PartzInsight is an AI Platform for your Part Catalog providing Amazon’s features, at Prime Subscription pricing
The pandemic has transformed the parts catalog business from the counter to e-commerce.
No longer are consumers and mechanics heading to the nearest outlet store to hang around the counter to shop for parts. Within a few months, the industry instantly transformed as a result of the pandemic and shutdowns ordered. As a result, Online sales of auto parts has already grown by 30% this year. Rapid growth in reaction to customer demand for online is causing a tremendous disruption in the marketplace. Now the Customer Experience (CX) is primarily electronic vs in person. Once you focus more narrowly on eCommerce transactions, you rely on digital systems (website, mobile app, etc.) to provide your customers with a similar personalized experience to what that they received when visiting the counter, minus the wait time & calendars.
Our increased reliance on digital is accelerating transformation. Transforming your CX to a digital-first experience presents new challenges; beyond launching a store, populating catalog, and putting some American Flags on the website. Aftermarket e-commerce catalog search has some unique challenges. Will the part fit a customer’s vehicle (compatibility and fitment)? Are customers using terms to find parts that match your descriptions of them in your store (linguistics and synonyms)? These are challenges handled by Amazon and other large e-commerce sites that follow best practices and enjoy the increased revenue growth that these methods carry. While technically challenging and expensive to implement from scratch, your customers do not care. They are typically active buyers looking to find and buy a specific part for a particular vehicle. The journey to purchasing is direct: search box to navigation to the product detail page to checkout. Removing friction from this new last-mile could mean the difference between growth or irrelevance.
SKU search is not every search in Parts Catalog E-Commerce
Hopefully, your current system returns the correct item when the user enters a specific Stock Keeping Unit (SKU) number. Effective SKU search is a good start, but what percentage of all searches does that represent a minority? Commonly, searches are only off by a couple of characters or terms, which is less specific than SKU “TRX400EX rear shocks,” for example. If your catalog had the part broken out in the following manner, “Honda TRX 400EX TRX 400X TRX 450R TRX 450ER,” there isn’t a match; therefore, you don’t sell the product. Another lost sale. How much better would the experience be for customers if the system suggested the category of “cables” and make of “Honda?”
To keep up with your users, top websites use Machine Learning to aid in the relevancy. Thoroughly detailed analysis of your query history via “Head to Tail” analysis. The computer can use machine learning to match low frequency searches to higher frequency searches to aid that user find what they are looking for. Natural Language processing can detect Year, Make, and Models from the query box as well in addition to other naturally occurring phrases (i.e. “Clutch Cable”).
Don’t lose customers by not providing personalization.
When you graph the distribution of search queries for most websites, you end up with a Zipf curve, also know ask ‘the hockey stick.’ There is an initial dense concentration of search queries that are generally predictable to a subject matter expert. These terms can be manually tuned, but as you look out further along the curve, it becomes not quite as clear how these should be performing and exactly what action to take to help with conversation.
Personalization is the Key for the Aftermarket.
E-commerce conversion is king. Better search can produce increases in conversion rates by 10%. Providing personalization can achieve an ROI of 100% in weeks, not months. Personalization improves CX, enhancing customer retention and affinity with your brand, leading to repeat traffic, and enhanced brand sentiment. Failure to own your CX will leave you bidding on ads against megacorp Google or hoping for Amazon/eBay shoppers’ orders which carry added commissions and lower margins. In all these scenarios your potential customers are comparing you directly to the general marketplace, and your competition.
Next-generation parts catalog e-commerce systems understand the customers’ intent rather than explicitly relying on the submitted search query. This is the foundation of providing intelligent experiences. Using this intent, you guide the user to the product that best suits their needs.
Parts Catalog recommendations boost basket size, and margin.
Traditional search systems provide results based on the term frequency-inverse document frequency model (IF-IDF). What if that term is not in the document? What if the form of the SKU query is different? What if the manufacture changed the name of the product? What if for branding or legal reasons, the aftermarket part does not mention the brand name. The ultimate judge of relevancy is in the purchase, and no restocking costs… Can you use that to influence future visitors?
That’s a lot of what if, but if people who issue a particular search buy a specific part, does the term frequency matter? What do people buy after they view this page? Can you use that to influence other purchasers, such as fueling your look-alike lead generation?
Amazon type of technology for an Amazon type price.
Faster, cheaper, and better. We saw the challenges that companies faced implementing modern technology. Most mid-market eCommerce, and even Fortune5000 corporations, do not have the staff dedicated to select, tune, and operate systems that are based on Machine Learning and Artificial Intelligence.
We took our experience working with large aftermarket part manufacturers, component manufactures, and small ‘mom and pop’ resellers to assemble a ready to run solution called “PartzInsight”. With PartzInsight you subscribe to a “consumption” based model. By a pay-as-you-go model (PAYG), your company does not need to invest hundreds of thousands of dollars to get started with a recommendation engine. A well-tuned recommendation engine can be your company’s single greatest margin-driver.
We’ve embraced the Software as a Service model (SaaS). There’s nothing to buy up from. Once you’ve been enrolled, simply upload your cart via our API and begin searching. If you go further and tag your pages to provide analytic feedback, then you can begin to benefit from Machine Learning!