How effective is paid social really?
Here's what we think...
This issue is constantly questioned by clients and advertisers alike when discussing digital campaign results; even more so when social channels such as Facebook are thrown into the mix.
Facebook is often referred to within the industry as a ‘walled garden’, meaning it works as a silo, with limited DMP or DSP integration capabilities.
What does this mean for social campaign measurement?
At IMA-HOME, we track results using multiple data sources, allowing us to piece data together and get a greater understanding of what results Facebook is actually driving. This includes:
- Click through conversion data from DoubleClick
- App Install data from platforms such as Appsflyer
- View/Click through data from the Facebook platform, using Facebook’s own pixel or SDK
If we can track click/view through conversions in Facebook, why not just report on these figures?
The issue here links back to the ‘walled garden’ term mentioned earlier. As Facebook has previously been unable to integrate with other DSPs, using results from the Facebook pixel is like marking your own homework. As the platform does not have a view of the other digital channels that are live, it will often attribute all conversions back to itself, giving an over-inflated view of performance.
Facebook recently announced that it is removing its 28-day conversion attribution window and reducing it to a 7-day window. This means that for many advertisers, results tracked through the platform will suddenly see a big drop off. For some, this will be viewed as a negative, as social conversions will appear to be heavily impacted. However, in the long term, this change is a move in the right direction, as conversion data will more accurately reflect the user journey.
###In the meantime, is there any way that we can accurately measure the true impact of social activity across performance marketing campaigns?
In recent years, the answer to this question would have been no. However, earlier this year Facebook started making some improvements to its own attribution product. The biggest change so far is the introduction of new server to server integrations. These integrations finally allow other key DSPs such as Google, Microsoft and DoubleClick to pass their performance data back into Facebook, meaning for the first time, advertisers can see how social performance – including view through data – is impacting activity further down the funnel.
Over the last two months we have started testing Facebook’s attribution platform and have already started seeing great results. When comparing click through conversion data to a first touch point attribution model, Facebook has seen a +450% increase in conversions. This data shows that Facebook activity is key for driving initial awareness and reaching new users, with the model then showing these users often converting further down the funnel through paid or organic search.
This is a big step for Facebook, with advertisers now being able to demonstrate to clients the value that social is contributing to sales; stepping away from last click conversion models which are not a true reflection of total performance. At IMA-HOME, this also gives us greater insight when allocating budgets during the planning stage of a campaign. We can now take into consideration the conversion uplift that Facebook activity will have on other digital channels, allowing us to plan investment more effectively.
Other recent developments, such as Bionic’s new ‘Trionic’ product, are working to further connect Facebook advertising with other platforms. However, this product works from an optimisation perspective rather than measurement, using paid search data to improve Facebook’s automated bidding.
Over the next few years we expect to see further developments in this space, as Facebook and other social platforms realise the need for greater transparency. Whilst there are still limitations in terms of server-to-server integrations, these should continue to improve over time, and will hopefully include other social platforms including Twitter, Snapchat and LinkedIn. The ability to optimise and measure social activity with all digital channel performance data considered will allow for stronger campaign results overall.