Despite its increasing popularity, measuring the effectiveness of digital video is still a minefield. News about incomplete and inaccurate reporting of video performance by Facebook and Google have led to questions about the ROI of video investment. In addition, many marketers continue to focus on video views and are struggling to correlate video campaign metrics and business performance.
At this point in its maturity, it is time to establish meaningful measurement of rich media’s contribution. As budgets soar, the need to close the loop in reporting will grow as marketers will need to prove the ROI of what’s poised to become the biggest and most strategic part of their marketing investment.
A good place to start is to separate the truth from the hype of what determines the success of a digital video campaign. Here are three things to consider.
#1. It’s about engagement, not exposure
When views are reported, marketers should always ask how it was measured. Was it a 3-second view or a 30-second view? Was it an unskippable ad or a click to play ad? Was it played on mute?
Views may be the simplest public scorecard but it is often a measure unrelated to actual business goals. Translating success as the number of times your video was scrolled past on a feed (likely on mute) is misconstrued and not a measure of actual engagement. As most social media platforms (Facebook, Instagram, Twitter and LinkedIn) all count a view at 3-seconds (except YouTube which counts only 30-second views), the number of views is a hollow metric. Average view duration, completion rates and click-to-play rates are better measures of actual attention.
#2. Optimisation happens during, not after a campaign
When video campaigns are used as a digital equivalent of TV ads, there is a tendency to apply an offline approach to measurement where the selected video creative is broadcast, and brand lift is measured pre- and post- ad campaign. This is a waste of the potential that digital platforms offer.
Most, if not all, major digital video platforms (including YouTube and Facebook) provide easy to implement brand lift surveys, which can serve as real-time evaluators of ad impact and enable actionability. Adding an A/B test of multiple video versions on a brand lift study during the campaign can provide insight into what consumers feel and remember and enable marketers to optimise both creative and audience engagement in a timely manner.
#3. Insights come from metrics and consumer behaviour
Marketers can obtain unique insights into trending behaviour if they broaden their attention to focus on the digital breadcrumbs left by consumers as they navigate their interests online.
For example, Google’s YouTube Ads Leaderboard for the second half of 2018 revealed that six of the top 10 ads watched in Singapore came from the public sector agencies. In these cases, government agency campaigns aiming to promote awareness on a topic may not have a direct call to action button with the video advertisement, but can monitor Google search trends to see if there is a difference in what consumers are searching for. Conversation on the topic could then be progressed through search ad campaigns, targeting those who engaged with the video, or even set social conversations and sentiments on key topics as a KPI for the campaign.
Metrics such as the video drop off point, when analysed across audience segments, can actually provide unique insights into the consumer. For example, a dial test on Nike's Colin Kaepernick ad exposed a generational divide among Americans. Similar tests can easily be done by creating meaningful audience segments and monitoring video view durations and drop of points - leading to insights on the effectiveness of the creative and inputs on what engages various sub-groups of your audience.
Social listening tools can, and should, be used to measure conversations triggered, comments and shares around the topic, and changes in public sentiment; all strong indicators of a campaign’s impact on public awareness and perspectives.
The writer is Swathi Eashwer, digital consultant at R3.