publicly available statistics such as PageRank, Pay-per-click advertising (PPC) costs and click-through rates, and traffic estimates to estimate the likely revenues of ecommerce sites.

The project is planned to take up to two years and involves my company, the OTHER media, too. As we have developed a number of successful retail sites ourselves including Paul Smith, Virgin Experience Days and Christmas Hamper site Food Parcels we can use the known revenues from these sites as a benchmark.

We suspect that customers increasingly reach some sites through search engines rather than because of a known brand. The likelihood that they will find this particular site, as opposed to similar sized competitors, can be predicted by PageRank. It is also possible to estimate the number of visitors to some sites using traffic monitoring services such as Alexa (and soon perhaps Google).

PPC pricing may also prove to be a predictor of revenues particularly for established sites willing to spend a fixed percentage to acquire a customer by comparison with average order value.

The project will make comparisons across a range of retail sectors and involve telephone surveys with retailers in each sector. In addition sites will be rated against factors that might reduce conversion of visitors into customers. These might include onerous registration processes or lack of decision support for customers. Finally we will include published results from Companies House to confirm the scale of revenues being generated.

The principle goal of the research will be to produce a model with the smallest number of measurements that can produce an acceptable prediction of scale of revenues. As a first approximation we will attempt to predict those sites generating more than £10 million annually, £1-10 million, £500,000-1 million, £100-500,000 and less than £100,000.

While it is not expected that we can generate a formula for accurate estimation we are aiming to be able to group companies by their expected revenues. This would be a useful tool for competitor analysis within the ecommerce marketplace.

If you would like to contribute to the research (perhaps with figures from your own retail store) and receive a full report of its findings then contact the author. Intermediate results and discussions will be posted here from time to time as they become available.

What do you think?







Add your comments