Product population

I am currently involved in an e-commerce project for a medium sized electronics retailer and we are having difficulties with product population.

As with most companies of this size their current stock system is quite old and constructed for use by the high street outlets, therefore the information it contains is not in depth or accurate enough for use in the site.

Product codes and costs are not an issue, description and images are giving us the dificulty.

Obtaining the information is not an issue but the information we require is a moving target as the stock that will appear on the site will be determined by a certain stock level. These stock levels are constantly moving as the stock profile for the warehouses that will supply the online store are under constant demand from the high street outlets.

Our fear for example is to have collected information for 2000 products only 500 of which are in stock when we go live but not have the information for a further 1000 that are in stock. Therefore end up with a site that is vastly under populated

Collecting and inputting ALL possible product information would not be viable and would take enormous resource.

I would be interested in anyones advice or experience in product population on this scale.

Regards
Sean Clark

Hi Sean

This is a very interesting area. Not that I'd suggest 'product popluation' as a matter of conversation down the pub, but it is in my view interesting for the following reasons:

1. Good product information increases e-commerce sales

It increases conversions and raises average basket sizes. I can say that with a fair degree of confidence because I've seen the management information of enough major e-tailers to be convinced this is the case. Equally, if you read the comments from e-business managers at the likes of Virgin Wines, John Lewis Direct and HMV (see NMA's e-tail forum) this if further confirmed.

There's 'added value' product-centric information (like user reviews, product demos, buyer guides, comparison tables etc.) that is very powerful in making the sale (especially for high ticket or complex items) but even basic product info is very important. Interestingly, you mention descriptions and photos - in my experience these 2 'basic' product info fields are *the* most important for conversion, for obvious reasons I think. I know of e-tailers who have performance metrics and targets based on the quality, even existence, of product data fields within their product catalogue databases.

2. Good product information saves costs

What are the majority of calls to e-tailers call centres about? Customers enquiring about products. They want more details, or perhaps some information was missing, or confused them etc. If you have good product information then the number of calls goes down and you save money.

3. Good product info is hard to come by and it risks getting worse

Now that we've got XML and web services, wouldn't it be nice to just 'plug and play' that product info into a site from a central source? Seamless integration, robust, up to date. "You're an e-tailer going cross-platform (most of the big ones are...)? Don't worry our XML product info has <mobile>, <kiosk>, and <itv> sub-sets, specially targeted to work on that medium..."

As we all know this is not currently the case and problem you point out is precisely the pain that is there at the moment - old systems, inadequate data, poorly structured data, out of date information, nightmarish systems integration headaches etc. Plenty of work for the consultants there...

Boring old 'product information' is interesting to me in as much as everyone is chasing after the likes of CRM and though there's nothing wrong with that there is so much still to be done, and so much to be gained, by concentrating on delivering good product information online.

None of that directly helps you I guess but I would say:

1. Whatever product info you *do* have must be good (the description and photos in particular I fear)

2. (Obviously) you cannot have a situation where customers are able to buy out of stock products online. E-commerce disaster.

I don't think customers will notice a decrease in range available online but they will certainly notice if the quality of what is there is poor, inconsistent, or the site is lying about stock availability. So I would concentrate on a sub-set of core products that you know will be in stock and do them really well. Then you can build out from there.

Ashley

Hi again Sean

In relation to your original post, thought you might be interested in a white paper we've recently added to the site titled "Product Data Integration in B2B E-Commerce".

You can find out more about it at http://www.e-consultancy.com/knowledge/whitepapers/whitepaper_view.asp?id=637 - it's probably too theoretical / technical for your purposes but could be of interest? If you're still hungry for more on product data ontologies and integration challenges then try the links from http://gunther.smeal.psu.edu/nrelated/0/4037

Ashley