Back to top

Ecommerce Search: Achieving Continual Improvement

With the growth of mobile search and the trend towards online orders for in-store pickup, search is becoming more important than ever in e-commerce. IT and business staff at e-commerce and retail companies are under increasing pressure to be agile in improving user experience and bottom line results from their online and mobile web site.

The problem, however, is that organizations often lack specific experience about how search works, how to diagnose what’s “broken”, and more importantly, how to fix the problems.

Search Technologies offers a staff augmentation solution to this problem: the Search Quality Analyst.

 

The Search Quality Analyst is a dedicated resource who can:

  • Communicate between business & search development to ensure business needs are met
  • Understand how to analyze query logs
  • Understand how to “fix” search by identifying necessary changes to:
    • Content processing rules
    • Query processing rules / query structures
    • Index & Query dictionaries, ontologies, synonyms, etc.
    • How exception cases are handled
    • Content text and metadata
  • Document improved results and compute ROI

 

The Search Quality Analyst service typically involves a full time consultant who will execute a continuous improvement process for search quality in the customer’s organization. Best results are achieved in multi-month engagements.

 

IMPROVING ECOMMERCE SEARCH

Whether you are talking specifically about e-commerce search or enterprise search in general, search is seeing a resurgence in importance, both to business analysts and IT stakeholders.

In e-commerce, this trend is being driven by a number of factors including:

  • The importance of online search even to large traditional retailers, who are implementing online purchase and in-store pickup programs
  • The increase use of mobile search users, whose limited screen real estate drives them to use search over other techniques such as browsing categories and thumbnails.

 

Old Problems with Search Relevancy and Quality 
For search teams, the problem has changed from “does the query work?” to “is the user satisfied?”

The problem with many e-commerce sites – especially for sites with large, complex product catalogs – is that users are often NOT very satisfied with the search experience. Across our customer base, Search Technologies notes common problems such as:

  • Irrelevant results – not showing results which the user wants and is most likely to purchase
  • False “zero” result summaries – i.e. an obvious product query that returns null
  • Improper synonym recognition – e.g. “monitor” vs. “display”
  • Bad query completion suggestions (terms suggested after users start typing)
  • Unclear results for out-of-scope queries – e.g. product FAQ / support vs. specification queries

These examples illustrate that relevancy is just the tip of the iceberg of the problems. There is a bigger problem with the overall quality of the search experience.

But it is one thing to recognize that search is “broken”. Figuring out how to “fix” it is another.

Typical e-commerce metrics such as “Abandonment Rate”, “Conversion Rate”, “Click Through Rates”, “Zero Queries”, etc. can identify that “search is bad”, but provide no useful guidance in how to fix it. It is the job of the Search Quality Analyst to bridge this gap.

Solving search quality problems involves a thorough analysis of query logs to understand and document problems as indicated by user behavior. This insight provides search Development teams with guidance on how to fix issues.

For organizations who lack the requisite business resources or expertise, a Search Quality Analyst from Search Technologies is an expert in search who is also capable of this type of analysis. Taking into account business needs and search technology architecture, the Analyst can also serve as a bridge between business and development and implement a continuous improvement process to help deliver better and better search experiences and business results.

 

READ ON: A Methodology for Improving Search Results Quality