Messenger Marketing is quite a popular word nowadays. Ordinary, rule-based Chatbots are being combined with Facebook ads and intended to generate new conversion paths.

Since Botoni focused in generating leads for their customers, they’re always looking to find new ways to improve the effectiveness of their campaigns. For over four weeks now, they tested a brand new Messenger marketing campaign while using and comparing two different chatbot builder platforms. Landbot and Manychat.

With Landbot they managed to get more than double their results, compared to another highly popular chatbot provider, Manychat. They did this by carefully watching and continuously optimizing the “Click-to-Message” KPI.

The Campaign: Gym Trial Workout

Botoni developed their test campaign for one of their customers in the Fitness sector. The goal was to launch a local Facebook campaign in order to find qualified participants who are interested in a trial training. Of course the participants were qualified by the chatbot and able to book the desired slot for the training right away using the chatbot.

Research and Development of their Facebook Campaign

Example of one of Botoni’s Facebook ads

Though they wanted to focus this article on the comparison of the two above mentioned chatbot providers, it might be interesting to understand how they set up the campaign:

In the first step they identified four different target audiences for their control campaign. Here they focused on the behavioral patterns of their target audiences.

In order to produce the content for their Facebook Ads, they identified at least one problem in each target audience, a problem they wanted to solve with the product (EMS Fitness). With this in mind, they produced one picture and one video add.

This way they had four ad groups with two different ads in each.

The Conversation Flow

By Conversation Flow they (Botoni) understand it is the dialogue between the user and the chatbot. For this campaign they decided to use a standard rule-based chatbot, that means no Artificial Intelligence. From their experience they assure that when it comes to Messenger Marketing, most rule-based chatbots are more successful in Lead Generation than those who make use of Artificial Intelligence.

With specific response patterns the conversation can provide suggested answers that lead the user to the ultimate conversational goal.

What is relevant for a Conversation Flow?

In their very first Messenger marketing campaigns they often made the mistake of designing long conversations to qualify customers with the purpose of getting relevant information and data. This is why they now focus on keeping the conversations as short and effective as possible. Of course it is still important to make use of longer and more explanatory sentences in order to develop empathy towards the user. It is important to find the right balance on this.

How they developed Conversation Flows

Botoni’s Conversation Flow example

For the development of their conversational patterns they used a very simple, but effective, approach developed by themselves. Initially focusing their efforts on the dialogue goal:

This is why they tried to answer the following questions:

  1. What should the user have done at the end of the dialogue?
  2. How does the user feel at the beginning of the dialogue and what is their knowledge about the product they are about to get involved with?
  3. What expectations does the user have?

As you can see, they took a customer-centric approach in order to better manage this assessment process.

This way they were able to better comprehend the user behaviour throughout the conversation, from the start all the way to the end. The dialogue is what connects the start and the end.

Example of their Conversation Flows

Goal: The user should have completed an appointment request for a trial training at the end of the dialogue. For this they would have gathered their first name, email address and phone number.

User: The user is interested in getting more information about the product once they feel that it could help them solve their problems. Throughout the conversation, users acquired more interest and knowledge about basic product features.

Now it is the time to develop the dialogue. For this, they set milestones that have to be reached before the user achieves the ultimate goal. In their example, these are the following:

1. The user understands what EMS-Training is and how it works.

2. The users were informed they can test the product for free and without any obligations.

3. The user has shown interest in the free trial.

4. The user has selected a specific time slot.

5. The user has provided all relevant data to complete a trial training appointment.

With these milestones they then developed the initial version of the dialogue.

Botoni’s Chatbot Setup

In their campaign they tested two different chatbot providers. Manychat allows for a native chatbot integration within Facebook Messenger, while Landbot allows for their chatbot to be allocated on your own website.

They admit that at the beginning they were convinced that an integrated chatbot version within Facebook Messenger had to be more functional and useful.

But they soon found out they were wrong!

In the following section they describe the technical differences between the two platforms:

Manychat

There are two particular features. On one hand there is the possibility to address to the user directly with his name on Manychat. Also they developed their own Extension (Webview) which makes it possible to select a time slot.

Botoni’s self-developed Messenger extension for booking a trial training appointment

Hint: If you start a Facebook campaign with a Manychat chatbot, don’t use their Growth Tool “Facebook Ads JSON”, but instead use the templates that Facebook offers you when setting up the ad in the Facebook Ads Manager. The advantage here is that you can address the user with their personal name from the very first message. Unfortunately this is not possible with the “Facebook Ads JSON”.

