Testing Welcome Messages in Facebook Messenger

I write about messaging for business and nonprofits at @Mssg (atmssg.com)
In this post we’re going to break down the Welcome Message — the most important part of the Facebook Messenger Destination Ads. When the user clicks a Facebook Ad, they can now be routed to Messenger rather than a URL. When the user routes to Messenger, they receive a Welcome Message, which is created during the Ad setup. Once the user engages with the Welcome Message, the Messenger campaign (or bot if you must) can take over the conversation.
This post will focus on best practice and testing the Welcome Message. If you need a refresher on Messenger as a Destination Ads or the setup, read about the basics of Messenger as a Destination Ads or an overview on Messenger Ad setup in previous posts.
The Welcome Message is the single most important interaction for any messaging campaign. The Welcome Message arrives at a critical time. The user has already clicked the ad, which is paid for, but the user doesn’t reach the developer’s servers until they engage with the Welcome Message. It’s the absolute top of the funnel.
So if the user receives the Welcome Message and doesn’t engage, that means that the advertiser paid for the click, but never sees the clicker. It’s the worst possible situation. Increasing Welcome Message conversion (we call this the Conversation Rate) is the single best way to increase results from Messenger campaigns. Testing and optimizing the Welcome Message should be the first objective of any campaign.
What can we test and how do we test? First, it’s important to build or use a platform that can track what ad, audience and Welcome Message the user has engaging with. Facebook tells the advertiser how many people click each ad. The messaging platform should track number of conversations that started and connect the conversation back to the ad that was clicked.
The Variables To Test
Audience: Facebook says that the ad will only be shown to users with Messenger installed. This is a useful filter, but the campaign should test what audiences are most likely to engage after the click.
Traffic vs. Conversion Objectives: Facebook allows Messenger as a destination for campaigns with both Traffic and Conversion objectives. Messenger conversations do not include a pixel, but Facebook would presumably know the conversation starts.
Image: The Welcome Message can include an image which will have a large impact on user engagement. Obviously different images can be tested with different audiences. It also makes sense to test if the same ad image works better or worse than a new image.
Message/Copy: The Welcome Message can include a Title and Subtitle, each can have up to 80 characters. It’s possible to test questions, calls to action, surveys or any type of copy that gets the user to click.
Response Type: The user can engage with the Welcome Message by clicking template buttons or quick reply buttons. The user can also reply to the Welcome Message, but it’s a little harder to track where they came from (but possible).
Best Practices: Success varies widely and can probably be best compared to launching Facebook Ads themselves. The value is there, but the first few dollars spent on ads might not present amazing results. More pithy — you need to test and optimize.
Tests have shown conversation rates as low as 5% and as high as 75%. Best practices are hard to come by in such a new space, but we’ve learned a few things.
1. The Welcome Message should be a relatively-lower ask. Meaning, the Welcome Message should be an easier ask than the Call to Action from the Ad. Anyone that clicked the Ad should undoubtedly click the Welcome Message.
A good flow might be:

In a bad flow, the Welcome Message ask would be a bigger commitment than the ad click.

2. Rhetorical Questions can work better than surveys. In the Army example, the question of whether the person wants to improve is rhetorical. Of course everyone wants to improve. This seems to work better than a multiple choice question of “What would you like to improve?” Strength / Discipline / Skills
Even a seemingly easy response might add a little friction.
3. It’s too early to tell, but using Conversions type Facebook Ads might work better than Traffic. This is definitely worth testing.
4. Confusion might be one reason that Ad-clickers don’t engage with the Welcome Message. All of this is new, and most Facebook users haven’t gone through the click-to-Messenger flow yet. In the Ad itself, it’s a good idea to mention that the user will route into Messenger so they are expecting it.
The really good news is that once a conversation starts, the results are amazing. Response rates are over 50% and overall engagement is off the charts. Messenger and conversations are a new canvas to drive value. This post is written in the spirit of sharing early learnings and hoping to advance the overall space.
Please like the post and leave a comment below.