Is site search less important for niche retailers than larger ecommerce sites?
It's an interesting question, and one which came up when I was moderating the Site Search and Naviagtion roundtable at Digital Cream last week.
However, when it is used by visitors, it more than pays its way...
Screen Pages looked at onsite search usage on 21 ecommerce websites, most of which are niche retailers selling own-brand goods, including Loake, Finisterre, Christopher Ward, Micro-Scooters, Hush and Dove Spa.
A few were "boutiques" selling goods from known brands and suppliers.
The number of products on offer was typically in the hundreds and low thousands. The websites share common constrictions. They are running on the Magento platform and use the software's standard search facility.
The analysis was conducting using the search metrics from Google Analytics.
- The average usage of onsite search was 5.75%, within a range of 18% to 1%. The two companies at the higher end of this range sold products (footwear) where size, style, colour selection is an important factor in decision-making.
- However, when used, search performs very well: visitors converted at 4.63% versus the websites' average of 2.77% - that's 1.8 times more effective.
- Consequently, visitors using search contributed 13.8% of the revenues.
- That said, there is room for improvement. 20% of people who used search went on to refine their searches (submit another search) and 21% exited the website from the search results.
Screen Pages also looked at the use of filters on these websites. Magento includes "attribute filtering" as a standard feature, where retailers can choose which features/fields can act as filters, generally as part of the left-hand navigation on category pages.
Whilst the average usage was 6.6% of all pages viewed, there were two distinct groups of websites. On 10 of the 16 websites, pages where filters were less than 3% of all pages viewed, indicating that browsers prefer other methods of navigation.
On the websites where filters were used by more than 8%, filters were used prominently alongside categeories more or less of equal visual weight (as opposed to a set of options below the main category menus).
Percentage of people using search and the percentage of revenues influenced by search.
Whilst much is made of onsite search, for niche brands in particular site search is not as important as other forms of navigation such as browsing via menus or actions driven by onsite merchandising (banners) or offsite marketing (email and search engine marketing).
When used by shoppers intent on finding the "right" product, search punches above its weight in terms of sales and conversions.
Given this data, it is wholly appropriate that retailers of these characteristics avail themselves of the "standard" search technologies provided by platforms like Magento, as opposed to investing in third-party search engines at additional cost. But efforts can still be made to optimise search performance.
Recommendations for quick wins include:
- Improve the merchandising and messaging on search results pages. Many retailers leave these pages as dead-ends. For example, show popular products and searches explicitly, with links to the relevant pages: this will reduce exit rates from one in five.
- Study search terms used and improve overall navigation to capture popular searches before they are needed. One retailer's navigation was based around branded "collections" as opposed to product types, driving visitors to search for these "absent" product types.
- In Magento, retailers can determine which fields/attributes are searched: content managers should define a set of keywords for each product which can steer search results. Specific searches can also be re-directed to distinct URLs (e.g. "returns information").
Screen Pages is hosting its 5th E-commerce Forum on 2nd May 2013 in London. The event is targeted at niche retailers and will focus on ecommerce best practice, case studies and trends. Speakers at the event will include: Donald Russell, Atterley Road, Piper Private Equity, Nomensa, Predictive Intent and Magento.