Articles | December 16, 2020 | 3 min read
What’s the Difference Between Last-Click Measurement and Deterministic Measurement?
Few companies in the AdTech/MarTech ecosystem can execute an omnichannel acquisition campaign with complete deterministic measurement. This deficiency means few marketers can understand the true, incremental ROI (and relative contribution) of each channel utilized at the conclusion of a campaign. Instead, marketers are left relying on flawed, outdated attribution models (one of the most relied-upon of which is last-click). To help marketers reverse this trend and move their media programs towards attribution methods that mitigate media waste while increasing campaign response, Zeta would like to explain the difference between last-click measurement and deterministic measurement.
Last-click measurement is a method of attribution in which marketers assign credit to a media campaign for driving a conversion when a consumer:
#1 – Clicks on an advertisement or link
And then…
#2 – Converts on the marketer’s website (in the same session and on the same device).
Make sense?
While there are many possible reasons, marketers typically use last-click measurement because it’s easy to understand and easy to implement compared with other approaches.
There are three major shortcomings when it comes to last-click measurement.
Most consumers are exposed to a variety of advertising touches over time. Not only do they rarely click on those touches, but most consumers’ buying journey extends across different devices, browsers, and sessions. These factors mean last-click measurement often assigns too much credit to the final touch (i.e., the source of the last click), and too little credit across all the other advertising touches.
People who click and covert via the various available purchase paths are very low in the sales funnel. As such, some of these people will likely convert without any exposure to advertising. Last-click measurement doesn’t account for this reality, which means the customer acquisition cost reported by this attribution method will not reflect the true cost.
Last-click measurement is a conservative attribution methodology that limits marketing campaign costs at the expense of business growth.
Deterministic measurement is a means of attribution where target audiences are defined by personally identifiable information, before being segmented into homogenous test and control groups. Once separated, these groups can be activated via digital media (and the results measured) using personally identifiable transactional data from the advertiser.
There are four main benefits to using deterministic measurement.
When properly set up and executed, deterministic measurement establishes why people really convert. Whereas last-click provides a fuzzy correlation between a measurable input (i.e., the last click) and a desired outcome (i.e., a conversion), deterministic measurement leaves no doubt by proving causation.
To know the true cost of acquisition, a marketer must know the incremental effect of their campaigns. Which is to say, they need to know if someone would have become a customer if the marketer’s campaigns never existed. Deterministic measurement is the only way to illustrate the incremental impact of advertising.
Unlike last-click measurement, deterministic measurement can be applied across multiple conversion types including ecommerce, in-store, and call center. This robust accommodation alleviates blind-spots created by last-click measurement (which can only be used for ecommerce conversions).
Deterministic measurement is a critical component of the iterative multi-variate modeling process required to drive campaign performance and scale over time.
Knowing the difference between last-click measurement and deterministic measurement is just a starting point. To actually start seeing better success with your ad campaigns, you need to know how to implement a deterministic strategy.
Without a very strong identity layer that can move with real people as they float between offline and online channels deterministic measurement can’t work, so start there. In most cases, obtaining this identity layer will require parenting with a technology partner, like Zeta, that can offer both deterministic and probabilistic methods of identity management, as well as encompass key consumer identifiers such as name, address, email address, cookie IDs, mobile advertising IDs, and IP address.
Once identity is solved, use the data obtained to measure every campaign and every channel to the highest of possible standards. Among other things, strive to determine what the incremental return on investment was, and follow through with an in-depth analysis of the value and quality of the acquired customers.
To succeed with deterministic measurement, the desire for adoption must come from the very top (the executive team). From there, it will percolate and permeate an entire marketing organization. More importantly, if adoption starts at the very top, it will always receive the resources and support it needs to succeed.
If you have additional questions about the difference between last-click measurement and deterministic measurement, please reach out to Zeta with your comments and questions here.
What is last-click measurement?
Last-click measurement is a method of attribution in which marketers assign credit to a media campaign for driving a conversion when a consumer:
#1 – Clicks on an advertisement or link
And then…
#2 – Converts on the marketer’s website (in the same session and on the same device).
Make sense?
Why do marketers use last-click measurement?
While there are many possible reasons, marketers typically use last-click measurement because it’s easy to understand and easy to implement compared with other approaches.
What are the challenges with last-click measurement?
There are three major shortcomings when it comes to last-click measurement.
- Inaccuracy
Most consumers are exposed to a variety of advertising touches over time. Not only do they rarely click on those touches, but most consumers’ buying journey extends across different devices, browsers, and sessions. These factors mean last-click measurement often assigns too much credit to the final touch (i.e., the source of the last click), and too little credit across all the other advertising touches.
- Doesn’t reflect the true acquisition cost
People who click and covert via the various available purchase paths are very low in the sales funnel. As such, some of these people will likely convert without any exposure to advertising. Last-click measurement doesn’t account for this reality, which means the customer acquisition cost reported by this attribution method will not reflect the true cost.
- Impairs long-term growth
Last-click measurement is a conservative attribution methodology that limits marketing campaign costs at the expense of business growth.
What is deterministic measurement?
Deterministic measurement is a means of attribution where target audiences are defined by personally identifiable information, before being segmented into homogenous test and control groups. Once separated, these groups can be activated via digital media (and the results measured) using personally identifiable transactional data from the advertiser.
What are the benefits of deterministic measurement?
There are four main benefits to using deterministic measurement.
- Measures causation vs. correlation
When properly set up and executed, deterministic measurement establishes why people really convert. Whereas last-click provides a fuzzy correlation between a measurable input (i.e., the last click) and a desired outcome (i.e., a conversion), deterministic measurement leaves no doubt by proving causation.
- Accurately reflects the true cost of acquisition
To know the true cost of acquisition, a marketer must know the incremental effect of their campaigns. Which is to say, they need to know if someone would have become a customer if the marketer’s campaigns never existed. Deterministic measurement is the only way to illustrate the incremental impact of advertising.
- Accommodates all conversion types
Unlike last-click measurement, deterministic measurement can be applied across multiple conversion types including ecommerce, in-store, and call center. This robust accommodation alleviates blind-spots created by last-click measurement (which can only be used for ecommerce conversions).
- Informs modeling & optimization
Deterministic measurement is a critical component of the iterative multi-variate modeling process required to drive campaign performance and scale over time.
How do you deploy deterministic measurement?
Knowing the difference between last-click measurement and deterministic measurement is just a starting point. To actually start seeing better success with your ad campaigns, you need to know how to implement a deterministic strategy.
- Start by solving identity
Without a very strong identity layer that can move with real people as they float between offline and online channels deterministic measurement can’t work, so start there. In most cases, obtaining this identity layer will require parenting with a technology partner, like Zeta, that can offer both deterministic and probabilistic methods of identity management, as well as encompass key consumer identifiers such as name, address, email address, cookie IDs, mobile advertising IDs, and IP address.
- Then measure business outcomes
Once identity is solved, use the data obtained to measure every campaign and every channel to the highest of possible standards. Among other things, strive to determine what the incremental return on investment was, and follow through with an in-depth analysis of the value and quality of the acquired customers.
- Solicit organizational commitment
To succeed with deterministic measurement, the desire for adoption must come from the very top (the executive team). From there, it will percolate and permeate an entire marketing organization. More importantly, if adoption starts at the very top, it will always receive the resources and support it needs to succeed.
I have more questions about the difference between last-click measurement and deterministic measurement…
If you have additional questions about the difference between last-click measurement and deterministic measurement, please reach out to Zeta with your comments and questions here.