By Per Pettersen on February 10, 2014
Marketing Attribution was the hottest thing in 2013 and still a hot topic in 2014, but appears to be losing some momentum as the latest and most important trend in digital marketing. There were several venture backed attribution technology companies, and both Google and Adobe made noise about their offerings. Marketing attribution and “Big Data” analytics where everywhere you looked.
As the craze has slowed, attribution has entered the mainstream as a topic of conversation and a pragmatic approach and increased fundamental understanding appears to have emerged.
Attribution is now frequently brought up by customers in almost every conversation as they try to:
• Get more out of their existing ad spend by allocating more money towards the most valuable media
• Figure out how media and CRM plays a role in winning back customers.
• Understand the value of the affiliate channel and types of affiliates.
• Understand the impact and journey across multiple devices for the consumer.
• Justify existing spend to stakeholders.
As you embark on your journey into getting a marketing attribution capability and mindset activated in your organization there are a few words of advice for the journey.
1. Adopt a Crawl-Walk-Run mentality
By not aiming for the stars immediately, but rather getting going with some moderate goals out the gate and aiming for “quick hits” it will be easier to build internal buy in down stream because there has been initial wins that everybody can see. Build the organizational capability over time and increase sophistication in methods and approach once you have mastered the level before. Simple optimization steps might give you a lot of the benefit/lift that you are looking for.
2. Focus on data quality
As you start going it is important that you canvas your own data sources and the capabilities and accuracy of your vendors data capture and cleansing so that you can come up with a plan to get accurate and true data as the basis of your analysis. It might be good, even if you use vendor tools, to be prepared to process and quality to control the raw data itself in the early part of your process. A good way to organize the data is by grouping all touch points that lead to a conversion together such as every impression and click or even mobile app events if you have that. If you have a good in-house analytic team getting them the raw data and maybe some initial path or allocation models they can play with to learn is also recommended. As with every other business intelligence initiative – junk in means junk out.
3. Iterative refinement is the key to success
Instead of focusing on getting your attribution model 100% right up front. Pick a model and approach that is reasonable and implement adjustments to the media mix as a result. Make sure the adjustments are big enough that there can be measurable deviation in results. While you are waiting on the actual results from the tweaked mix you can continue to analyze historical data. Once the results come in with statistical validity you can then implement a more sophisticated approach informed by the results of the first test. And keep this going. With every test iteration the knowledge in your organization and the confidence in methods and results increase. So test – learn – test – learn…
We are still in the early stages of adoption of attribution technologies and methodologies and organizations that start adopting the mind-set and approach should have realistic expectations and timeframes. Tools and methods are still evolving at a good pace and understanding how to attribute the entire consumer journey, including mobile and offline ad views, is still in the very early stages. As with everything though – those who do not try will not gain.