By now, most marketers agree that multi-touch attribution is the correct way to measure the success of their digital campaigns. And fortunately, advances in technology and analytics have put multi-touch attribution within reach for most organizations.
Multi-touch attribution provides marketers the ability to view how each type of digital advertising vehicle (display banners, video ads, search, social) affects customer behavior. This information is invaluable to marketers, who can use it to better plan, optimize, and evaluate their media campaigns.
Historically, three factors have contributed to marketers’ reluctance to adopt attribution as a standard practice across all campaigns:
- A lengthy data collection process
- Confusion around the types of attribution models available
- Not getting timely results and ROI
For marketers to get a handle on their attribution, they first need to take ownership of their first-party data. Adopting a central data hub where customer engagement data from all channels can be collected, centralized, matched, and accessed is the key to getting that control.
Here’s how to overcome some key obstacles to accurate multi-touch attribution:
The Problem: Lengthy Data Collection Process
Today, marketers have typically already amassed the large amounts of customer data needed to fuel attribution solutions. However, this data lives across numerous independent data silos.
So when a marketer partners with an attribution solution, they should expect that a significant amount of their effort will be spent trying to access the data housed in these scattered silos. For example, just to get access to display banner campaign data, an attribution company will have to:
- Get introduced to a marketer’s ad server company and the data delivery contacts
- Agree upon optimal data delivery schedule, method, and security protocol (queued API request, real-time API request, Static CSV file, FTP, etc.)
- Determine the ad campaign data points that will be delivered (campaign ID, creative ID, time stamp, geography information, etc.)
- Test data delivery method to ensure delivery format is correct and, if using FTP, that data is not corrupted
- Perform standard maintenance by altering the data points and formatting of data delivery
Just building data collection relationships between the attribution solution and each independent data silo becomes a large-scale project that can take months to finalize.
The Solution: Centralize Data Collection and Normalization Once, Deliver to Many
Recently, marketers have turned to companies like Signal to help them tap into each of their data silos to centralize and match their cross-channel data in a single location. Through Signal, marketers can seamlessly decide the attribution partner to which they’d like to deliver their matched cross-channel data. These marketers can now have commitment-free partnerships with attribution solutions, because they no longer have to worry about repeating the lengthy data collection process each time they decide to work with a new partner.
The Problem: Confusing Models and Methodologies
Marketers who are evaluating attribution solutions for the first time are often surprised at the numerous approaches to advanced multi-touch media attribution. At first glance, it looks as if there are only two methods for attributing digital media campaign success: last-touch attribution and multi-touch attribution. Unfortunately it’s not that simple. Google Analytics alone offers marketers the ability to apply seven different types of modeling Additionally, companies specializing in attribution such as Convertro, Adometry, and VisualIQ boast of their own proprietary self-optimized algorithmic attribution models, and identifying the differentiators across these models is even more difficult than evaluating Google’s standardized models.
The reason that choosing the attribution company and model is so difficult is because multi-touch attribution is itself highly complex. Also, marketers cannot treat attribution as if it were a point solution, like partnering with an ad network for a media buy. An attribution solution needs to be interwoven through an organization for it to be most effective. To date, marketers have had to evaluate attribution companies and models by going through a standard RFP process where they are unsure of the actual differences. Fortunately, there is another way.
The Solution: A/B Test Multiple Attribution Solutions to Find the Best Fit
Marketers who utilize Signal as their data ownership solution can bypass the complexity of attribution models, and cut straight to the results and insights. Signal-powered marketers can conduct A/B testing across attribution vendors, where they can provide the same cross-channel matched data set to two or more companies or models and evaluate them based on both the customer insights generated and the effectiveness of insight-driven optimizations.
Additionally, as marketers begin to refine their media plans and marketing business decisions, they will begin to look for more complex and complimentary attribution models to develop even more granular insights. Only after a marketer has centralized and matched their cross-channel data can they conduct A/B testing of potential attribution partners.
The Problem: Not Getting Timely Results and ROI
Typically, months will pass between the time a marketer signs a contract with an attribution company and when they begin receiving in-depth insights about their cross-channel customer data. An attribution solution has to do significant data collection and then launch the media campaign before generating enough data to reach statistical significance. Only then can insights be generated by the attribution solution, and even then a marketer cannot yet measure whether they have netted a positive return on investment, because the insights generated have to be leveraged into media optimizations that also have to be evaluated for effectiveness. The extreme disconnect between paying for attribution and seeing it enhance campaign performance leaves many marketers frustrated.
The Solution: Leverage a Data Foundation for Cross-Channel Messaging
Attribution solutions have done their best to address these concerns. They have integrated with programmatic media vendors and layered in third-party data in an effort to show marketers better and faster results. However, there is an opportunity to go further. The next step is to provide marketers with an intelligent cross-channel activation platform that enables marketers to see insight such as “customers who visited the reviews section of the website after clicking on a display ad had a higher propensity to purchase in-store” and then to take action by targeting customers who did not convert in-store with a display ad or promotional email.
Marketers who leverage Signal as their data ownership solution are able to see faster attribution-driven marketing results in two ways: decreased time spent on data collection set-up, and real-time messaging via Signal’s Live Audience tool. Live Audience gives marketers a way to create their own customer segments that are based on both offline and online data and can be activated through various programmatic display advertising technologies, email messaging platforms, and other digital marketing solutions.
Own Your Data, Own Your Attribution
Applying effective multi-touch attribution to every campaign is achievable, as long as marketers have a strong data foundation.
Signal’s Fuse Open Data Platform gives marketers the ability to break down data silos and own their cross-channel data by collecting, matching, and activating their cross-channel customer engagement data. We can help you and your marketing platforms generate better customer insights and boost marketing performance by providing you with the tools to build your own data foundation.
Originally published March 12, 2015