Neil Thackray’s Business Media Blog

Just another weblog

Personalisation of Content as a Route to Revenue

Image representing Cognitive Match as depicted...
Image via CrunchBase

In earlier discussions on this blog we have summarised the two big challenges facing business media in the online world as being user engagement and the development of an effective advertising model.   I recently had the pleasure of being introduced to Cognitive Match which goes some way to addressing both these issues.

Imagine that the publisher has a set of content parcels, each of which may be of interest to different parts of the community.  Let’s create a real example by pretending that we are the owners of a food and wine web site for for the catering trade.  Let’s imagine that we have written a series of wine reviews.  To keep it simple let’s assume we have written just four.  One is a review of fine champagne that sells at £50 a bottle.  The second is a good quality Merlot that sells for £25 a bottle.  The third is a dessert wine, a muscat perhaps and the fourth is a blended £5 a bottle Sauvignon.   Let us further imagine that we have cleverly sold relevant advertising against each of these wine types.

We are so excited by these wine reviews and our potential revenue (which has all been sold on a CTR basis) that we devote  a third of the site page area above the fold to it on every page of our site.  All we have to do is to decide which wine review to show at any given time to which user.   Now this where Cognitive Match gets clever.  Using some mathematics which I am not going to try to explain, the Cognitive Match engine collects annonymous informaion about each user and shows, on the fly, the content most likely to be of interest to that user.   A user looking at dessert ingredients content might be shown the Muscat, whilst someone who had looked at a job advertisement for sommelier in a Michelin star restaurant might be shown the Champagne.  As with all search solutions this is about probability mathematics.  If the probability that the content is relevant and personal to the individual user is increased the value of that engagement with the user is enhanced.

For retailers, Cognitive Match claim that the basket attrition rate will fall dramatically.  For business media companies, the ability to match content to user interests increases the chances of an ad click and is likely to encourage the user to spend some more time on the site.  It provides a key to unlocking profitable  CPA deals too.  We have all done CPA deals, but how many have exceeded our expectations?  Matching the right offer to the right user is clearly a useful approach.

Matching content to users is a famously hard trick to pull off.  Ask my team who built foundography (a vertical search engine) or anyone who has experimented with vertical semantic or intelligent search.  The Cognitive Match team appear to be close to a model which is easy to deploy for their clients.  It has an impressive academic team developing the application and an interesting pipeline of blue chip prosepective clients. I am promised a live demo in the near future.  I’ll keep you posted and let you know if it does what it says on the tin.

Reblog this post [with Zemanta]

April 15, 2009 Posted by | Advertising Sales, business media strategy, Search | , , , | 1 Comment