This article is part of a series discussing what could be the perfect CSE for 2010. In previous articles, I discussed the opportunity for CSEs to implement a true product search, using product clustering for both hard and soft goods; and explored how to improve browsing through those products. In this follow-up, I would like to see how CSEs could better help users to decide where to buy.
This may be a simplification of the current situation, but I’m tempted to say that there is only one interaction pattern implemented by CSEs to let users choose a product and get a list of stores selling it: a. you browse products; b. you click on the “compare prices” button; c. you land on a product page displaying – among other information – the list of merchant offers.
And I would say that all those product pages look the same whatever the CSE is… The offers are displayed using a tabular presentation, with at least offer description / store details / price columns. And users are usually left with prices and store ratings to decide which store to buy from. There usually are few sorting options (usually by price / store ratings) and rarely any filtering option.
A classic product page at Shopzilla:
Such product page is ok when there are less than – let’s say – 20 merchant offers for a single product. Users can scan the list, compare and decide. Or they can directly compare the product pages on store websites… But, whatever the strategy, the price is rarely enough to take a decision.
Less than 20 merchant offers for a popular product? I feel the strategy from Twenga or TheFind to achieve comprehensiveness is the way to go. I believe displaying the most popular stores (or said differently: the ones that give money) is not enough: it is important for users to have an overview of the market for the product they are interested in (number of stores selling it, price ranges…) in order to choose a store.
For shopping engines that try to achieve a true comprehensiveness, the number of shops displayed on a product page can become overwhelming – i.e: 96 stores for the Canon EOS 500D on Twenga FR. It is therefore important for users to be able to refine the list by their needs:
Those are some simple questions that could make users choose a store or another; but unfortunately, those data, when available on CSE sites, are usually displayed on store pages, but rarely used as filtering options on product pages.
I wonder whether CSES are right when they only allow to choose stores when users have selected a product. Couldn’t it be useful to select stores when doing product filtering?
Anytime I want to buy English books or CDs, I go to Amazon. But the problem I have is that I never know if I should buy from Amazon FR, Amazon UK or Amazon US… I always check the three sites to see which one has the lowest price and the shortest delivery time… I don’t check Amazon DE or Amazon NL though as I don’t speak German or Dutch…
Isn’t it a problem a CSE could tackle? Even if I’m loyal to Amazon, I would be even more loyal to any CSE that would save me the extra time to know which Amazon site I should buy from. If this CSE could also warn me of custom taxes I may have to pay, I would be really grateful :)
Vasilis Dimos, from Skroutz, a Greek shopping engine, recently left a comment on this blog, suggesting the following:
Deeper integration into the checkout process. It is inevitable i think that in the future people will shop through the CSE, without leaving the site. Don’t know if any CSE (which isn’t a boxmover itself) has managed this so far.
Google has its Checkout program; Shop.com has tried to manage a shopping basket from certified stores (I have no clue if this program is successful or not)… As Vasilis suggests, this is clearly an area where CSEs could bring an extra value.
I think that solving the “where to buy” problem for customers is the raison d’être of CSEs. Still little to no improvements have been done in this area for years. As CSEs try to improve their user loyalty, bringing innovative solutions to this question may help.