Leveraging Customer Support with a Help Center Bot and Human Combo

We have recently launched a Help Center bot inside the Landbot app for around 80.000 users. The bot centralizes all our help resources — Knowledge Base, Video Tutorials (Academy), Community, FAQs, and also options to contact our Support, Sales, and Product teams — in just one place.

These resources were already available across our website, but we could often see that clients reached out to Support because they were not sure how to find help on their own. The Help Center bot makes our help resources more accessible to users and serves them in specific cases, instead of having a Support Rep intervention at all times.

The plan is to roll out the bot to more users soon, but for now, let’s take a closer look into it.

Bot Context and Challenges

Landbot’s Customer Support team is known for the amazing service we provide to our clients — if you don’t believe me, check out our G2 reviews. Some of the main responsibilities of the Support team are to maintain the relationship with current users, guide new users through their first steps using our builder, and tackle bugs or issues that may come along the way. Oh, and all that while working on projects, OKRs, and maintaining our customer satisfaction levels over 95%!

At the same time, the Product Team keeps working on new features and improvements, and the clients keep developing their bots in new and creative ways that still amaze us! This is great, but we have to face the fact that having a sophisticated product also requires more dedicated time from the Support Reps in each request.

Since clients that engage with Support have a much higher retention rate than those that don’t, we have always been cautious when it comes to the availability and reach of our Support to different plans. However, as we scale, so does our client base, which means that the Support team receives more queries each day.

Enter the Help Center Bot

Under this context, the main challenge I had in hand was to find a solution to maintain the quality of our services as we scale, using the resources available, of course. More importantly, given the opportunity, how could the Support team play a key role in helping our company grow?

So I decided that a Help Center Bot was the way to scale our Support while allowing live agents to focus only on cases that really need a personal touch. The idea is to have a bot providing fast support and information for the most common queries we get, instead of relying on Support Reps to solve all customer queries.

I know this is the moment we raise the discussion about providing support by bots instead of humans: Bots x Humans, who should “win” this battle? Is there a real incompatibility?

I don’t think so.

If we provide accurate and helpful information through a bot, there really is no need to contact the Support Team and wait for an answer. Don’t get me wrong… I still think that the personal touch of a Support Rep is necessary, just not for all cases!

Our Support Team was used to getting many “simple” queries about the platform and bots. By “simple,” I mean tickets that could be solved just by searching for an article in the Knowledge Base, watching a video tutorial, or finding the correct section inside the app.

For example, when a user just needs to access their invoices, instead of opening a support ticket, they can follow the “My Billing” flow in the bot. With just two clicks, they will be able to find the information needed as we direct them to the Subscription section of the account.

In this example, the information is exactly the same as a Support Rep would provide, but the Support Team can use this time to focus on more challenging tickets, while the bot has perfectly guided this user. Needless to say, the option to contact the Support Team is still available when users can’t find what they’re looking for — or need further help, so that they can have the best of both worlds!

‍

As most of the Customer Support Managers out there, I was on the frontlines of Support for a long time, so I understand how important it is to focus on the quality of the service provided instead of handling a huge amount of support tickets as fast as we can. I’m also aware of the effect a high volume of queries can have on the quality of the support provided, and the wellbeing of our team.

With that in mind, I wanted to improve our self-service ratio to maintain quality support for the cases that require more time and care, while the bot takes some load off our Support reps’ backs. This way, the Support reps can leave the most simple and mechanical work for the bot, and handle the most challenging cases personally. By doing that, we can reduce the number of tickets handled by each agent so that they can focus on the high-level queries, improving their skills and knowledge. It’s a win-win situation!

Now, considering the growth opportunities, the Support Team was used to getting queries from new users that just wanted to see a demo of our Product in action, as well as pre-sales queries. The Help Center Bot can also help with this by showing how Landbot works in action, leading the user to ready-to-use templates for different use cases, and allowing them to test a demo Whatsapp bot.

After engaging with our Product, the bot itself presents opportunities to upgrade, provides more information on our pricing and plans, or connects the user to the Sales team. Naturally, the same logic can be applied to any other company and product out there, showcasing what is relevant to clients and the company at a determined moment.

