As we approach the later stages of the year, the pressure mounts on retailers to ensure they’re maximising their revenue with people taking advantage of Black Friday sales and getting their Christmas shopping done. COVID-19 has seen businesses put more emphasis on e-commerce than ever before, with Accent Group’s chief executive, Daniel Agostinelli, predicting 30 per cent of retail sales will be online. And while many businesses have a vastly improved e-commerce offering when compared to the start of the year, there’s still room for improvement.
A key part of ensuring a business’ e-commerce is effective, is its customers’ ability to find what they’re looking for and getting the most out of its search.
Identify the biggest problems
There are two types of bad search results: false positives and false negatives. A false positive is when a search returns “relevant” results that are actually not relevant. This is the most common problem most businesses focus on in their search, and wrongly so.
A false negative is when a search omits relevant results, meaning the customer can’t see what they actually want, and the business loses a sale. Instead of false positives, false negatives are where e-commerce leaders should focus their efforts. While it’s not ideal including irrelevant products into search results (i.e. false positives), the business is ultimately leaving revenue to the side by excluding relevant results (i.e. false negative). Including false negatives must be a priority that e-commerce leaders address immediately.
Focus on the right areas, in the right order
Various components are important to getting e-commerce right, but it’s essential to prioritise these components to deliver a strong e-commerce offering. Failure to do so will leave a clunky and disconnected process, which will see customers become frustrated and exiting the website with a full cart. Here’s how retailers should prioritise addressing their search challenges.
- Fix searches that incorrectly yield nothing
A search should never return zero results. If the store doesn’t have what a consumer is looking for, it should provide suggestions that are connected or similar to the query. This tells the consumer that while the store doesn’t have what they’re looking for, they might still find other similar products interesting or useful.
2. Include false negatives
As mentioned earlier, false negatives don’t provide consumers with what they’re looking for, despite the product being available. Be sure to include false negatives into the search results, even if that means including false positives as well.
3. Rank relevant products first
If a consumer searches for a black t-shirt, put the black t-shirt first. The most relevant results should always be prioritised at the top—not in alphabetical order or any other order. By entering a search query, the customer is asking a direct question…give them a direct answer.
4. Fix user interface (UI) / user experience (UX) issues
Many retail leaders focus their efforts here in the first instance. While this does have a commercial impact, it’s not the most important aspect to be focusing on. There’s no point in having a nice UI/UX just for the sake of it—make sure it’s backed up by a search that accurately directs customers to the right products.
5. Get more precise by focusing on false positives
Once a number of other issues have been fixed and consumers are able to see all relevant results, now the business can focus on delivering more precise results. This means removing the false positives, which act as clutter as consumers try to find the products that are relevant to them.
6. Strategically position products
With an appropriately optimised search that only delivers relevant results (i.e. true positives and true negatives), the business should strategically position products on the page. Products the business prefers to sell (this could be because they deliver a higher ROI or perhaps are perishable) should be positioned at the top of the page and the last slot as these are the positions consumers take note of the most.
7. Move from lexical to semantic
This is what retailers should aim to achieve as they continue to develop their search. Semantic search results will see businesses suggest products related to the search query, yet don’t share any keywords. For example, if a user searches “The Meg” on a retailer’s site that sells movies, naturally the movie The Meg will appear (assuming it is stocked). However, semantic search will see other similar shows and movies be recommended such as Jaws and/or Open Water. This allows businesses to take their personalisation efforts to the next level as they begin to understand more about user intent.
The time of year we’re now approaching will be like no other for retailers, with bricks-and-mortar stores experiencing record lows, while online stores experience record highs. The environment in which retailers are forced to operate in has now changed—they’re now competing online, rather than next door. It’s important they change and deliver a modern and optimised e-commerce experience, otherwise risk experiencing a grim Black Friday and unhappy Christmas.
Peter Curran is general manager for digital commerce at Lucidworks.