THE RKG SOLUTION:
As we see it an attribution management system needs to meet the following criteria:
- It must be comprehensive, incorporating information from all tracked marketing initiatives;
- It must be sufficiently robust to give directional confidence;
- It must provide actionable insights;
- It must be reasonably priced
RKG's system tracks across all media using the advertiser's existing web analytic system tracking parameters. It smartly parses organic traffic into brand and non-brand buckets, and ties into our paid search reporting and bid management at the most granular level.
Additionally, we incorporate data from display advertising platforms for view-through data and offline direct mail programs to better understand the interplay between marketing channels.
We provide both a flexible heuristic system, and custom-built advanced statistical models for those who feel the need for the latter.
We have a sales-based reporting interface, with interesting path analysis reporting, detailed attribution reporting, and "what if" scenario modeling for those using the heuristic controls. At this point we do not intend to fold cost data into the reporting interface as pulling in redundant information would add significantly to the price tag without materially improving the product.
A beautiful dashboard that controls nothing is a waste of time and money. We see no value in staring at attribution data day-in day-out. The attribution data should be fed to the platforms that actually manage marketing programs so that the focus can remain on the core marketing activities, not on the data correction mechanism.
We believe the actionable insights will be largely limited to paid search, comparison shopping engines, and affiliate programs. Direct mail, display advertising and email already have much more accurate mechanisms for testing their value, namely hold-out tests. The results of those tests can help tune the model, but the models probably shouldn't replace a sound testing regime.
RKG's system exports order-level attribution to advertisers and their agencies at the most granular level provided by the web analytics tags. This allows the managers of those programs to drive by the appropriate metrics.
With our friend Kevin Hillstrom, we can also offer consulting projects to provide forward-looking guidance on more complex questions of marketing mix.
Because the attribution system itself manages nothing, we don't think huge ongoing fees for the SaaS are justifiable. We charge a fair price for our work.
Consulting arrangements with Kevin will be priced by the job.
We think we've built the right-sized solution for the problem. What do you folks think?