Finally, as any good Help Center should be, if the user does not find what they are looking for, they can still contact our Sales, Product or Support Teams, according to their needs.

I know it seems obvious, but it was very common for the Support Team to receive many requests from users trying to contact Sales or just send feedback to the Product Team. In those cases, the Support Team would redirect the message to the other departments, but let’s face it, that is not a very productive use of our Support Reps’ time, and it creates unnecessary steps for the user until they get to the right department.

By facilitating the contact to other departments through the Help Center Bot, the user can send the information directly to the correct department, avoiding friction and saving Support Reps’ time.

Bot Setup

Now, how did the Help Center Bot come to be?

I built everything without coding, naturally, using only Landbot’s native features and a couple of workarounds.

I basically created a Main Menu with the Help options available, and developed each part of the flow according to its specific needs.

Using a Persistent Menu, I added quick access to most options offered in the Main Menu, making the access even faster for users. The buttons either open another tab with the desired option (like in the Knowledge Base case) or send the user to a specific point in the bot flow.

I used Bricks for each specific branch of the flow, to keep the builder organized and working smoothly. It is possible to add as many options as necessary inside each Brick, focusing on each help area/resource at a time. In the Knowledge Base flow, for example, the information was pretty straightforward, so the flow was simple. However, you can make it as complex as you need.

‍

Throughout the bot, we give the option to click on buttons and hyperlinks to open the content on another tab (or window) so that the user can have access to the information while keeping the Landbot App open.

Support Flow

Since I’m in the Support Team, and it holds a special place in my heart, I want to dive a little deeper into the Support flow of the Help Center Bot.

I created a sub-menu offering additional Support resources, such as direct access to the Subscription section, recommended use cases, and information about our Status Page.

In cases where the users need to report an issue/bug, request setup assistance, or contact support on another subject, the bot directs them to a flow where they can contact a Support rep.

In this flow, we used the Brick Workaround Business Open/Closed (ready to use, with minimal adaptation required), which separates the flow according to our Support working hours. This way, the bot will respond differently depending on the time the user contacts us, informing them how long they should wait for a Support Rep to get in touch.

After that, we added a Multi-Question block to create a form and request the basic information to pass on to Support. We use the same approach for when the user needs to contact the Sales team.

We save the answers in variables, which we then add in the “Send an Email” (Sendgrid) block, so that an email is sent to Support with all the necessary information.

This email is then sent to our Support system, and the team will handle the query as soon as possible.

Super easy, right?

It is also possible to use our Human Takeover block and native Chats section to answer the clients directly in the Landbot platform, but we decided to center this operation on email for now.

Finally, regarding tracking, we used the Goal block marking several points in the flow to track the paths taken by the user, which we can then review with the bot’s activity in the Metrics section.

Why you Should Have a Help Center Bot, Too

The Help Center bot was the perfect solution to leverage our support while we scale, giving customers quick access to our help resources and allowing them to be independent when they need to, but still counting on the personal touch whenever it is necessary.

The bot allows users to engage in a real conversational experience, instead of having just boring buttons or exclusive access to a Support Rep. It further allows us to centralize all the help resources offered and increases the reach to other departments.

With the bot in action, we can reduce the number of unnecessary tickets handled by the Support team so that they can focus on what really matters. As a result, we can improve the quality of support provided and the work quality of our agents.

I have to point out that the main thing when building a Help Center is to just create a coherent flow. It is not necessary to have a technical background, and when I say that tech is for everyone, this is the proof of it.

The key for the bot to be effective is having a deep knowledge about what your company can offer users, and how you can help them. Know your strengths, your pain points, and use the bot automation to tackle these problems, leaving the rest to the Support Reps.

We are still measuring the results of the Help Center bot, and I plan to open it for more users soon, but it was very well received so far!

The bot complements our Support team’s activity and will take our services to another level. Instead of engaging in the “Bots x Humans” war, I think we are progressing with the “Bots + Humans” form, offering best-in-class support and ensuring that our customers are served fast and efficiently.

And based on this, of course, I definitely recommend that you implement your own Help Center Bot too!