Top 7 tips when deploying a bot

There’s no shortage of reasons why your business should consider deploying a bot. For large enterprises, in particular, a chatbot can bring a number of important benefits. These include a reduction in the costs associated with running call centres, deepening customer relationships, and creating a variety of upselling and cross-selling opportunities. Perhaps most compelling of all is how quickly the technology is being adopted. If you’re not investing in it now, you risk putting yourself at a major competitive disadvantage within just a few years.

>See also: Wimbledon 2018: How AI and chatbots are serving innovation

That’s particularly true in financial services, and one of the reasons why we’re seeing banks all over the world embrace chatbots. And while that’s great news for those banks — and their customers — it’s important to note that you can’t implement a chatbot, walk away, and expect to get results. Successfully standing up a chatbot takes a thoughtful, disciplined, iterative approach.

7 Tips for Chatbot Success

While there’s no one-size-fits-all way to implement a chatbot, there are some best practices that you should always keep in mind. This will help you get the business benefits you seek while creating better experiences for your customers.

  1. Have clearly defined goals.
    Although it’s true that more and more businesses are launching chatbots, this alone shouldn’t be the reason you choose to do so. Instead, you need to establish clear goals that implementing a chatbot will help you achieve. As part of that process, you need to articulate the problems you’re hoping to solve with your chatbot. This will allow you to be much more focused and therefore more likely to succeed.
  2. Give your chatbot a distinct persona.
    No matter whether you’re launching a chatbot to help people with their banking or simply to order a pizza, you need to personify your bot to help bring it to life. What name will it have? What type of personality? Will it be serious and matter of fact or more casual and playful? It’s important to sort out these details early on because they will help set the tone for how people interact with your bot. As a matter of priority, make sure these interactions are not only consistent but also aligned with your brand.
  3. Develop the right channel strategy for your audience.
    There are lots of different places where bots can live. You can install them on your company website, app, or third-party conversational platforms such as Google Home, Alexa, Facebook Messenger, or WhatsApp. When deciding where to host your chatbot, it’s important to understand which channels your target audience tends to use most. Think carefully about which channels your audience is already in and start there. Often, it’s a good idea, to begin with just one channel and then expand over time.
  4. Give security the attention it deserves.
    In a world where data breaches regularly make headlines, security is imperative. All the more so when it comes to highly sensitive financial data. That’s why it’s essential to take the necessary precautions to ensure you’re safeguarding your customer data. This includes simple steps such as always authenticating users before they can access account information and limiting the types of changes they can make to their account. In addition to taking these and other steps, it’s important to be up-front about your approach to security to help reassure your customers.
  5. Figure out how your bot will handle questions it can’t answer.
    There isn’t a bot in the world that won’t make a mistake from time to time. These mistakes often arise when it simply doesn’t understand the intent behind what a user is saying. The key here is to know how to handle the situation appropriately. In most cases, this will involve acknowledging that your bot hasn’t understood something about that particular topic and immediately connecting the person to a customer service representative to resolve the issue.
  6. Roll out your bot incrementally.
    When it’s time to release your bot, we typically recommend taking a phased approach. You might release your bot to internal stakeholders and friends and family first, followed by a soft launch to just a few channels or a subset of your customers. Only once that’s done, should you go ahead and make a big above-the-line splash. The reason for this incremental approach is that at each step along the way you’re going to gain customer data and feedback, which you can use to improve and refine your bot. That way, by the time you’re ready for a mainstream release, you’ll have already worked out any kinks.
  7. Put a strong onboarding program in place.
    Don’t just dump your new chatbot into the laps of your customers. You need to think about their first experience with your bot and how you can drive adoption early on. To do so, you must educate them about your bot and the benefits they can derive from using it. You also need to help your customers understand how to best take advantage of it, focusing on some of the basic functionality from which they can easily benefit. The idea is to rack up quick wins with your audience in order to promote adoption.

Creating Great Customer Experiences from Day One

All of my advice is aimed at one thing: Creating great experiences for your customers the very first time they interact with your bot. This is critical because people are impatient. If they have a bad experience with your bot or it just doesn’t work the way it should, most won’t give it a second chance.

>See also: A CIO’s guide to chatbots: Everything you need to know

Keep in mind there’s no single solution to getting your chatbot right nor should you think of it as a one-and-done experience. Instead, look at it as an iterative multi-year project that you can continue to enhance and improve over time. That said, if you follow the tips I’ve outlined above, you’ll be in good shape to ensure that every customer’s first experience with your bot is a positive one.

Sourced by Ben Terrill, VP of Customer Success at AI-powered chatbot platform Finn AI.

Avatar photo

Andrew Ross

As a reporter with Information Age, Andrew Ross writes articles for technology leaders; helping them manage business critical issues both for today and in the future