Product images in site search
Predictive or suggestive search is rightly held up as a must-have for ecommerce retailers that have many categories and products. What is often not mentioned is how helpful imagery can be in search.
As you can see below, the Screwfix search functionality includes both a suggested text listing and several suggested products with images (much like mini product listings).
This allows for much better product recall when customers are trying to find something they have viewed before (and may reduce the number of mistaken clicks from search).
These panels can also be personalised, using my behavioural history in combination with popular products and, most importantly, the text I have entered into the search bar.
Shop by persona
Screwfix already includes hundreds of categories and possibly thousands of filters to allow customers to find what they want without headache.
However, Screwfix doesn’t stop there in its attempts to help users cut through the noise. Two of the main categories (tools and safety & workwear) include the option to ‘shop by trade’.
This means builders, mechanics, decorators etc. can exclude products designed specifically for other trades.
Though there is much hype around personalisation, with the associated danger of funnelling customers away from products they may well be happy to discover, the process of thinking about customer shopping personas is a valuable one.
In fashion, this could be as simple as allowing customers to filter by style (e.g. Boho), rather than by product type.
Rich product listings
Catalogue retailers have a mammoth task transferring great UX to a website.
The enormous number of disparate products with such detailed specifications is a nightmare for categorisation and navigation (see this great case study from RS Components).
One of the ways of overcoming this is by including as much detail as possible in filtered product listings.
Look at the screen shot below. Three bullet points are used to sum up the key features of each power drill. This mini product description is perfect for quick comparison (with a compare tool available, if more detailed weighing-up is required).
Click-and-collect and next-day delivery are surfaced as buttons on the listing, allowing users to jump straight to these options (rather than clicking to a product page and then clicking again). The next-day delivery button is changed appropriately, when products have longer lead times.
Number of reviews and average rating is displayed, price is given accurately (wth option to turn VAT on or off in the header menu), and icons are used to show ‘trade rated’, ‘new’ or ‘Screwfix exclusive’.
All in all, these listings are designed to allow me, the user, to skim as lightly as possible over the catalogue, before I need to dive into a product page and out again. This saves time and means I’m less likely to abandon my search.
Product page video
The drill product page shown below has a video button that launches a Screwfix YouTube channel frame in window.
This is exactly the sort of content a customer wants for medium- to high-value purchases where spec and reviews are very important.
Product page video is not needed for every listing (not much use for a screw or a light switch), but retailers in any industry should prioritise important products, with video a tool for improving conversion rates.
Filter best practice
This should be dyed in the wool by now for any online retailer, but the elegance of filter best practice is demonstrated so well by Screwfix, it would be remiss of me to remind myself.
Filters should be collapsible, with each filter option showing the number of products within it. Pricing should be filtered by pre-set ranges, but also a custom range that the user can set.
There should be the option to ‘clear all’ applied filters in one click. Filters should include customer reviews and attributes entirely relevant to the category in question (in the instance below, lawnmowers can be filtered by collection capacity, propulsion type and cutting blade size).
One thing that Screwfix doesn’t do is highlight a filter heading when an option has been checked within, however this is only a problem if the user applies a filter and then collapses the filter menuin question.