Argos are one of the largest catalogue merchant here in the UK, and offer a variety of goods with everything from Sofa Beds to Swing balls. In each of their retail stores, they have physical terminals and kiosks setup so you can check stock in store via a unique reference code from the catalogue. This same code can also be used online, again to check stock from the comfort of your home before either purchasing online, or reserving items instore.
Although it is a relatively clinical shopping experience for the consumer, its an effective one for the retailer, and facilitates a vast range of items to be offered to the consumer, whilst maintain smaller shop footprints than their competition without as much risk of shoplifting.
Argos are also a champion at long tailing offline in their physical stores.
One of the many advantages that Argos has, are in its implementation of stock checkers. The entire business circles around that 6 digit code that you enter to find if something is available. Without realising it, as you and thousands of other shoppers check stock daily, you are sending a signal to a server somewhere that helps determine real time demand.
If the systems fails to find anything or an item is out of stock for an arbitrary number or more people in any given day, then its a pretty strong signal to get that item into stock and rolled out across the stores ASAP. The kiosks offer a real time feedback engine that allow central office to stay abreast of consumer demands, and respond accordingly all whilst maintaining that small footprint instore.
Your real time feedback engine is your own website.
Everyone that runs a website has similar trend data at their finger tips, often without realising it. If you have a 404 page, a search box or any other ‘no results’ or ‘failure’ page, then you too have the capacity for finding consumer trends and demand on your website. For e-commerce stores online, this is a data gold mine, which all too often is buried under a mountain of keywords in a log file somewhere.
If you aren’t paying attention to the pages your visitors can’t find, or you haven’t yet created, you are leaving money on the table and losing visitors. Often the highest bounce rates occur on failure pages such as these, and offer insight into exactly what people are trying to buy, what articles they are trying to read and can provide useful data overall on how to better serve visitors.
If you’ve a site search box or equivalent – setup this. It will take you 30 seconds max. Seriously. Better still, on a no results page log the exact referral keyword or phrase used in the failure page directly inside Google Analytics when it occurs. The following snippet can be added to your GA code to fire a custom event to do just that.
'_trackEvent', 'Failed Result', 'Search Results', 'Green Widget with cherry on top'
'_trackEvent', 'Failed Result', 'Out of stock', 'Green Widget'
Once in place, you’ve at least started to see where your site isn’t serving a good experience to visitors, and take steps to put it right. You can go even further with this by logging failure keywords and phrases directly in your own database.
Whilst this might seem like overkill, and not worth the effort now – you’d be surprised at how useful this sort of data can be when understanding your overall performance over time, and indeed building out an automated, data driven solution to the problem in the future.
For example, you could compute that many searches which currently fail often bear a close resemblance to other products in your range, and offer these as alternatives. You could choose to capitalise on failed queries, and substitute alternative website links when they occur (perhaps generating an affiliate income as a result). You could send yourself visitor failure reports monthly to see how the site is performing over time, and measure performance. Or, if you are an e-commerce outlet like Argos – build much more realtime stock handling and shape your store around demand.
Google and other web giants have gained significant competitive advantage in the past number of years, with similar techniques, simply by recording the minuta of failed results.
The possibilities are only as limited as your imagination, and considering we are in the era of Big Data, with large cloud based storage more and more accessible to every business, failure data is priceless information that you can’t afford NOT to store.