By Travis Biechele, Director of Digital Media
January 28, 2015
As digital media has evolved over the years, we have seen advances in creative technology (rich media, mobile, video, etc.). These tend to receive a lot of attention (and rightfully so) due to their visibility in the marketplace and how innovative these new types of creative units are.
In digital media planning, we have also seen technological advances. Algorithms, what I call the ‘unsung heroes’, work behind the scenes to drive better results by finding the right audience while optimizing toward the best placement and creative. An algorithm used in media buying is a mathematical formula that performs calculations, processes data and can perform automated reasoning based on repetition. These formulas are adaptable to current conditions and make a campaign stronger by adjusting and concentrating budget on areas that are driving the best results.
Digital Media Planning the Old Way
In the past, digital media (mainly banners on websites) was purchased and optimized by separated targeting attributes. These targeting attributes—which were purchased individually and could not be combined—included:
- Demographic (age, gender, household income)
- Geography (city, state, country)
- Behavioral (web-browsing behavior)
- Contextual (content on websites or webpages)
- Daypart (days, time of day)
Performance was manually monitored and optimized. Reports were pulled daily or weekly. Strengths and weaknesses were identified and budget was shifted from one targeting attribute to another (ie. contextual targeting budget was reduced and reallocated to behavioral targeting). This process involved contract revisions between Agency & Client, as well as Agency & Vendor, which could take 2-5 days to complete.
Digital Media Planning the Smart Way
Digital media buying and optimization happens more quickly now with the advancement of media buying technologies. Media vendors now use complex algorithms to drive stronger performance for their clients. Real-time bidding (RTB) is an evolved algorithm and is currently the most commonly used in regards to banner and video placement buying today.
RTB combines all desired targeting attributes and allows advertisers to purchase inventory when it is needed instead of purchasing placements ahead of time. This makes a campaign more effective and efficient because the purchasing of media happens in 1/10 of a second. Optimizing campaigns with RTB is just as effective and efficient because optimization happens any hour of the day and occurs just as fast. Poor performing placements are constantly eliminated as the algorithm learns which combination of targeting and placement drive the best end result.
RTB uses the same targeting attributes we have used in the past along with a few additional attributes (listed below), but combine them instead of targeting each separately.
- Day of the week
- Time of day
- Device (desktop, smartphone, tablet)
Choosing a Vendor
Different media vendors have developed their own media buying algorithms. The best way to determine if one vendor is a better fit is to test multiple vendors during the same period of time with the same creative. This will alleviate as many variables as possible for an A/B test. Be sure to measure all vendors against the same key performance indicator (KPI) to determine the strongest partner to continue working with in the future.