The behavioral analysis of Internet users, during a session on an e-commerce site, for example, shows significant use of the search engine.
On average 40% of Internet users use the search bar during their navigation. The tendency is accentuated as soon as the site proposes a catalog rich in references. In the age of merchant sites with an increasingly complex product referral, not to mention those that combine a marketplace, it becomes difficult to navigate to the desired product without exceeding the rule of 3 clicks.
The internal engine becomes a must for all Internet users, and a priority for e-commerce sites. Well optimized, it can be a sales accelerator, and thus improve the ROI of your traffic acquisition channels (SEO, AdWords, emailing …) via the increase of the conversion rate. Conversely, if it is not efficient, it may be a vanishing point for your shop and increase your bounce rate.
Audit the relevance of the results of the top 100 keywords
Before deeply modifying your internal search engine, it is important to analyze the existing, ensuring the relevance of the results. For that, it will be enough to extract the keywords the most plebiscites by your Internet users, in order to analyze them individually. If you use an internal search engine solution, you can easily find these queries in your back office. In case you use the native engine of your e-commerce solution, it is possible to go back to the requests of the users’ thanks to the tools of web analytics or via your back-office e-commerce.
The objective is to find out if the answers provided really correspond to the will of your company: Does the request show enough results? Are these results relevant? Are high- margin products or promotions better than others?
To perform an audit, here is a typical action plan:
- list the 100 most searched keywords (on an Excel file for example)
- individually test each request and give it an overall score. You must take into account the criteria that interest you: relevance, number of results, exclusion of products out of stock …
- count the number of satisfactory requests on the total.
If you have at least 20% of your top queries with a low score, that is, below average, you are currently losing sales. It is, therefore, necessary to thoroughly review the operation of your internal engine.
Define a keyword optimization process
The first step to optimize an internal e-commerce engine is to improve the relevance of the results. There are two main types of internal engines in e-merchants:
- the native search engine (Magento, Prestashop, WooCommerce, Shopify …)
- the use of an external search engine solution (open source, SaaS …)
In the case of a native engine, it will be necessary to modify the product sheets to add keywords, to highlight them on the queries. If this is not enough, it will also provide for the creation of additional fields, and modify the operation of the engine to take into account these new fields. This process can be tedious and does not guarantee a relevant sorting of results. In the case of an outsourced engine, all these features already exist, and there is nothing to develop: you can, however, optimize the display of results, keyword by keyword, focusing on top searches. This work may seem time-consuming but it will optimize to the nearest millimeter the most wanted terms. For your information, the algorithm developed by the solution will also allow you to act on the other keywords, which constitute the “long tail”. You can, for example, cross additional criteria (data from your ERP in particular) or create synonym lexical fields very quickly, to make the results more relevant automatically.
Implement merchandising: the gondola head e-commerce version
As a consumer, we are all faced daily with the rules of merchandising supermarkets: gondola heads, promotional offers, new products positioned at the entrance of the shelves, etc. This lever is a strong sales generator and can multiply sales by a factor ranging from 3 to 5 depending on the attractiveness of the offer.
Searchandising uses the same process. This is to display products, not only taking into account the entry of the user, but also using other criteria (margin rate, available stock, deal negotiated with the supplier, product rotation, etc. .)
Take the example of an e-commerce site specialized in computer hardware. For the “500 GB hard disk” request, the semantic query will only display products with this storage capacity. However, we may want to position at the head of the results a selection of 1to and 2to hard drives benefiting from a commercial rebate, to promote the sale of these references and suggest to consumers more advantageous offers. If the A brand 1to hard drive is a few euros at the same price as the B brand 500GB hard drive, the consumer will be tempted to take the A brand model.
It is therefore essential to establish a strategy of searchandising. How? After you have completed your keyword audit, you will need to define display rules based on business criteria. For this, there are two options: the development of a searchandising module, or the use of a search engine technology including searchandising.
The home-made solution can be expensive in terms of project management and development time. It will also have to be maintained by your teams, especially during updates on your e-commerce heart. Opting for an already developed internal engine technology, such as SaaS for example, offers the advantage of not having to worry about the technology (server loading time, development costs, updating of the solution, etc.). : everything is managed by your solution. In addition, this will allow you to customize the display of products by crossing many additional criteria, in order to increase the relevance, and assign an individual scoring to your products that will condition the display compared to other references.
Organize relevant and useful faceted navigation
The last point not to be overlooked is the faceted navigation. Simply put, these are the sorting criteria relevant to a product typology. Let’s take the example of our e-commerce shop specializing in computing, and the request “hard drive”. Faceted navigation will allow the user to specify his request, by selecting “facets” relevant to his need. This may be :
- storage capacity (250 GB, 500 GB, 1to, 2to, 4to, 8to …)
- the type of hard disk (SATA, SSD, IDE …)
- the size of the hard drive (2.5 inches, 3.5 inches …)
- the brand (Toshiba, Western Digital, Samsung, Maxtor …)
- the price (less than 50 euros, between 50 and 100 euros …)
These criteria are important because they make it easier to guide the user during his search. As they generally take the data sheets of the products, it will be necessary beforehand to have well defined and informed fields of information for your data sheets. The use of specific filters is also recommended to popularize a range of products when targeting a large audience that is not necessarily expert in computer hardware. For example, creating a “Type of use” criterion and “Portable PC”, “Desktop PC”, “Server” … filters will make it easier to choose the user according to his needs.
To conclude, it is essential not to consider your internal search engine as a simple query tool, without meticulous monitoring. Many merchant sites focus on the ergonomics of their homepage or product sheet, without thinking of the intermediate stage that represents the internal search engine. By balancing the relevance of the results, the strategy of searchandising and the intelligent use of faceted filters, you put the best of luck on your side to obtain the best return on investment, and especially to guarantee Internet users successful shopping experience.