In our previous article, we presented the results linked to deploying AI in e-commerce customer service. These results depend directly on first identifying the right friction points. Here is how to identify yours.
These 5 frictions cost you sales every day. Yet none of them appears in your standard analytics.
An e-commerce dashboard that is often limited
First, when sales stagnate, the natural reflex is to optimize what is visible. First the visuals, then the promotions, the price or even the page copy. These actions are certainly useful, but they do not solve the invisible frictions. These arise between the purchase intent and the click on “order”, in the space of a few seconds of silent hesitation.
Before deploying an agentic AI project, it is worth considering some support.
Indeed, an audit of your friction points will not seek to improve what already works. It will serve to identify precisely what is blocking. The point is to pinpoint the key moment and which type of product the friction occurs on. To help guide your thinking, here are the five frictions encountered with the e-commerce players we support.
1. The technical question with no immediate answer
In the first place, on products with technical specifications (electronics, sports or tools, for example), between 25 and 40% of abandonments are linked to a question about compatibility or a specific feature. Often, the product page does not answer visitors’ questions (Forrester 2024 data).
Unfortunately, the information exists in your knowledge base but it is invisible. It is simply not accessible at the right place or the right moment.
How do you measure it? In general we recommend analyzing the internal search keywords on your high-traffic pages. You can also look at the queries on the product pages themselves. They reveal exactly what visitors cannot find.
This is precisely the first point that Genii solves thanks to its proactivity and its dynamic formats that integrate perfectly into the acquisition funnel.
2. Uncertainty about delivery at the moment of decision
Indeed, the question of delivery is not one you ask after the purchase. On the contrary, it is one of the last questions you consider before ordering. A visitor who does not know whether their parcel will arrive before an important date does not order, even if the product matches their need exactly and the price suits them.
In practice, sites that display personalized real-time delivery times convert 12 to 18% better on gift products and urgent purchases (Federation of e-commerce and distance selling, FEVAD, 2024). The difference between uncertainty and transparency is sometimes worth more than a commercial discount.
Thanks to a virtual assistant, the e-merchant offers a channel for instantly answering this question that is crucial to the purchase decision.
Did you know? With the Genii solution, you can use various widget formats to create a discreet, innovative presence that is useful to your visitors.
3. The return policy that cannot be found
In third place, returns are one of the top five purchase decision criteria for 73% of French shoppers according to a FEVAD study. Yet most sites relegate it to a page accessible only from the footer or via the terms and conditions.
As a result, a visitor who cannot find it in one click from the product page will interpret this absence as a negative signal. You lose their trust before they have even clicked on “add to cart”.
4. Abandonment during the checkout funnel
68% of abandoned carts are abandoned during the checkout funnel. Half of these abandonments occur over a simple question left unanswered: the visitor has a doubt about their order and has no way to get an answer without leaving the funnel.
In response to this, an AI agent available during checkout is able to answer the visitor. A simple answer to “can I change my order after confirming it?” reduces this abandonment rate measurably!
5. Recommendations disconnected from context
Finally, your search engines display “similar products” or “customers also bought”. These generic recommendations have a low click-through rate because they do not take the potential buyer’s context into account.
Indeed, an AI agent that knows the conversation it has just had with the visitor can make a genuinely contextual recommendation. This level of personalization increases the average cart by 15 to 25% in the categories concerned.
Improve your customer experience with AI
tolk.ai offers its clients free trials and personalized support to stay as close as possible to each company’s business goals.
We help you analyze your analytics, your abandonment journeys and your support question flows to identify the best use case. The first results are measurable within the 30 days following go-live.
To schedule your audit, contact our team. Try the Genii solution for free for 14 days.
