If they can’t find it, they can’t buy it…One of the most crucial aspects of e-Commerce search is the way users perceive a lack of relevant search results. Is it because the site doesn’t carry the product, or is it because the site doesn’t support the query type?
• Reduce frequency of “no search results found”
• Improve usability
• Increase number of supported query types
I referenced Nielson Group and Baymard studies on eCommerce Search to get a better sense of supported Search Query types across the industry as whole and for interface best practices. I then looked at their Search usability scores through the lens of their global commerce ranking (see Top500 Database) to see if better usability might have a bearing in the success of their online business.
I did a deep dive audit of our own Search platform, pouring over customer videos, reviewing search metrics and submitting over 100 search queries to identify areas of confusion, opportunities for improvement and supported query types.
I presented the findings to our key stakeholders and got their support to make the initiative a priority and began working on competitive analysis of top performers in eCommerce Search usability benchmarks as well as some consumer favorites to get a better understanding of what made those experience great and what expectations customers brought from their personal life to their shopping experience.
Following the deep dive on data and competitive analysis I began speaking with Engineers and Product Owners about our technology stack, its restrictions and settling on a roadmap. However the roadmap we agreed was accelerated by leadership who to launch the improved search experience in parallel with a site re-architecture, new content index and article views. It was the first of many unforeseen bumps in our roadmap. We scaled back our initial experience to be a minimum viable product (MVP) and altered our plan to include iterative improvements post-launch.
Testing and Validation:
Partnering with our team’s Usability Analyst we set up a list of several Search oriented tasks for new and return customers to complete within the mobile and desktop InVision prototype(s) I had built.
The tests were largely positive, customers felt the new search experience was cleaner, more intuitive and less pushy about product up-sells, however there was little engagement with site-level scope changes in the left rail and few customers used the Search Scope adjacent to the search field. (typically a power user feature)
Given that customer’s expectations are colored by the experiences they have with devices and products in their daily lives, we felt it would be good idea to test Google’s tab-based approach for our site-level scope changes — it was one of the few successful approaches we’d seen from a vendor with both Content and Commerce. Surely putting them front-and-center above the search results would drive more engagement? Luckily it did, and with the marked improvements we’d made we could proceed to building out a high fidelity designs, developing and QA testing before the eventual launch.
Search launched with much fanfare and we saw:
• Improved usability
• Decent strides in our support of Exact Search, Thematic and Product-based Search
• A decrease in support of Natural Language.
• Better integration between Search and Filters
What I Learned:
Aside from become the resident eCommerce Search expert, I’ve walked away with a cautionary tale about boosting keywords, a better understanding of the limitations of Oracle’s ATG and Endeca and have a much greater appreciation for the magic that happens behind the scenes of Google and Amazon’s search fields.