Implementing advertising efforts in an increasingly cross-device based age has had an exponential impact on how we understand and measure the success of individual ads. Multiple placements, devices, and channels can leave advertisers with skewed results when measuring through the traditional models of measurement.
Traditional models of measurement rely solely on cookies and last click, which limits an advertisers view of the customer's path to conversion. With these models, only two marketing touchpoints are tracked and credited for the conversion while the remaining are overlooked. This leads advertisers to continuously choose the same ineffective placements, devices, etc. and disables them from receiving true insights into what works and what doesn’t.
Attribution = Data + Model
This all changes with the implementation of Facebook’s new data-driven attribution model. The term attribution simply means the assignment of credit to your marketing touchpoints, which are any efforts that touched a consumer. It’s comprised of two components: the data and the model. Attribution is a product of combining the two. Data simply refers to consumers interactions with your marketing efforts, such as impressions and clicks, across all channels and platforms. The Model is the lens used to analyze that data. Models interpret data and determine credit for the conversion which is then assigned to the touchpoint that correlates with the result. Think of the data as the input and the model as what determines the outcome.
There are two different types of attribution models, rule-based and statistical. Rule-based models assign credit according to the parameters imposed on the data by the specific rule selected. Statistical models use algorithms and historical data to determine and assign credit to each touchpoint.
Statistical models are also referred to as Multi-Touch Attribution, which refers to their ability to capture multiple touchpoints along a customers path to conversion. Statistical models produce dynamic results and learn from historical data giving you a more authentic and unbiased measurement. Rule-based models rely on the type of data you track and the rule you impose on it. They also give you the ability to track and credit one or more touchpoints. There are 5 different rules that can be imposed on the data, Last click, Last Touch, Even Credit, Time Decay, and Positional.
5 Types of Rules
Last Click - Assigns credit to only one touchpoint, the last click an individual made.
Last Touch - Assigns credit to the last Ad or other media an individual interacted with.
Even Credit - Assigns credit to impressions and clicks evenly.
Time Decay - Assigns more credit to impressions and clicks the more recently they occurred.
Positional - Assigns the most credit to the first and last touchpoints and distributes the remaining credit to the other touchpoints.
Statistical Model - Assigns credit to each touchpoint based on historical data of ad impressions, conversions, and clicks.
MTA
Multi-Touch Attribution or the Statistical Model allows for a complete and more holistic view of the conversion path, key touchpoints, and effective channels. Other models can only track one or two touchpoints leaving the remaining to be written off as ineffective. By using algorithms and historical data, MTA is able to allocate credit to each touchpoint by determining the incremental value of a touchpoints contribution to the overall ROI produced by the campaign. MTA also runs controlled experiments against your ads to show the difference in result if you weren't running them. For MTA to work, you will need two components. The first is the input which is all clicks, impressions, and media spend costs. The second is your outcomes which are purchases or other conversions. Outcomes are gathered through the Facebook pixel, app events, offline conversions, online sales, in-store sales, etc. With MTA you can: Identify which ads are driving action by assigning credit to touchpoints on and off Facebook, measure across the Facebook family of apps and services, measure across devices, publishers and channels, and assign credit based on estimated incremental value with the data-driven attribution model.
Cookie vs People Based Measurement
Both model types are beneficial to advertisers in that they allow for better insights into what is working, where it’s working, and how it’s working. The third factor that affects attribution is the use of either People based or Cookie Based Measurement. Cookie-based measurement only allows the tracking and crediting of one touchpoint, which is the last click on the device where the conversion took place. The conversion is then credited to the device/Ad where the last click occurred. The major disadvantage of cookie-based measurement is its inability to connect separate touchpoints, that happened on different devices, to one person. People based measurement differs by having the ability to connect separate touchpoints across devices to one person. This gives you the ability to view the entire conversion path from touchpoint to touchpoint and device to device no matter the time frames between them. With a more broadened view of the conversion path, advertisers are able to make more accurate decisions about ad performance, editing, and implementation.
How it works
When combining the last touch rule and cookie-based measurement, you will see credit given to the device in which the conversion happened on and to the last ad or other media that was interacted with. When combining the last click rule and people based measurement, you will see credit given to the touchpoint and device where the conversion and last click happened. Say a person performed a branded search on their home PC and then interacted with a Facebook Ad on their work PC. Using the last click, people based measurement, which touchpoint would be given credit? It would be the branded search on the home PC because the Ad interaction at work would be counted as a separate person.
On the other hand, the combination of Multi-Touch Attribution and People Based data gives you a different result. Combining these two will allow you to allocate credit to multiple touchpoints, across multiple devices, and all to a single individual. That being said, the advertiser can use this result to understand which ads are driving action, For example, you view a mobile ad on Instagram, then a week later interact with one on Facebook on your home PC, and then finally see another ad at work 2 weeks later that closes the deal. The conversion happens and credit is then distributed based on incremental value to the touchpoints leading up to the conversion.
Facebook Attribution MTA is a game changer for digital advertisers. At Grayscale, we've seen over a 300% increase in tracking return on ad spend over 90 days. We have taken our clients from an average 5.7x to an average of 18.2x.
What does that mean? It means that for each $1.00 a client gives us to spend toward their digital advertising on Facebook and Instagram we, give them $18.20 in return. A big leap forward fro the $5.70 we were averaging just June of 2018.
Sure, the ad creative is always dynamic and new. However, it was not the ads that shifted the ROI, it was the tracking.
If you want to gain a clearer picture of your digital advertising spend and return, understand your customer path to purchase, and fully understand how many touchpoints it actually takes before conversions happen, then call one of our Facebook Blueprint Certified professionals and take your business to the next level.
Comments