Why Online Product Search needs to become more human

This article is brought to you by Retail Technology Review: Why Online Product Search needs to become more human.

By Twan Vollebregt, Chief Executive Officer and Founder at Traverz, a provider of e-commerce personalisation technology.

eCommerce retailers need to face up to an uncomfortable truth, which is that today's search and discovery mechanisms are based on technical and user experience concepts developed over 20 years ago. They leave consumers frustrated and unengaged, resulting in churn or missed sales opportunities.

Not true! Is the collective cry. So let’s analyse it and see why (and how) eCommerce retailers need to rethink their approach to how consumers search for products on their websites.

So much change - yet so little change

At the heart of product discovery is search, guided by facets and search-bars. These mechanisms reflect a technical search engine perspective - for good historical reasons. Once upon a time in the early days of the Internet, a software engineer was asked to fetch products from a database. The simplest way to achieve this was to ask the product designer to add some input parameters to the UI, which acted as database filtering parameters. Voila - the concept of ‘filtering’ was born - and it is still with us today, but not because it provides a great user experience.

Yes, both mechanisms have evolved over time. Facets are now being selected dynamically, and may be based on the specific user’s journey. Search bars have become more and more sophisticated, adding capabilities such as autocomplete, spellcheck, faceted search, synonyms, and boosting - as well as providing their results using a more appealing drop-down UI. 

However the fundamental nature of both mechanisms remains unchanged. They are static one-shot approaches to search that are not always easy to find or operate - especially on the small mobile screens that are dominating today. Neither mechanism involves any feedback capability or iterative element. 

This poor user experience (do you recognise the frustration?) means that generations of eCommerce consumers have too often resigned themselves to simply browsing the product list. This results in higher churn and missed conversions.

Today, just like 20 years ago, the human is expected to adjust to the needs of the machine. This leaves no room for iteration and it doesn’t enable the provision of feedback.

Personalisation through guessing

In recent years we have seen the advent of personalisation technologies based on data-driven machine-learning algorithms. These technologies are trying to bridge the gap for eCommerce retailers to predict what consumers want, and make up for the failure of facets and search bars to provide clear purchase intent understanding. Personalisation algorithms order the product list based on assumed intent, derived from harvested data or consumer segmentation. 

One of its key advantages is that it doesn’t have any impact on the UI or UX - it is an algorithm operating behind the scenes. But that is also its primary downside. Such ‘personalisation’ perpetuates the arms-length approach to consumer engagement, and the technology-driven user experience. The ‘machine’ is assumed to know best. 

Yet the ability to guess intent successfully is limited by the information available. An increasingly important issue is the large amounts of site-specific data (third/second/first party data) that are required to train machine learning algorithms. With tighter data protection legislation and an increasing reluctance from consumers to hand over their details, these machine learning algorithms are being starved of the required data, which in turn reduces their ability to achieve a significant amount of uplift.

The problem with a technology-led user experience

When technology leads the search and discovery experience, retailers and consumers alike lose the human touch, and this is to our collective detriment.

The focus in guiding the “consumer search journey” has been fully on just one of these three words - search. eCommerce providers are in danger of forgetting the elephant in the room - the consumer – who is on a multi-step journey. 

At each step the consumer is learning about the available products and developing a better idea of their preferences, and thereby forming or reshaping their purchase intent. This is a dynamic that has largely been forgotten in the rush to bring new technology into eCommerce.

Filters, search bars and personalisation do a poor job of allowing consumers to express their needs and wants. In turn, eCommerce retailers have insufficient understanding of consumer purchase intent to enable a positive and productive user experience. The net result is consumer frustration and missed sales opportunities.

Becoming truly customer centric requires engagement

It’s time eCommerce retailers embraced change like they did 20 years ago and rethink their approach to how consumers discover products on their websites. We need to shift the focus to understanding rather than guessing consumer intent, and provide a richer and more engaged experience that benefits both the consumer and the eCommerce website. 

This shift will allow us to escape the technology-led foundations of product discovery and become truly consumer-driven. The good news is that the route to shifting the product discovery paradigm is not as radical as you may think. 

In order to embrace consumer engagement and interactivity we must simply give the consumer mechanisms that encourage them to specify what they want as they progress through their journey. Then we must process those needs and wants in a manner that reflects fuzzy human thinking. So simple, yet so revolutionary!

A new discovery paradigm

How does this look in practice? Here are the core requirements. 

• Consumers must be given the ability to react and provide feedback, immediately, at any time - without needing to laboriously find a filter bar or start a whole new search. 
• Their reactions must build up a profile of their preferences, which encapsulate their purchase intent. 
• New search results must reflect those preferences, even when they are inconsistent or they conflict. 
• Rather than filter, we must flexibly search and present the best matches first. 
• We must explain whether the products shown are actually meeting all of the preferences - and when not.

Consumers think in terms of goals, not database queries. Their preferences develop and adjust through exposure to products, even (or maybe particularly) when those products do not yet meet those goals. Through a process of iterative engagement we gain highly valuable insight into their commercial intent - which allows accurate personalisation using zero-party (“volunteered intent”) data rather than trying to guess their intent from third, second or even first-party data.

The benefits for all parties are clear. Consumers gain a more productive, enjoyable search and discovery experience. Their increased trust and confidence leads to better conversion and greater loyalty for the eCommerce retailer. A win-win.

Does this mean total upheaval for the existing platform technology? No. All that is needed is an additional layer of consumer engagement. Much like adding a shop assistant into a brick-and-mortar shop, we can add engagement into the eCommerce experience.

A new approach to solving an old problem

Implementing new technology in eCommerce often involves long timescales, and as a result sticking to the “tried and trusted” method is often the chosen strategy. This has led to a focus on exploring smaller incremental adjustments through A/B testing. It is therefore not surprising that the underlying filter approach is still core to the search and discovery process today.

But in this crowded and highly competitive market, can eCommerce retailers afford to deliver a hit-and-miss experience? 

New innovations around product discovery enable the delivery of a richer and more tailored experience for consumers - leading to improved conversion. This paradigm shift is underway.

We need to remember that people are at the heart of all of this. Yes, we have the technology to deliver various algorithms to help with online shopping experiences, but without the interaction, engagement, and the human touch, we simply cannot begin to guess at what the consumer really wants. It is time to engage them, and know.