Chatting with bots: online conversations are good for business in the real world

Chatbots are so hot right now

One of the hottest topics in the world of retail at the moment is conversational commerce, or “utilizing chat, messaging, or other natural language interfaces (i.e. voice) to interact with people, brands, or services and bots that heretofore have had no real place in the bidirectional, asynchronous messaging context.” according to Chris Messina.

Translation: more ways for customers to talk to brands and buy stuff.

Conversational commerce makes sense for the always-connected crowd. How many chat or messaging apps do you have on your PC or phone right now? If you want to connect with friends or colleagues chances are you’ll hit them up over Messenger, Whatsapp, Kik, Slack or Skype. In fact, according to ComScore, your preferred messaging app is probably one of the three most used apps on your phone. For most of us engaging with brands via chat is evolutionary rather than revolutionary.


Rise of the machines

Broadly speaking, conversational commerce runs the gamut from the more traditional human-led (store associates, personal shoppers and concierge) experiences to purely computer-driven, and the latter is really starting to pick up speed. Chatbots are a conversational AI, capable of interfacing with both humans and other technology – in retail applications usually drawing data from connected systems including PoS, CRM and inventory. They’re also a potential answer to the not-quite-age-old question – does our brand need a mobile app?

A lot has been said about consumers’ fickle relationships with retail brand apps: some get used regularly, most don’t, and it’s doubtful even Starbucks could claim to have one of the top 3 apps on any given customer’s phone. If instead of forcing people into an app, we allow customers to connect with brands without leaving whichever messaging app they’re already using, it’s not a stretch to imagine they may do so more often. This also allows brands without the budget or means to develop a custom app to engage in connected commerce simply by not-coding a bot or hooking up an existing AI – of which there are many: Pandorabots,, Conversable, Shopify,, and all of these guys.

So how might a conversational transaction with a chatbot play out? Obviously you’d be aiming for a human-ish conversation; if you didn’t need that level of back and forth you could get away with a way less clever technology.

Obviously that’s a fairly simplistic example, but have a think about what needs to happen in the background to make this experience natural and fluid:


Be where your customers are

There’s not much point being somewhere your customers aren’t. (Marketing 101). Luckily for marketers, chatbots as a tech are fairly equal opportunity and can be deployed on most popular social platforms. Recently Pizza Hut announced it will start taking orders via Twitter and FB Messenger, while TacoBell has opted for Slack delivery for its awesomely named not-quite-launched TacoBot. In China the chat app of choice is WeChat, where conversational commerce is likely to rely less on bots and more on human assistance (check this extensive commentary out).  Wherever and however your demographic is chatting, that’s where you need to be.


Introduce efficiencies without losing the human touch

Bots should be able to respond immediately; to engage the customer while doing the behind-the-scenes heavy lifting. This isn’t always possible with human brand ambassadors – they may have several requests to deal with at once, might be out to lunch or in a meeting or generally otherwise occupied when a query comes through. While it still makes sense to route sensitive, complex or high value requests to a human assistant, the ability of bots to near- instantaneously assist hundreds of shoppers introduces some serious efficiencies. Especially if some queries are just that; unlikely to ultimately result in a sale but still a good opportunity for a positive brand experience.

Personalization is expected

Bots need to respond differently to repeat customers vs people you have no data on (many of whom may become repeat customers if you play your cards right…)  So the bot will need to check the CRM for customer data: personal information, previous purchases, sizing information, favourite pizza toppings, preferred store, loyalty program enrolment and status. In the same way we’re taught to pre-populate emails and forms with already known details, bots should be able to respond with a customer’s basic information without having to repeatedly ask for it.


Connect stores for that omnichannel experience

Finally, your stores need to be connected so your bot can return accurate information on inventory levels at each location, staff on duty, current promotions, opening hours. You also need to be able to push back from the bot to stores in order to reserve stock and notify staff once a customer’s placed an order.

Because our focus is on retailers with a physical presence, this ability to direct customers from a digital channel into a store is of particular interest. We know that almost 2/3 of customers who opt to click and collect purchase additional items once they come to store, and we know every visit gives us additional data points that we can potentially use to personalize future interactions. By using technology to support and enhance the in-store experience we get the best of both worlds; customers who get what they want via a frictionless experience, and stores getting visitors through the doors.

Check out the video to see how Plexure makes chatbot magic a reality for retail marketers: