Attribution is by now a widely discussed topic in the performance marketing world. While challenges and concepts are the same everywhere, the level of implementation and adoption vary hugely from market to market.
Being a Germany-based international performance marketing agency, nonstopConsulting has insights into several diverging markets. The following reflects our stance distilled from the current state of discussion in Germany, which we see to be relevant elsewhere as well.
An integrated evaluation of all online marketing activities using customer journey analysis is quickly becoming the de-facto standard for performance seeking advertisers.
Customer journey analysis gives valuable insights into dependencies and cross-contribution of all marketing activities used. Marrying those insights with purely data driven marketing instruments such as RTB to create campaigns that automatically maximise their ROI is a big trend for 2014 and beyond.
Affiliate marketing needs special consideration
However, attribution models and customer journey insights are very individual to each advertiser. This makes it very hard to reflect them in established systems such as affiliate marketing.
By design customer journey contributions are not compatible with typical “the winner takes it all“ setups like the last-cookie-counts principle that is widely used nowadays. There are clear technical obstacles on this, as neither affiliate networks nor tracking systems employed are sufficiently equipped to handle customer journey contributions.
Also all kinds of very popular publisher models such as loyalty and cashback sites would not work anymore in a customer journey setup. Not convinced? Then try to explain to a cashback user, why she is only getting a fraction of a promised cashback due to her previous clicking of different ads.
In that light, simple and transparent attribution models that always reward publishers with a foreseeable commission for any sale referred will continue to be needed.
Intelligent use of customer journey insights
How do you solve this? Rather than trying to force the full logic of customer journey attribution models onto performance marketing, the relevant learnings taken from customer journey analysis should be implemented in the commission models used. This can be done step by step in a structured process:
- Categorise all publishers by business model (voucher, cashback, loyalty, price comparison, content etc.)
- Track and measure the typical contribution in a customer journey context for each business model (using one’s own preferred attribution model).
- Measure the true value of business driven by each model (using KPIs such as customer retention, AOV, goods returned etc.)
- Calculate an average commission for each business model, reflecting contribution and value measured.
- Assign publishers individual commissions based on their business model.
Typically, this would result in reduced commissions for publisher models that are relevant towards the end of a buying process, such as voucher or cashback sites. By contrast, commission for content driven business models would go up, reflecting their higher impact in engaging new customers.
Too complicated? Do it step by step
Even for advertisers where customer journey analysis and business intelligence are not available, the outlined approach can work. Instead of using customer journey contributions, just reflecting specific KPIs can already make a big difference.
A lot of data is already available in the affiliate networks (such as AOV, returned goods, sold product categories) and more can be easily added in the course of sales validation (e.g. new/existing customer, GP per sale, use of vouchers ...).
Using this data and running a number of pivot reports on it will yield quite enough insights for initial commission model adjustments. As more data becomes available, further refinements can be done.
While not trivial, gaining valuable insights from customer data and using it for performance marketing is not rocket science. Data and tools that can assist you are already available and access gets constantly easier.
There is no need for a fully fledged implementation of any sort, insights can be gained and used gradually. Every time new results are available they can easily be implemented by simply adjusting the commission models used.
And as a side effect of this process, the diversity and number of contributing publishers in a typical affiliate program will improve dramatically.