Big Data for Better E-Commerce Search and Conversion Rates
In the Trenches with Big Data & Search – A Blog and Video Series
You've heard a lot about e-commerce personalization for a better online shopping experience and greater bottom line. But how can big data make advanced, real-time personalization possible?
Watch the case story below to find out
What Causes High Abandonment and Low Conversion/Retention?
According to a recent e-commerce search experience report by the Baymard Institute, 34% of searches on the top 50 e-commerce sites don’t produce useful results, and 70% of those search engines can’t produce relevant results for product synonyms. Statistics may vary from one research to another, year after year, but the reality is simple: If your online shoppers can’t find what they need, they can leave your website, abandon their carts, or worse, turn to your competitors for future purchases.
Some shoppers may search on the fly on their smartphones, some may browse on their computers for hours before typing specific queries in your site search box. But in any cases, good e-commerce search will eliminate these dreadful scenarios:
- Irrelevant results – search results are not what the user is looking for or most likely to purchase
- Improper synonym recognition (example, “monitor” vs. “display”) – you have the product in your inventory, but it won't show up on searches because the user types in a synonym rather than an exact match
- Bad query completion suggestions – the auto-suggested search terms are irrelevant to what the user is looking for
- Unclear results for out-of-scope queries – product FAQ / support vs. specification queries
- And the worst offender? A single line stating “Your search returns 0 results,” without recommendations for related items or a way to browse for similar products.
Using these scenarios to evaluate your e-commerce search performance, you can identify the causes of failed queries and lay out the criteria for improving your user experience, conversion rate, and Net Promoter Score (NPS), starting with...
... Knowing Your Online Shoppers
Think of the volumes of activities and transactions performed on your e-commerce site. There are enormous opportunities to leverage big data in the e-commerce space in order to address the scenarios described above. Knowing what your shoppers search for, how they browse your catalogs, what products they purchase, how much revenue a product generates, etc. is great (find out how to better understand your shoppers with big data log analytics).
But how can you use that valuable data for higher e-commerce conversion and customer loyalty? Big data and machine learning have provided an automated yet powerful and accurate way to understand, attract, nurture, and guide your online users through their entire shopping journey.
Machine Learning and Predictive Analytics for A Multi-Device E-Commerce Search Strategy
On desktops and tablets, product recommendations can be displayed expansively and there’s more leeway on how you want to display them to shoppers.
On mobile, however, every pixel is critical. So to appeal to your shoppers and improve your conversion rates, put your best foot (in e-commerce case, top margin products) forward.
Here’s how a big data architecture can optimize multi-device e-commerce search:
- Cleanup and enrichment of product metadata for showing on different screen sizes, including product descriptions, image tags, categories, ID numbers, etc.
- Intelligent automated suggestions and/or synonym recognition for user’s specific search queries
- Queries chaining for contextual, long-tail search queries
- Individual product suggestions derived from preferences of users with similar characteristics
Through automated machine learning and predictive capabilities, a big data e-commerce platform can show products to an individual shopper based on the probability or likeliness that the shopper would purchase. It can even multiply that probability with your revenue estimate, then sort your product display order by revenue size (remember putting your highest margin products forward?). Just like how online marketers are obsessed with associating website activities with dollar amounts, your e-commerce search strategists would want to tell how much potential revenue can be generated by big data's personalization for e-commerce.
See how machine learning and predictive analytics work to produce a more intelligent e-commerce search box.
Real-Time Personalization – the Future of E-commerce with Big Data
The next-generation big data architecture is revolutionizing e-commerce personalization with its ability to produce real-time user profiling by analyzing real-time rather than historical log data. Eventually, personalization will rely on very little data entry (users typing in a search box) but more on machine learning and relevancy ranking algorithms, as we discussed in Big Data, Personalization, and the No Search of Tomorrow.
At Search Technologies, we have been working on a sophisticated big data architecture using Apache Spark to provide real-time user profiling with clicks, products added to cart, search queries, etc. With remarkable data processing speed, Apache Spark can input massive user log data into a real-time personalization engine that will customize search results, catalogs, and product recommendations on e-commerce sites.
Imagine being able to serve the exact products that your online shoppers are likely to purchase, in real-time. With open source big data, real-time e-commerce personalization is no longer a distant future - it's here now and will be a competitive advantage against other legacy systems.
E-commerce search is a use case in our “In the Trenches with Search and Big Data” video-blog series – a deep dive into six prevalent applications of big data for modern business. Check out our complete list of six successful big data use cases and stay tuned for more video stories of organizations that found success from these use cases.
Sign up for our newsletter to get the latest updates on "In the Trenches with Search and Big Data" video-blog series.