Attribution is defined as the process of identify the marketing touchpoints which influence a user's conversion. If there were multiple marketing touchpoints prior to conversion, the conversion event or value of the conversion can be attributed amongst the touchpoints using one of the following techniques
Single Touch Attribution models assign 100% of the conversion credit to a single marketing touchpoints, primarily based on the position of the marketing interaction such as
First Touch : The first marketing touchpoint that the user has with your organization within the Attribution Window
Last Touch : The last marketing touchpoint that the user has with your organization within the Attribution Window
First Touch, Non Direct : The first marketing touchpoint, excluding Direct and Organic web sessions, that the user has with your organization within the Attribution Window
Last Touch, Non Direct : The last marketing touchpoint, excluding Direct and Organic web sessions, that the user has with your organization within the Attribution Window
Multi Touch Attribution models assign fractional credit to each touchpoint for every conversion and aggregates all such fractional credits at an overall level. Factors today supports the following multi touch attribution models
Linear (or even spread) attribution wherein every touchpoint in the Attribution Window gets equal credit for the conversion
U-Shaped attribution wherein the first & last touchpoint in the attribution window gets the maximum credit for every conversion (50% each)
Time decay attribution wherein every touchpoint gets credit for the conversion with the amount ‘decaying’ or becoming lesser as the touchpoints age. In other words, more recent touchpoints get more credit for the conversion
Building an Attribution Report
In this section, we will provide step by step screenshots for building an Attribution Report.
Step 1 : Select the Conversion Goal
Conversion Goal in Factors can be thought of as the final user action which a campaign is expected to drive. The success and ROI of the campaign will be measured basis the number of users who perform this action and the cost incurred for the same.
Typically conversion goals are high value actions such as
Schedule a Demo form submission
Discovery Call Done
Within the conversion goal drop down, all the events received by Factors would be listed down for the user to chose from.
In addition to selecting any event, users may also chose to apply filters such as Industry, Country or any other event or user property to narrow down the scope of the report.
Step 2 : Select the Marketing Touchpoint
A user in his journey with your organization may be exposed to different types of marketing touchpoints such as
Paid Media Campaigns
Offline Marketing Events such as Webinars or Field Events
This section allows you to select the specific marketing touchpoints which you would like to include in the attribution report. Within Paid Marketing, given there is a hierarchy of levels (Source -> Campaign -> Adgroup -> Creative / Keyword) the user can select only one of the hierarchy elements for the attribution report.
Post selection of the Marketing Touchpoint, filters may be applied to include or exclude specific touchpoints. This capability becomes especially handy if you want to exclude a specific marketing tactic or set of tactics from the attribution report (for ex: excluding 'Branded' search campaigns).
Step 3 : Select the Attribution Model
The attribution report can be run using a single attribution model or users could select multiple attribution models for a side by side comparison
Example with a single attribution model (the icon on the right represents the comparison feature across the product)
Example which compares two attribution models
Step 4 : Set the Attribution Window
Attribution Window may be set from 7 days to 90 days. The attribution window determines the earliest time prior to the occurrence of a conversion event within which the marketing touchpoint should have happened for it to be considered in the attribution analysis.
For ex: If a conversion event occurs on 15th April and attribution window is set to 90 days, marketing touchpoints prior to 15th Jan will not be considered in attribution.
Step 5 : Set the Attribution Timeline
Attribution Timeline refers to two different widely used approaches for attribution, namely by Interaction Time or Conversion Time. The difference between the two models is caused by the fact that users might interact with campaigns in a certain month (say June), but the conversions may happen in another month (say July). In this case, in addition to attributing the user to the appropriate campaign (say Campaign ABC), we would need to determine whether the conversion should be attributed to the Campaign ABC's spend in June (interaction time) or July (conversion time). The two options can be described further as follows
For these examples, assume the Attribution Analysis Timeframe (selected in the calendar) is 01 June to 30 June and attribution window is 30 days
Conversion Time : All conversions which happened during the timeframe of 01 to 30 June would be attributed to campaigns with a 30 day attribution window (implying the marketing touchpoint for a conversion on 01 June could have happened on 02 May to 01 June). However, the campaign spends would be computed for the same timeframe of 01 to 30 June, even if some of the touchpoints happened in May. This approach would also not credit marketing touchpoints which happen in June for future conversions that happen in July or August. The pros of using this approach are as follows
It is easily interpretable, albeit not exactly accurate. However, it gives a useful mental model for quick analysis
It is easily verifiable : The sum of conversions and sum of spend in the conversion time report would add up to total conversions and total spend in the attribution analysis timeframe
For any specific month, this analysis can be run immediately at the end of the month without having to wait for incorporating future conversions which happen due to the spends in the current month
Interaction Time : In this case, the model would consider all the marketing touchpoints which happened in 01 to 30 June and look forward up till 30th August (considering a 30 day Attribution Window) and attribute conversions. The pros of using this approach being that it would give an accurate picture of cost per conversions and conversions that result from marketing touchpoints (and hence spend) in a specific month. However the cons of using this approach are
It is a more complex attribution model and hence takes approximately 2x longer to run as compared to conversion time models
To assess the true picture of conversions resulting from spends in a specific month (say June), users would have to wait till July or August depending on their conversion cycle
In general our recommendation is that for short conversion cycles (say less than 3 days), the conversion time analysis gives a sufficiently accurate mental model of attribution. However, in case of longer conversion cycles (in many weeks or months), it is better to use the interaction time to get a truer picture of the impact of the spend and cost per conversion
Step 6 : Select Linked Events
In this section, you can select downstream events which users are expected to do post the conversion event so that you can get an accurate picture of cost per downstream conversions.
An example would be using 'Lead Created' as the conversion event and adding Linked Events as 'Demo Completed' or 'Opportunity Created'. In this case, our attribution models would compute the cost per conversion for each of the down stream events as well. This could help unearth insights such as 'Campaign ABC brings in a lot of leads at acceptable cost per leads, but very few of them convert to Opportunities result in substantially higher cost per demo and cost per opportunity'. You may apply filters for each of these conversion events as required.
Please note that the conversion events should happen within the Attribution Analysis Timeframe.
Step 7 : Run the report and save to dashboard
Attribution Analysis is the most expensive analytical module within Factors and hence takes time to run depending on the Analysis Timeframe. We recommend that users run the report and add it to a dashboard so that they do not have to wait for the report to run (which may take a few minutes). Once added to the dashboard, the attribution reports would be updated on a daily basis.