blog

When I recommend Recommendation Systems… And when I don’t

Thu 30 April 2009
  • Email to a friend
  • Print this article
  • decrease font size
  • increase font size

Following an AIMIA presentation a few months back on ‘Recommendations Systems’ – where I effectively waived businesses away from implementing such Systems, or not at least until they had the bigger priorities on their websites sorted – I was interviewed by NetRegistry’s NETT Magazine on the topic.

 

In the true sense of the defintion, Recommendation Systems refers to technologies that make real-world, intellegent recommendations to users of websites, based on past-behaviour and smart predicitions.

 

For instance, if a website sees you looking at the Die Hard DVD set, it might assume that you are probably male, into 1990’s police/comedy/action and would probably consider DVD box sets for similar TV series – say Seinfeld. It would then look at what other people who purchased the Die Hard DVD set purchased, the time of the day, the rate of the Thai Baht and a million other factors and present you with logical, beneficial and relevant recommendations.

 

“If you like the Die Hard DVD set, you will also like the Lethal Weapon DVD set while separately, there is a limited-time special on the Die Hard X-Box Game for only $29.00 when you purchase any three DVDs with free shipping in your area.”

 

There’s just no question, these systems are great. I don’t need to go into the benefits, though it’s a win-win for everyone; the website, and the customer. You sell more, they’re happier. And Recommendation Systems are by no means limited to eCommerce.

 

The issue is that pushing these technologies onto most corporate websites, who haven’t yet grasped who their users are, let alone what these users are looking for, just isn’t advisable; get the basics right before biting off even more.

 

In my interview (and in my presentation) I make the point that it is quite feasible to engage in the benefits of Recommendation Systems (i.e. making good recommendations to users, holding their hand through the website) without needing to actually invest in the technology. Analyse each page, who is reading it, and try to work out what might interest them next… and then provide those links.

 

Of course, there comes a point where it makes complete sense to invest in a Recommendations System technology, though if the time is right, you’ll know it… and not before.

 

The article and my interview is at http://nett.com.au/technology/ecommerce/you-might-also-like/11377.html.

 

Robert’s Recommendations
If you liked this article on recommendations systems, you may also be interested in:

1. Other articles by Robert Beerworth

2. Subscribing to NETT Magazine

3. Contacting Robert to see how Recommendation Systems might work on your website

 
posted in: BusinesseCommerceTechnologyWeb 2.0 Bookmark and Share
 

Got something to say? Post a comment about this blog below.

GET THE NEWSLETTER

Find out about new events

On the 3- 6 April 2009 I presented in Melbourne at APEC Conference on Government, Intellectual Property and the Internet. Read more about the APEC Melbourne Conference.


On the 17 Feb 2009 I presented at the  AIMIA Web 3.0 Gets Personal on recommendation systems and how they can be tailored to our personal tastes.  Read more about the AMIMA Event.


I presented on the 15 Jul 2009 at the  AIMIA "Playing by the Rules"  covering such topics as legal risk and compliance issues on the website.   See the presentation


I presented at the Online Retailer Expo and Conference, in Sydney on the 18th-19th of August 2009See the presentation.


 
Your Name:
Your Email Address:
Friends Name:
Friends Email Address:
Email Friend Close

Sending ...