First let’s will look at why it is important to understand your customers intentions when they search and interact with your Ecommerce site.
Many studies show that shoppers using on-site search are, on average, two* times more likely to convert than shoppers that do not search. So, when someone searches on your site it is important to understand what they want and get them to it quickly.
Secondly, mainstream consumer technology and voice assistants such as Alexa and search engines like Google are creating new “Human Computer Interaction” paradigms and expectations from consumers. This will become table stakes to compete in the world of ecommerce creating a knock-on effect for traditional ecommerce search interactions too.
Defining the best digital Journey
Helping you navigate around this and any new expectations can be simplified by breaking it down into bite size chunks. The first chunk to look at is how you can understand your customers intentions on your site, once you do that it can help you curate an amazing and relevant digital experience. This does raise the question, “how do we know what sort of experience they need”? This can be easier than you think. Human behavior in the retail context can be very predictable and mapped out in journeys. So of you may have been in one of these fun workshops:
Once you have mapped out your customers journey the next step is a nice approach that I stole from a retail guru and mentor which has turned out to be very useful. 1) Take the context of the consumer be it: online, instore, mobile, with pushchair & screaming baby, and so on. 2) Take the product or service category: is it essential, discretionary, influencer led & lifestyle signaling, and so on. Armed with these two key points on context and category you can create a wonderfully relevant customer journey. This is what many successful retailers do, they give you what you want, at the right time, in the right way. So much has been written about exceptional customer journeys based on categories and services. It does not take much reading or research to create something special.
Now we have the journey nailed, here is where in the ecommerce world its gets hard. Once you have created that amazing journey, like the one above, when a customer interacts with your digital ecommerce presence, how do you know the point they are at in their journey?
Understanding your Customer Intentions “old school”
It was much easier in a brick and mortar shop, human to human interaction made it easy to understand customer intent. This was arguably turned into a science with data from observed behavior when Paco Underhill wrote, “Why we Buy”. This masterpiece was first published in 1999, Paco provided key lessons for anyone willing to read his book. He showed how consumers in physical stores were likely to behave in differing category and service scenarios. Given these from findings from Paco Underhill I would argue that customers are in fact highly predictable. Amazing to think that before his work most retailers used gut feel and were largely right. Paco Underhill put some credence and data around what retailers did out of “gut instinct”. Adding to this my own observations, based on growing up in retail helped me understand that “gut feel” too. I worked in retail from an early age in a family run bakery and shop, I loved serving customers, it was fascinating. The smell of freshly baked bread and cakes placed openly, not behind glass, but on top of display units are just a couple of the simple merchandising tricks used to entice customers. However, if they had feedback or wanted to make up a regular order or spend 30 minutes reviewing a very important wedding cake design, it was pretty easy using human to human interaction to work out their intentions and service them appropriately with a friendly smile.
Understanding Customer Intentions in Digital
In the world of digital we need new tools to help us understand customer intent.
To help us understand the challenge in any approach, let’s take to some examples. Imagine you have an online DIY store. Let’s say you have a scenario with two different customers: Amit is on his mobile device and searches for “bayonet bulb” on your ecommerce site and Sarah searches for “Spring bulbs and planting ideas for my garden”. The first is at the end of a purchase journey the second is at the start of the journey. The first needs a product and checkout, the second needs content rich editorial inspiration with glossy pics like this:
Most ecommerce sites with traditional token-based matching could manually create redirects to help with “bayonet bulbs”, some retailers don’t do this and end up providing “flower bulbs” as a search result – try it out! Most sites with effort can cater for this product search. We can use search tuning to help direct to the relevant product.
In the second example, bringing back products now we would make the journey less relevant and arguably bad, possibly losing the customer. This happens due to many ecommerce search engines tokenizing Sarah’s search phrase and creating a lot of irrelevant search noise. You are most likely to get “plant pots” returned, again try it out at any DIY site! You and I can both see that Sarah needs content rich inspiration and she is at the beginning of her purchase journey.
Digital Tools to enable relevant Customer Journeys
To better server these two customers in your ecommerce store you need two key capabilities in the your Ecommerce store, they are:
- Search, as a digital merchant your ecommerce search engine should recognise context better, to do this you will need a few tools in the kit bag. Natural language processing (NLP) capabilities are one of the “killer apps” to help here. What does this mean? Well traditional search engines as we can see in our examples would match words as tokens, with no context. This creates a lot of search noise and too many results. NLP understands parts of speech and helps reduce search noise significantly. It does this by labelling each token in the customers search term with part of speech identifiers such as noun, adjective, and so on. NLP can also help your store understand named entities common ones such as towns and cities or any entity in your Catalog. In simple terms this more accurately identifies the subject of your customers search phrase. Try it yourself here https://corenlp.run/ put some phrases in to see how Part-of-Speech and other NLP features annotate search phrases. NLP helps make search results more relevant.
- Content management capabilities integrated into search. This is the next “killer app” to improve yet further and differentiate your ecommerce eourney from a “catalog retailer” like experience. Fair enough if you another strategic advantage that’s a fair argument, it might be, supply chain or unique product/service, but on the whole to survive in digital most merchants will need the ability to let loose business and creatives. You will want them to create editorial content relevant to your customers seeking guidance and inspiration. For that you will ideally have on hand a Web Content Management System (CMS) capability to manage your content. Helping out with ever changing consumer tech, multiple languages, multiple view ports/devices and so on. This should then allow you to create immersive and rich content led experiences that flow seamlessly into purchases. This content should also appear in on site search too. Historically a Content system was and for many is still a siloed system not integrated into Ecommerce. This should not be the case; we need Sarah’s search term to return a relevant blog or editorial piece. And here is where the NLP indexing of your content helps too, because all your content has been tokenized and tagged with Part of Speech and any named entities tagged too, that same tokenizing is applied to your customers search phrase means that relevant results that are content rich results are more likely to be returned for Sarah.
HCL Commerce has both these capabilities, check out this video to see it in action:
In conclusion, I like to think of digital retail like I think of any relationship, it is very important to understand what the intentions of the other party are…
*Authors own observations based on data from several Ecommerce retailers. 2 X is also a conservative interpretation of this data.