Initially conversational commerce revolved mainly around chatbots, but with the proliferation of smart speakers the ‘conversation’ part in conversational commerce is becoming more natural, so these days conversational commerce is (according to its Wikipedia definition): “commerce-commerce via various means of conversation”, with live chat, messaging, chatbots and virtual assistants as primary examples.

Conversational Commerce is taking such a big leap that the web is full of ‘increase revenue with conversational commerce’ type articles. This leads to companies rushing into the field trying to create a virtual salesperson to sell whatever it is that the company normally sells. Be aware however that people generally don’t like to talk to salespeople. Especially with virtual assistants it’s quite easy to create an obnoxious virtual sales assistant because, with the current state of voice interpretation, virtual assistants won’t understand many nuances in a conversation, and where a real salesperson might know how to tone down a conversation if the customer gets annoyed, the virtual sales assistant might just push on, chasing the customer away.

Our approach is to start with a Conversational Service and let the Commerce part come in naturally. We do this from the perspective that when using a voice assistant, the user probably already knows what they want. When for example people are talking to the assistant of a major cinema chain that we recently released a voice app for, we start from the assumption that the user’s intent is probably to buy a ticket, or at least to get information that may later influence his or her decision to buy. This means that we don’t have to actually sell something to the user, we can probably just guide the user along the process.

In cinema terms, we try to make sure the virtual assistant does not mimic the salesperson that sells you a ticket, but rather the usher that lets you in, guides you to the ticketing area, or points you to your seat. We’ve noticed that it really helps to do it that way so that the sales process comes natural to the users and they don’t feel like we’ve been trying to sell them something.

It’s not a coincidence that it’s called an ‘assistant’ – that’s really how we need to think about it. When we develop an assistant app, we first look at ‘conversational assistance’ and see how we can best assist the user in reaching their goals. When we do that successfully and the user truly sees a benefit in the assistant and feels comfortable using it, we add the commerce capabilities, in a way that fits in naturally.

We’ve launched the cinema assistant that I mentioned above a couple of months ago, and we added commerce capabilities to it just this week, so it’s a bit too early to back up the above approach with actual evidence, but we’ll monitor the performance of the app and will later write another blog post if we come to different insights