There’s a lot of confusion and misinformation swirling around the ever-encroaching wave of business automation. Depending on whom you ask, you might come away believing that it’s all overhyped smoke and mirrors, or that we’re on the precipice of every job aside from AI handler being made obsolete. The truth, unsurprisingly, is in the middle.
Automation keeps getting better, but it has plenty of limitations that aren’t likely to be overcome in the near future. The ecommerce chatbot exemplifies this perfectly: it’s a powerful tool when used well, but a disappointment when deployed with unrealistic expectations.
If you’re an ecommerce merchant, then, and you’re considering implementing a chatbot to support your store, you need to understand what it can do to help you, and what it can’t do. Let’s run through what you need to know.
Ecommerce visits and sales don’t fit neatly into standard opening hours. Given the relative ease of international shopping today, and the temptation of impulse buying from a smartphone in the early hours of the morning, you can pick up sales at any time of the day. Consequently, it isn’t enough to have people ready to provide support between 9 and 5.
What do you do when a prospective buyer has a vital question at 3am? You could respond to their email later that day, but by that point their interest might have waned. You could hire someone to be ready and willing to help during night hours, but that adds significant expense to your operation and eats into your profits.
Consider that we live in a time of lean startups for a reason. The online sales industry is hotly competitive, and the more money a merchant can save while getting their business started, the stronger their chance of being successful will be. Disaster can strike at any time, remember — you must find the right balance between investing in the future and saving for it.
Chatbots allow you to maintain a basic support operation at any time of the day, and even in any language (provided you install the right language libraries), at very low cost. If someone wants to know some more details about a product, or chase an update on a pending order, they can get the information they need without requiring manual human involvement.
Replicating the traditional sales process is complicated in a digital setting. When a sales assistant is helping someone in a brick-and-mortar store, they have more time to speak to them: they can gauge body language, ask nuanced questions, and walk them to particular aisles. Online, there are so many more products to be aware of it, and it’s so much harder to know when it’s appropriate or useful to bring something up.
Chatbots, however, can achieve comparable effects through machine learning alone. When a shopper interacts with a chatbot, it can rapidly draw from their purchase history, take myriad other factors into account, and display the cross-sells and upsells that are considered most likely to drive sales.
Could a human assistant with a comprehensive knowledge of the product range and some creative freedom do better? Probably — but the cost would be so much greater, as well as the risks. If a person provides bad recommendations, the shopper can feel justified in complaining. If a chatbot provides bad recommendations, it’s harder to gripe — after all, we all know that algorithms aren’t flawless, so why expect otherwise?
The protection and use of personal data has become an exceedingly-thorny issue in recent years. In particular, the implementation of GDPR brought a lot of mainstream awareness to how ineptly many businesses store data, and how unethically they exploit it for profit. When a shopper reaches out for support and starts being asked to provide personal details, it’s understandable that they be reluctant.
One of the most useful parts of using chatbots is that they can take human interaction out of the equation when it comes to handling data. When you talk to a person, even the most secure chat can’t be that secure because the person you’re talking to will see the information — but when you provide ID details for a chatbot, it can be suitably parsed within the system without anyone other than you getting to see it (what’s more, a chatbot is much more reliable when it comes to confirming necessary data, which is vital for avoiding identity fraud).
This has two benefits for you: firstly, it lowers the threat of private data being leaked or misused by an unprofessional worker, and secondly, it can make your customers feel more comfortable about sharing their information (you will need to inform them that your chatbot is safer, though, because they’re unlikely to assume it!).
Should you dismiss the bulk of your staff and commit fully to your new chatbot army? No, of course not. The advantages we’ve looked at are certainly valuable, and you have every reason to add chatbots to your strategy — it’s simply ill-advised to conclude that human assistants are lesser somehow, because that’s far from the case.
What human assistants lack in scalable low-cost efficiency, they offer in versatility and empathy. A person can learn without being prompted, upskill over time, change their role. They can understand linguistic nuance and find creative solutions to stubborn customer problems (given sufficient managerial backing, of course).
Furthermore, customer preferences vary based on the circumstances. Sometimes you just want a prompt and accurate response without having to deal with a person, making a chatbot perfect. Sometimes you want to know that someone is actually listening to your perspective, and dealing with a chatbot would be infuriating.
Optimally, you should be using chatbots to automate basic tasks and improve efficiency, but with a clear escalation system for handling the complex cases that warrant human intervention. That way, you can get the best of both worlds: human ingenuity and machine performance.
Chatbots aren’t magic bullets that can make all your customer service problems go away and drive huge improvements in your conversion rates, but they are exceptionally effective when used correctly. Deploy them alongside well-trained support assistants and they’ll save you time and money.