The results of the triggered email audit For the online store, the total revenue share from automated abandoned category, abandoned view, next purchase recommendations campaigns and product articles, increased on average from 2.3% to 5.06%. We achieved this by: Reducing the number of triggers from 133 to only 46. We removed overlaps and duplicates when necessary and opted to get rid of emails […]
Travel Affiliate Network Automated Direct Marketing and Reduced Affiliate Churn
Travelpayouts Affiliate Network is a global affiliate program focused exclusively on travel offers. It allows the user to choose from 100+ affiliate travel programs and to monetize traffic on their websites by paying for clicks on affiliate links. The offers cover flights, hotels, insurance, excursions, and car rental. The company was founded in 2011 as part of the Aviasales travel metasearch company
Company size.
More than $20 million paid out to affiliates over 10 years
300,000 users connected
100+ affiliate travel programs
300,000 users connected
100+ affiliate travel programs
Challenges
Reduce affiliate churn using direct communications
Solutions
Combine fragmented customer databases and send all campaigns from a single platform
Personalize bulk campaigns and set up 300+ triggers
Personalize bulk campaigns and set up 300+ triggers
Results
The churn of affiliates decreased year-over-year
The CRM manager saves an hour a day as thanks to the automation of manual campaigns
The CRM manager saves an hour a day as thanks to the automation of manual campaigns
Integrated with:
The website, iOS and Android apps, as well as backend developed by Travelpayouts
Marketing Automation Results
-
1 hour/ dayis saved by the CRM manager thanks to the automation of manual campaigns
-
↓Churnof affiliates decreased year-over-year
We can’t say exactly how much we earn thanks to Maestra, nor is this our goal. The main thing is that thanks to marketing automation, we have managed to reduce the churn rate of affiliates, that is, to increase their retention and convert them into making their first payment.
Although this didn’t happen immediately, we found the right criterion for determining the churn rate — namely, when the customer stops logging in to their account. The period when this may happen depends on the stage of the customer’s life cycle: seven days for new users, two weeks for those who made their first payment, and a month for permanent users.
We also look at the effectiveness of each specific offer, whether affiliates are starting to visit their accounts actively, and whether the time it takes to make the first payment decreases. This is how we are gradually trying to analyze automation results in concrete facts and figures.
Without Maestra, we would not have been able to work with such specialized segments in manual promotional campaigns, launch trigger campaigns for local promotions, and track user activity in them. We could set up triggers for large segments in any platform, but only here can we automate specialized offers.
How marketing has changed after automation
Before automation was introduced
After automation was introduced
Bulk campaigns were sent from one platform, trigger campaigns from another, transactional emails from a third, and web push notifications from a fourth. The communication load on the customer could not be controlled
All communications are sent from a single platform. Thanks to highly specific segmentation, the customer only receives relevant offers and is not bombarded with messages
Customers weren’t able to unsubscribe from campaigns and they just kept on coming
Customers can unsubscribe from campaigns with one click
When creating segments for emails, we had to manually upload contacts from different sources and ask analysts for help. There was no information about the actions, activities, or earnings of affiliates
All customer data, including information about sales, campaign interest, profile logins, activity, and earnings are stored in a single profile. The segment for each campaign is created by the CRM manager in just a few clicks and is updated automatically
We had to turn to the IT department and wait a long time for a new trigger campaign to be created, so there were only two campaign sequences
The CRM manager is able to create new trigger sequences in a few hours. This allowed us to produce more than 300 trigger campaigns
In order to analyze the results of each campaign, we had to ask analysts for help pretty much at every step
Simple analytics, such as campaign enagagement and interest are available on the platform, so we no longer need to involve analysts
Initially, we relied on another email marketing service, but it did not have the functionality of a Customer Data Platform (CDP). For instance, the same person who registered for four webinars was recorded on in our platform as four separate customers that were completely unrelated to each other.
We realized that we needed a system that could build customer profiles without any issues. In our case, Maestra turned out to be the system we were looking for. The final argument in its favor is that almost all the emails I receive are created using this platform.
How customers are segmented in manual campaigns
Manual campaigns are a very important channel for us. We never send them to large audience segments. Even if we’re launching a big promotion for all affiliates, we still show different parts of it to different segments.
Manual campaigns work well, and there will only be more of them in the future. We will never be able to automate all communications because we have too many promotions, webinars, and excursions for specific business verticals.
It’s crucial that the platform allows us to create very specific, targeted segments — for example, taking into account the stage of the customer’s life cycle, the business vertical, and the date the customer last logged in to their account. We simply build segments according to the conditions we require inside the platform, thus eliminating the need to involve analysts or the IT department.
Thanks to filters, highly complex segments are available to us — for example, affiliates who earn from X to Y amount on a specific vertical or customers who went into churn in 2019, but came back to us at a later date.
Which automated campaigns are sent to customers
Travelpayouts has more than 300 triggers configured. The segments that are used are highly specific, making it impossible to conduct A/B tests: getting a statistically reliable result on such small samples would be very difficult. Below we will tell you about some of the most important trigger sequences for our business.
Onboarding. Affiliate marketing is a complex tool. New users need to be taught how to use it, so this email series is very important for Travelpayouts.
News regarding offers. Travelpayouts has 90+ affiliate programs, there are different verticals of users. Because of this, we need to send several news articles regarding offers every day.
Previously, we had to manually send up to 10 digests on offers per day. This took about an hour, however, our email marketer had to constantly prioritize and postpone less important news.
Now, urgent news are sent promptly: the trigger checks if there are any new offers, and automatically sends an email every day. The marketer only needs to make sure that the campaign has been formed correctly.
Training on securing offers. The goal of Travelpayouts is to allow the affiliate to receive their first offer as soon as possible. The training email chain was set up and its effectiveness was immediately verified with the help of a control group: part of the segment was excluded from communication in order to compare the behavior of other users with it. The test showed a statistically reliable result. The rate of receiving an offer increased for the test group.
Special projects. Travelpayouts’ special projects are complicated. Onboarding is provided for each of them. It also tracks whether affiliates check their rating, whether they are subscribed to campaigns, and whether they log into their account. All these conditions affect which emails they will receive.
Churn reactivation. The main goal is to motivate the customer to carry out a simple action that does not involve any obligations. For example, clicking on the affiliate link and making sure that the transition is displayed in their account. The email sequences differ depending on the customer’s lifecycle stage.
Reactivation works very well. For example, we recently sent an email to several thousand inactive customers offering our help. We got 150 responses. We helped a lot of people, and they came back to us.
Next steps
Develop more triggers for user segments, taking into account each user’s lifecycle, earnings, and activity.
Make reactivation campaigns even more personalized: take into account not only account logins, but also transitions from segment to segment.
Finalize onboarding by expanding the current basic workflow, branching it out and creating a different series of events and campaigns depending on the user’s behavior (i.e., what they did or didn’t do). Our goal is to help each specific customer. We want to explain the basics to those who do not yet understand affiliate marketing and to offer more complex solutions to those who received a payment in the first month.
Set up triggers to remove inactive users: together with analysts, define criteria showing that affiliates cannot be reactivated, and automatically delete them to avoid overpaying for database storage.
Create a set of standard dashboards that can be applied to each company, for example, to evaluate the transition from segment to segment. The default reports are not suitable for Travelpayouts' needs as they were designed with e-commerce in mind.
P. S. In this article we discussed our Customer Data Platform (CDP) and Multichannel Communications modules. Learn more about each module by clicking the links, or reach out for a discovery call.