Unbxd Product Recommendation System

Machine Learning Versus Automation in Site Search for eCommerce

Machine learning, AI, Deep Learning are trending buzzwords now. Especially with Site Search and Online Commerce. But, how do you tell that you need machine learning? How do you know that a particular solution has machine learning?   So before learning all of that, let’s understand what machine learning site search is. How it differs from automation-based site search?

Machine Learning Site Search Vs. Automation Site Search

Though the terms automation and machine learning are used interchangeably, the most significant difference is that automation does not require any teaching. Automation is all about using machines to automate what humans would otherwise do manually. For instance, in E-commerce Site-Search, a good automation-based solution will use past purchase information to rank products with high sales numbers at the top. It might even use ‘clicks’ to automate showing highly viewed products to the top. It is advantageous because many search systems today require humans to input rules to boost the products’ rank to the topmost position. The flip side of automation is that it can go wrong with ranking within search results (like in the image above). If a ‘legging’ is selling really well, it’s great if it shows up when somebody searches for ‘pants.’ The system is geared towards ranking the most popular products to be shown the most. What about 80%+ visitors who want to see pants, not just the most popular legging (which technically is also a pant)? It ends up catering to the top 20% of best-selling products to 20% of visitors shopping online. So what is Machine Learning? 

What is Machine Learning?

If you need a system that not only automates human tasks but also sifts through a lot of data to identify patterns of predictability in data sets, you need Machine Learning. Machine Learning systems require intensive computations (a lot of CPUs) and need training data sets — a lot of data. They will use that to predict, for instance, the likelihood of a user to click on a particular type of product after a specific search. Another use case would be to predict what other queries are similar to one particular query with different search results.  The limiting factor for machine learning is the data — the more data it ingests, the more likely outcomes will be realistic. Machine learning models cannot predict outside of the tabulated data that is fed in. It is an important thing — If somebody is offering machine learning and is not ingesting a HUGE amount of data, they are just automating; there is no machine learning in play. In a typical use case of well trained and excellent performing machine learning system, it ingests data source like:
  1. The source of traffic
  2.  Search queries
  3. Clicks on filters
  4. Clicks rank of products
  5. Product clicks
  6. Time spent on PDP pages
  7. Add-to-cart events
  8. Purchases
  9. Exits
  10. Repeat visits
  11. Non-search clickstream events
  12. Web bases social trending data

Machine Learning Site Search Versus Automation Site Search for eCommerce

While relevancy plays its part with machine learning, it can determine that there is sufficient deviation in behavior when people search for ‘pants’ v/s ‘leggings’ independently. It’s not programmed or automated, as is evident in the image above.

How is machine learning site search leveraged for eCommerce?

At Unbxd, when we compared a machine learning site search system (CTRs and conversion rates) on an A/B test with the best-automated systems, they both behaved similarly on the top 5% of queries. However, the Machine Learning site search outperformed automated search by 2x (100% difference in conversions) with the rest 95%. It is because sales data heavily drive the top 5% of search queries. The remaining 95% of questions need more predictability — not just automation. One thing to note, though, is to be realistic about what machine learning can deliver. You need significant data to make training models work. If your site has less traffic — it might take a while before it starts kicking in. To summarize, you can use this simple checklist to verify if there a site search has machine learning or not:
  1. Does site search ingest at-least five different types of signals?
  2. What machine learning models does it use?
  3. What are some patterns that these models have predicted?
  4. Is there a dedicated data sciences team?
  5. What’s the technical architecture for machine learning in play?
This article is the first in a series of super cool posts by Pavan, which he pens down when he’s in a mood to share. You can read it on Medium, too. 

More To Explore