Marketing Attribution Model Alignment
One of the most exciting aspects of unifying marketing data is the opportunity to measure conversions uniformly across channels. The primary purpose of these posts is to focus on this point in time where siloed multi-channel budgets are now being aligned into one omni-channel budget. This will alter many aspects of marketing departments and many technology opportunities exist as a result. While digital advertising has had attribution technology for many years unfortunately there is not a single point of attribution due to privacy laws and the rise of walled garden media. This further underlies the need/opportunity for cross-channel, unified attribution.
From a data standpoint however it might be possible to align measurement across channels in a simplified way and/or it could be time to implement a robust solution aligning conversions with an “apples-to-apples” methodology. At the core of this opportunity is defining the point of attribution or conversion to determine success from the marketing effort. For an e-commerce advertiser this might be the “thank you” page. In this example the advertiser or marketer is saying that their spend/effort will be a success when a new user makes a purchase.
There can be more than one success metric in place and the user conversion path can be fully captured or only the last media touched. With marketing budgets mostly siloed to date there have been different methodologies used per channel. As the industry evolves and incorporates the entire conversion path per user it will be critical to understand and apply the best attribution methodology.
To restate there is an opportunity for a technology vendor and/or market research vendor to re-evaluate attribution as a concept. Ideas already exist where points can be applied to each user interaction with more or less being applied for specific points in the process for example where market researchers already know are most important in user decision making process. Even in this concept of points there are many differing points of view so likely each company will craft their own unique attribution methodology, eventually.
In terms of new attribution technology/concepts it would be ideal to be able to offline and historical data (i.e. POS sale data) as well as non-paid media data (i.e. public relations, social posts, etc.) so that value can be realized from all marketing efforts. It seems to me that ideally the data should ideally be connected to a user ID and/or CDP. The wish list likely is very large but the main key is comprehensive attribution measurement and not sample based but it is reasonable to understand that initial solutions may not cover 100% of the data.
My suggestion is to move forward today in crafting a plan of how without any new technology would we combine attribution data to provide a holistic view. This will help to determine later what technology makes sense to use as well. Data sets and the attribution points can be very different from each other which is an understandable outcome from siloed budgets. The effort at this point is high level to capture on a shared document a list of the systems and point of attribution.
It might be possible to have some success manually combining data once per quarter to at least start with a unified approach to success metrics. It might require cross-team support from IT, market research and other departments but such effort will also show the complexity of the task. Organizational change is always painful and cross-team work can give people a sense of what we are trying to accomplish so it does not come as a surprise later should we need more support.
To get started aligning your channel data first list the conversion metric or attribution point for each of the following marketing initiatives:
- Advertising — paid ads (list separate for each of the following if different: search, social, display, TV, radio, billboard)
- Advertising — non-paid ads (i.e. social posts)
- Outbound Marketing
- Public Relations
- Search
Once the list is completed it might also be helpful to identify where metrics or concepts can be aligned so that conversions are more uniform across channels. Some departments/budgets may have been defining an attribution point that is relevant to their initiatives but not to others for some reason and that is the point of this exercise to determine if channel goals can be aligned at least in concept or not. If not perhaps things should be left as they are but if the conversion mechanism can be aligned it will help bring data sets together more quickly.
Next it would be smart to identify the systems that hold data for each channel, you can list the technology name (system and/or database) next to each channel or sub-channel. This will help in determining if/how data sets might be joined. It will be good to connect with a market research person to determine how analysis, combining data sets, might be done if possible. With this work you are prepared to answer a question from management how you might align attribution across channels.
You will want to determine how quickly a new solution can be implemented and if there will be additional cost. Because of the work you have done in understanding and listing conversion metrics per channel you can start talking to vendors of all types to understand how the industry is answering these questions. Be direct and ask each vendor how they suggest solving your alignment goals with the information you compiled. Do not hold back on meetings with vendors of types, you do not have to purchase a product just for taking a meeting.
Keep in mind that some vendors will provide analytic solutions for attribution while others provide a software product to capture each conversion. You could also find a database solution as where you retain existing technology and push data into a central location. A cross device graph or vendor will surely come up but this may not be needed especially if the advertising buying/targeting is done outside the organization.
The best solution for measurement and the best for cost may not be the same so you should additionally craft out expected time of each solution for your company to be able to fully realize the opportunity. For some organizations this is daunting work to be sure but absolutely worth the effort to be able to thoroughly answer how you can measure marketing effectiveness across channels in an “apples-to-apples”, fair way. You will at the same time be enhancing your resume to cover marketing vendor selection and management.
