When a recession hits, companies turn cautious with their expenditure. The customers try to buy smartly and cut costs to things they don't want, which causes companies to cut their own costs just to stay in business. Many companies that are spending money on advertising suddenly have this dilemma: let's cut advertising expenses, but from which channel should we cut so we don't lose our momentum in a long run? Without accurate attribution data from their traditional & digital media channels, most companies would choose to cut overall ad spend down or shut down certain channel they deem "not worth it". The effect of the budget cut, historically, is devastating! For instance, in the first quarter of 2009 after the 2008 Recession, advertising spend in the U.S. dropped over 10 percent. And without knowing the right attribution to give the correct percentage of credit to each channel, traditional or digital, for every purchase made, proved to be detrimental in momentum for those companies later that year.
Luckily, even if there is an economic downfall in the near future (I am not pointing fingers here), our advertising tech stack has reached a point in which we could finally tie all media dollars together into one dashboard, and learn how each channel affects a consumer's decision to make a purchase. Yes, you read that right, I mean to tie traditional media, such as, National TV commercials, Postal flyers & even radio into the path of conversion with the regular online channels to tell the FULL story of a customer journey. Sounds unlikely? It probably means a lot of work, but I believe, to a certain point of accuracy, it is possible. No doubt that this project would be immense and costly, but it is an investment major brands are pursuing as we speak.
In fact, the actual concept is quite simple. Currently, working with a DMP is probably your best way to attribute your offline purchases to your online campaigns. Now, let's say an advertiser could mail out a promotion flyer on June 1, 2017 & start a display campaign to target the same households a week later. Using the same concept, I could potential connect the postal flyer recipients to my online campaigns and set a chronological order that shows the household receives the flyer first & is hit with display impressions a week later. If this flyer recipient then converts offline, our DMP partner will be able to tie this multi-touch journey into one path, explaining how much spend each channel took before the purchase event. For TV commercials, it could get a bit inaccurate to estimate the number of households that actually watched those commercials, but with the TV stations' help, I believe we could come up with a solution to report the number of households in certain regions that was hit by a TV commercial impression. Set top box data is currently the most popular solution to accommodate this challenge, although the data won't be able to tell if people are actually watching the TV. Then the process repeats again: we tie household data back to DMP and integrate into our attribution with the rest of the online channels! Once the business tackle down all the major traditional media spenders, it will have itself a pretty decent traditional & digital multi-touch attribution model.
And this is where the fun part begins. The potential disruption of this practice will change the media world forever. We have reached a level of tech stack that would allow us to essentially identify how effective the traditional media versus that of digital is on the individual customer level. This, ladies and gentlemen, is a significant feat and I believe we now have the capacity to do!
The other day, I was talking to one of my ad tech buddies about this and he pointed out a weakness in my argument that I found it very intriguing. Clearly, I won't be able to accurately prove if the household did actually read the postal flyer or saw that one TV commercial right after the 9 o'clock news. Therefore, I shouldn't be allowed to count all flyer recipients or TV watchers in a certain geographic area as "touched" by traditional media and to integrate that with the online channels. It was a good challenge that I thought about for quite awhile. Eventually, I decided my rebuttal to this argument is that, just like online ads, viewability is always problem. But the reason you would spend hundreds of thousands of dollars on TV is so that, hopefully, everyone watching it will see your product. Therefore, there is a predetermined media cost per person within the broadcast range. Adding that cost of TV impression to the customer journey does make sense to determine a more accurate conversion acquisition cost.
Thus, I believe it is now the time for brands to start dipping their toes into that sweet, sweet attribution modeling to combine all channels, traditional and digital together. If there is another recession or economic fallout, this attribution model could save companies millions of ad dollars. Even without it, business would be able to actually pinpoint channels with saturated spend and reallocate among channels to optimize for conversions. Potentially, the completed multi-touch attribution would be able to tell advertiser how to run with a lean marketing spend. Yes, media vendors may hate me for it, but this is where we are heading into whether they like it or not. I can see the next 3 years to be the most crucial stage for experimentation and the mastery of a comprehensive multi-touch attribution model.
To learn more, follow our Twitter, @BangPixels
Jim Chan is the Founder of BangPixels Interactive, a data-centric global digital marketing agency. Contact us to see what BP can do for you!