Another important difference, when using Manychat, is that the user can be contacted again via the messenger. With Landbot, it is only possible to follow-up as long as the user provided their e-mail address or phone number.

Unfortunately, nobody responded to Botoni’s follow-up messages sent to the users who used their Facebook chatbot. However, this could also be because those messages were simply too “bad”, therefore they could not be certain that this does not work.

Landbot

Example of one of Botoni’s Landbot chatbots

In Manychat they get to see the customer first name, last name, gender and language information, when a user interacts the first time with their chatbot. Using Landbot, that is not possible.

With Landbot there are easier ways to book appointments, get phone numbers or other customer details. Botoni did not have to develop their own solution, they realised that the ready-to-use Landbot templates were good enough for their test.

It is important to mention that Manychat also provides templates to add a feature where the user can book an appointment, but in their opinion these were not sufficient for their purposes.

A decisive advantage for them was the tracking of events with the Facebook pixel when using Landbot’s solution, which is not possible through a native Facebook Messenger chatbot like the one provided by Manychat. This way Botoni was able to track each of their milestones, which allowed for them to better evaluate retargeting strategies along with the quality of their target audience.

The results compared

As already mentioned in the introduction, this article refers mainly to the result of the KPI “Click-to-Message”.

What is Click-to-Message KPI?

For Botoni it refers to the relationship between clicks on the Facebook ad and people who start interacting with the chatbot, i.e. sending at least one response.

In their campaign, they registered a click-to-message rate of 21.37% for the Manychat chatbot campaign.

With their Landbot campaign they reached 55.56%.

Botoni’s campaign figures

Why was the Landbot campaign much more successful than the one with the native Facebook chatbot?

They discovered several reasons for this. Of course, everything is based on their subjective perception. It is important for them to mention that this cannot be statistically proven.

Users did not click on the Call-to-Action button on the Facebook Ad

When evaluating Facebook ads, a difference can be observed between “Link clicks” and “All clicks”. Here they could quickly see that there was a notable discrepancy between these two values in the Manychat campaign.

They noticed that a number of users attempted to click on the text or ad copy placed above the main ad image instead of going for the CTA button below such image. Therefore they wanted to add an in-text link to the chatbot.

With Manychat this is also possible with a so called “Ref URL”. But then the first message cannot be defined. The messenger opens simply with a “Welcome-Screen” of the corresponding chatbot.

Because Landbot is in the end a normal website, they could solve this much easier, inserting the link twice at the beginning and at end of the ad text.

The user expects a website

The user clicks on a link, expecting for a website to open and not the Facebook Messenger. This expectation could certainly still be managed by the content of the Facebook ad, but they believe that they would then have even lower click rates on the ads.

Landbot made it very easy for them, because a website opens and the chat starts there automatically.

The users know they are anonymous

With Facebook Messenger we all mainly communicate to our friends, relatives and colleagues, and they acknowledge that Messenger has become a very trustworthy communication tool.

Botoni believes users are more open to interact with an automated chat because of the anonymity they have with a Landbot. They subconsciously realize that, by communicating via messenger, they are revealing their identity without knowing who is on the other side.

Botoni’s Conclusion

Botoni believes Facebook Messenger for marketing purposes is still in its early stages. Users are simply not used to interacting with chatbots. That’s why they believe that Landbot is a very good alternative to a native Facebook chatbot, especially when using chatbots for marketing purposes.

They think that a messenger chatbot should be considered as a new alternative for smartphone apps. So don’t just think about the one-time use, but about the added value that a chatbot can offer when being used permanently.

In this case, they too think that the platform (Facebook Messenger) we all use to communicate with friends and family can clearly score points.

When they started their initial messenger campaigns, they were certain of doing it with a standard Facebook chatbot and not with a built-in website chatbot. Their key learning here: Think outside the box and try everything before you make up your mind!

If you also want to get better results for your business, Go Ahead and Create Your Own chatbot today. Get started by creating your free Landbot account today!

This article (see original in German) was written by Christoph Paterok, Co-Founder of Botoni. Botoni is a chatbot agency based in Düsseldorf (Germany), which develops innovative chatbot solutions as well as high performing messenger marketing campaigns.