Prioritizing Bot Deployments: The How and Why in 6 Steps
Wondering how in the world you’re supposed to know if you need a bot? Which bots do you put on which pages? What questions should you ask the user? In our latest podcast with Amanda Stevens and Billy Bateman, they discuss the 6 key steps to prioritizing chatbot deployments. Those 6 steps might just rock your world…
The Benefits of Bot Deployments
Designing and managing a bot can require some high-level understanding of your bot. But we promise it’s worth it. Invesp guesses that more than 67% of consumers worldwide used a chatbot for customer support in the past year. That’s 2/3 of shoppers using chatbots to enhance their customer-business interactions. And businesses are eating. this. up.
Why is the high demand for chatbots so good for businesses? Well for one, they save costs. [Chatbots can cut operational costs up to 30%.] They also have a really high-resolution percentage. [69% of total chatbot chats resolve from start to finish] And when things go wrong? [90% of businesses report faster complaint resolution with bots!]
Obviously, there are major benefits to deploying a bot. But what does it mean to prioritize your bot deployments? We are going to talk about understanding how your business works and how to tweak your bots to be: personalized for your company, activated on specific pages to cater best to your users, simple enough to capture users until the finish, and allow pull from multiple platforms to make life easier for the user. AND MORE! These adjustments will help your stats fly, and create not only the best use of your company bots but also the best experience for your users.
1. Are people talking about this today? Is it relevant?
Something to consider when deploying a bot is the demand behind it. Is this something that people are talking a lot about today? If they aren’t- you can’t expect that there will suddenly be an influx in traffic once you put the bot in. Use a bot to manage questions that the data shows that people are actually asking, and then fulfill those needs.
“Do you ship for free?”
“If I live in [Country], can you ship to me?”
“How long does your warranty last?”
This is the stage when you can start looking through transcripts or tweets and social media posts to find out what people really need information on. But don’t let it be an intuitive thing. USE DATA TO BACK IT UP! Bot deployments can cost so use pie charts, line graphs, tables- whatever you have to do understand what people are wanting to know more about. These questions will be more personalized for your website and thus perfect for a bot!
2. What is the complexity of the correct action?
We like ease. We’re designed to figure out the quickest ways to complete tasks while still maintaining efficiency. When deploying a bot, it’s crucial to first look at the complexity of the current state. Is it a brief interaction? Unnecessarily long processes are against what everyone wants, so why allow them for your business?
Be wary of use cases that are going to require 20+ interactions between the user and the bot. The more steps, the more people will drop out midway through, leaving the bot before they’ve completed the process. Understanding the complexity of the current state can help you simplify processes and then do testing to know where your information still isn’t coming through!
3. Does it require multiple interactions to complete this task?
In the last step, we emphasized that simplicity is key. BUT if the bot would only require one or two interactions to get the information across, you might be better off putting the answers on a FAQ page. Ask yourself: is having a bot actually saving time? You need a happy medium, not too many questions but still enough questions that the bot is necessary.
4. Does the user need to interact with multiple systems?
If the answer to this question is yes, PUT IN A BOT! Chatbots are a perfect way to integrate multiple systems to cater to the buyer. A date needs to be put on your google calendar? Bots can use integrations to complete that. Need it to send you a reminder email for your demo? The bot has got you. If a chatbot can pull data from one database into another system then it’s much easier for a user to digest that information and move in the right direction.
“Yeah, if a bot just lives on its own with no integrations, it’s not that useful most of the time. The beauty is that it can look at your calendar, it can look into the CRM or it can look into our marketing automation. And see oh yeah you were sent this email. It’s just awesome what we can do with that!”
This isn’t difficult. But with so many bots coming with multiple integrations now, people don’t need to do those easy things anymore. You’re making everything happen for them in one space, and that’s appealing to anybody.
5. What are the current difficulties around this use case?
The reason you’re using the bot is because there’s an obvious difficulty around this use case. How is asking the user 5 or 6 questions going to help this difficulty? Sometimes it’s helpful even just to prep a potential buyer, and allow them to think; “Is this an area I do need help in?”, “What am I willing to do for this solution?” “What efficiencies am I trying to drive?”
This prep can also be useful for building it as a conversational solution instead of a conversational state. How can we help you become better? A bot is great for that.
6. Do users feel comfortable talking about this to a bot?
Is your use case sensitive enough that someone might not feel comfortable typing their answers to a bot? This is something that is not often discussed but is very real. In some cases, people might feel much more comfortable talking to a live agent than a chatbot and that’s okay. When executing bot deployments, consider that sometimes having the option to switch to live if things get personal might be just what your users need.
Prioritizing Bot Deployments in a Nutshell
How are users responding to your implementations? How can you make improvements? Fine-tuning the conversational design of the way you book demos is huge for your user’s experience. The first step is to deploy a chatbot. The second is to prioritize your bot deployment. Consider the relevance of your bot, the complexity of the current action, how many interactions you have, if you can pull in multiple systems, the difficulties around the use case, and the topic comfortability. Adjust and test. Done!