loveholidays: Fixing Paid Social Measurement to Make Revenue Rocket

The problem: loveholidays’ paid social output was under performing, delivering a lower ROI than the business’s high growth targets required.

When we were tasked with improving the paid social output for the UK’s fastest-growing travel company loveholidays, there were a number of problems that we would uncover and later overcome:

  1. They were too reliant on retargeting campaigns to deliver positive ROI
  2. They were attributing paid social performance incorrectly, undervaluing the channel’s contribution
  3. Their internal attribution modelling had been setup with a bias for Google Ads

The first thing that we needed to do was to ascertain why the paid social (specifically Facebook) was being negatively credited by the company’s internal attribution modelling system.

Our approach was to first dive into the bookings that Facebook was claiming credit for. We did this by setting up an API with Facebook, as booking references and order numbers are not visible as standard within Facebook Ads Manager.

Once we had the API setup, we could see which bookings Facebook believed it should have been receiving attribution credit for. This allowed us to look at the attribution path of the booking to see what was happening – we noticed to problems:

  • Not all paid Facebook traffic was coming through with UTM parameters attached, so instead of being credited to Paid Social credit was being given as Referral
  • When Facebook and Criteo retargeting channels appeared in the same conversion path, they were sharing a single credit rather than getting their own credit: so if there were four paths in a booking they should in theory get 0.25 bookings each, but they were instead getting 0.165 bookings (or half of 0.33).

At loveholidays, their internal attribution engine was setup to give Organic Social credit to all traffic that had the source of Facebook and the mediums of either Referral or Organic.

Due to the fact paid Facebook traffic was being attributed to the Referral medium, this meant that loveholidays’ internal attribution engine was incorrectly giving credit for over 60% of Paid Social traffic to Organic Social.

So this meant that we needed to put a test in place to figure out what percentage of Organic social traffic was being correctly (or incorrectly) credited to Referral.

To do this, we simply switched off Paid Social activity for seven days, and added Bitly links to all of their Organic links, so that we had definitive evidence for how much traffic from Organic was being attributed to Referral.

It turned out that when we used Bitly links, 100% of Organic traffic was being correctly attributed to Organic. With no traffic coming from Referral. This meant that we could be confident that all Referral traffic was coming from Paid Social.

This allowed us to reconfigure loveholidays’ attribution engine to credit Paid Social with all Referral traffic. We then worked with loveholidays to configure the model to ensure that Facebook and Criteo were being awarded the correct amount of credit.

Being able to correctly attribute Paid Social properly allowed loveholidays to invest more in the channel, and gave them the confidence to invest more in upper funnel activity, easing their reliance on retargeting.

This resulted in a +68% increase in Paid Social revenue, +64% increase in transactions and +5% increase in ROI.

Can we help your business with its Paid Social performance? Contact us today.