Customer service chatbots: Dos & don’ts for eCommerce companies

Chatbots are the solution to a lot of the problems present in many customer service organizations. In particular, eCommerce companies have eagerly jumped aboard the chatbot bandwagon. 

It’s no surprise why: 80% of customer service requests can be automated and shoppers have a lot of questions. Chatbots can lend a helping hand—one that doesn’t log off at 5pm.

On top of that, customer service chatbots can significantly reduce the noise, that is, the low-value, repetitive questions that land in your conversation queues—freeing time up for your service team to do their best work with those customers who can’t be helped by a bot. 

But what kind of problems can a chatbot truly solve? 

The questions your chatbot should be able to take care of

1. Repetitive questions

If a question comes in multiple times a day, your chatbot should be answering it, rather than your agents. For eCommerce companies, a classic example would be “How much is shipping?” or “where is my order?”. 

2. Simple questions

If the question has a standard answer (for example, “which countries do you ship to?”) or can be answered by a quick search in your database (for example, “when will my order be delivered?”), then it belongs in your chatbot! Chatbots are perfectly suited to searches, manual inputs, and other information-fetching tasks.

These tasks are often tedious for the service team – so instead of having your service agents manually check the order status, your chatbot can fetch this information – usually with just a name, email, and the order number (or similar).

A general rule of thumb: if something is a couple of clicks away, the bot should do it.  

3. Questions that don’t require empathy or judgment

Bots and people can bring out the best in each other—when working hand-in-hand. Bots are clear winners when it comes to answering repetitive questions in large volumes, but they don’t feel empathy (yet!) and can’t make judgment calls.  Your service team should always handle sensitive, unusual, and other high-impact / high-risk cases. 

For example: a customer didn’t receive their order in time for their son’s birthday. You definitely want to have someone from your service team handle that request with the human touch that it requires. 

4. Troubleshooting questions

If you can narrow down to the root cause of a problem by asking a couple of questions – this is also a clear use case for a chatbot. 

The hardware industry has a lot of these kinds of issues, where the correct solution is identified with a series of questions: “Have you tried turning it off and on again?”, “Can you see a flashing red light?”, “Is the battery empty?”, and so forth. Again, no humans needed here!

5. Questions that need specific information in order to be answered by a support agent

Some customer service cases would require lengthy, back-and-forths with service agents to gather a customer’s information. 

For example, say an item was delivered damaged. A bot can ask basic questions before handing the customer over to a service agent: collecting the order number, a picture of the damaged goods, the delivery address, and other important information from the customer. Bot-human collaboration at its best!

So you have the use cases, maybe you even already have a bot – now what?

How to make the most of your customer service chatbots: dos & don’ts

Your customers come first, so their experience with the bot is the single most important factor to keep in mind. In the best-case scenario, their service experience with your chatbot will either increase the likelihood of a sale, or strengthen the customer relationship you already have because they get a fast and satisfactory answer to a question they have. In the worst-case scenario, their experience with your chatbot will be unhelpful or impersonal, damaging the customer relationship. Here are some dos and don’ts to avoid the second scenario.


A classic mistake we’ve seen way too often, is to think of your chatbot like an FAQ or a help center. In practice, this means chatbots that present the user with extremely long blocks of texts. Big mistake–huge! A chatbot is not an FAQ and should only provide short, precise, and to-the-point answers. If the answer to the question needs to be lengthy, it’s good practice to link to the FAQ solution from the chatbot. The rule of thumb is that a chatbot should give a specific answer to a specific question.


Picture it! It’s good practice to add gifs, images, and other visual aids to your chatbot solutions. This adds a dash of personality, and, can lighten the mood and create a sense of kinship with the bot. The old adage “a picture is worth a thousand words” also holds true here, and a gif or a screenshot can be extremely useful to your customers. This holds especially true when giving directions or explaining a workflow – sometimes an image here is more digestible than a block of text.


Decide on a tone of voice and a personality for your chatbot. This should align with your brand’s tone of voice and reflect the values you espouse. Whether your chatbot speaks casually or formally, consistency is key. If you feel your chatbot lacks charisma (a.k.a. “good vibes”), use emojis to give a specific touch to the conversation between the user and the bot. Your bot is your first line of direct communication with your customers—make sure it doesn’t disappoint!


Don’t try to “fool” your users by pretending they are talking to a real person. Call a bot a bot! You can of course give your bot a custom name and create an avatar for it, but your customers will figure out very quickly if you are trying to pass off your bot as a human. Make it clear that the interaction is with a bot and that you will hand them over to a human if needed. 


Don’t assume that a chatbot is a “set it & forget it” solution. That’s not true for your FAQ and your website, so why would it be true for your chatbot? Just like everywhere else, the content in your chatbot needs to be accurate and updated regularly. Which leads us to the last point…


Make your chatbot a priority! Assign resources within your service team to look after it. In most cases, once you’ve set it up properly, bot maintenance is low effort, so even busy service teams can plan for it. After all, a chatbot will keep cutting costs and contact volume, but only if it’s used by your customers. Considering the value a chatbot can bring to your service organization, it’s undoubtedly worth the few hours of tweaking content and training new solutions. 

Finally, if you don’t yet have a chatbot set up in your organization, this handy guide can help you get started. 


Francesca Valente

Share this article

Get CX insights, proven strategies, & the latest news in your inbox every month!

You might also enjoy

AI in Customer Service
7 min read

5 metrics to track the impact of AI on your customer service

Read more
AI in Customer Service
7 min read

7 benefits of AI knowledge bases for increasing customer service efficiency

Read more
AI in Customer Service
7 min read

5 ways to boost customer retention with AI and measure it

Read more