How to get valuable customer feedback online with chatbots

Équipe SnatchBot, 12/09/2019

How to get valuable customer feedback online with chatbots

“Your feedback is valuable to us…” is a common refrain among business feedback mechanisms, but what’s the point if a customer can’t actually say what they want, especially if it can’t be encapsulated in a 10-5 score. Chatbots can provide a better feedback loop and improve the customer relationship, while providing better data to the business.

Marketers, business leaders and data-based analytics tools like nice round numbers. That’s why most customer satisfaction surveys are fixated on the 1-5 score, be it numeric, stars or thumbs up. Even worse from the consumer perspective is the request to “please leave feedback” but then only provide a great/good/neutral tick list. And where did “could do better” or other feedback options go?

As a customer, even in cases of simple feedback, I’d like the businesses I interact with to do better or to acknowledge success. I love to heap praise on retailers that stick a little freebie in a package, or go a little extra for great service through serving staff, a service agent or other member of the team.

In other areas, such as health and education, patients and students are often asked to rate their level of service, and a chatbot represents the ideal mechanism to help generate feedback or boost engagement rates where a greater level of detail is needed to understand how those audiences are progressing.

Whenever businesses think about “How can I get customer feedback online?” increasingly, the answer is through a chatbot. But where can you describe those type of feelings and endorsements in a 1-5 rating? Would “3” or “4” be okay - and what if a company went all out to help you, where’s “5+” or a “6”? That’s why a chatbot is more likely to take over the customer feedback form, imagine:

“Hi there, how did we do with your last transaction. Was it great, did we do okay or can we improve our service?”

 

TYPE RESPONSE HERE…

Look at all the open space for a customer to think about how the pet store put in some extra treats with their order, or the dentist made a child feel relaxed and reduced the sense of fear, or the new garage spotted and sorted a random fault that a previous garage had missed for a couple of years!

How to collect customer feedback with a chatbot

Businesses are adopting chatbots for customer service, internal worker services and other uses in huge numbers. The non-technology consumer and business world is already used to chatbots, and the technology is not new or something that people fear.

According to a 2018 survey, 72% of chatbot users found them helpful and informative, a percentage that will have grown over the course of 2019 as bots get smarter, more approachable and useful.

One example of feedback apps include MHAssistant, a dedicated app for Malaysia Airline customers which includes the ability to add photos alongside text feedback and ratings to highlight the passenger experience.

Adding a feedback mechanism to a chatbot is a trivial task. Extending the original chat, or providing a reminder message at some point after the original chat and any arising transactions or interactions have taken place can be added to Facebook Messenger, SMS chatbots, your app or to get feedback via your website the next time a customer visits.

All it requires, assuming the chatbot keeps legitimate and legal logs of interactions and ensures the user approves and opts-in to future messages, is a positive “hi there, can I get your feedback for one minute please?” And then the key questions that provide data for the business to seriously improve its service.

Ensure you add balance or a positive factor to the chatbot, by badging it as a feedback/compliments bot rather than falling into the easy trap of calling it a “complaints bot.” With the ratings ticked, options chosen, and data or other feedback accepted by the chatbot, the business can collect and analyse the customer feedback.

Some businesses would still panic about how to metricate all that random, unsorted data. Well, the simple thing is to keep that 1-5 score as part of the chat, but perhaps after the customer has had a chance to think and respond. However, chatbots and AI can analyse the text and understand the sentiment, looking for negative or positive words to derive the overall tone and specific verbs or nouns to understand product or performance comments.

While positive thoughts will boost the score, negative ones at least come with a reason for the low feedback, be it:


“Delivery was too slow”

“The agent didn’t understand my problem”

“Your script sucks”

Using AI to understand the sentiment provided from a chatbot feedback questionnaire or

chatbot satisfaction survey can provide an analysis of a high-volume of chat messages. Or, if your business is smaller, a simple text log of feedback can highlight what your business is doing well, or not doing so good at.

Creative ways to ask for and collect feedback

If your business adds a feedback option and nobody cares, then you need to get creative about how the feature works.

Enticements such as entry into a prize draw are common among retailers, with most grocery stores, McDonalds and other fast food joints all offering this, highlighting just how important getting feedback is even to giant firms who can afford all sorts of research.

Other ways to encourage feedback are a discount voucher for the customer’s next purchase or use of your services. If you put a time-limited offer voucher up with the feedback session, people are more likely to make use of it which helps build a positive reinforcement loop.

Brands or those with more ephemeral businesses can gain feedback by creating a quiz for Facebook, for example, weaving in feedback questions among some points of interest and fun.

If they have negative things to say, then perhaps a “we’re sorry” discount might help convert that negative sentiment into something more positive. There are plenty of ways to frame the conversation, but most companies, even those building their first chatbot can build a useful chatbot questions list based on existing examples, or by talking to regular customers to find out how they would like to respond.

Finally, since this is their feedback, feel free to make the bot chattier and more personable than usual, using names and information to encourage engagement.

Statistics on chatbot feedback and how not to do it

When it comes to finding value in a chatbot, be it one providing a feedback form or other use, customers have a set of key priorities. Customer personalisation, understanding the value of the customer, and automated actions are all high on the list. Any business needs to appreciate these needs, no matter how simple or complex the chatbot and its questions.

Businesses also need to ensure the feedback bot is friendly and engaging, not a formal bot which might be the case for an introductory bot, or a formal sales tool.

What to do with customer feedback

Do something, do anything is the general desire from customers. Regular customers like nothing more than “our customers asked for, so we did xxx” messages in which they feel like they played some part.

The worst feeling is the suspicion that their feedback is going into a black hole, never to be seen again. Therefore, even as it plans the feedback chatbot session, the business needs to create a strategy to leverage the data.

Setting thresholds for positive feedback (perhaps rewarding the responsible staff or department with a bonus) and negative feedback (launching an inquiry into what is happening) are one approach. Seeking ways to improve business efficiency or product features can be based on feedback, if there is sufficient information.


All of which requires more than a 1-5 rating, no matter how much the likes of TripAdvisor and others promote that as a key metric. As chatbots become an everyday part of the business conversation, customers will expect to see the feedback part built into the bot, available on request or the tap of an option.

Preparing and building feedback into chatbots and building an ROI case if needed will help any business improve its performance and take the customer feedback loop more seriously. Chatbots can provide live data, alerting the business to a problem, or useful dashboards to help nudge the business in the right direction and keep its customer base happy.