Video intelligence is more than views and clicks - rethinking video analytics with WatchingThat
- Video marketing may have hit the mainstream, but video analytics leave a lot to be desired. Barb gets an inside look at how WatchingThat aims to address this problem - by integrating video analytics across channels.
Who is watching that?
I had an interesting conversation with Cameron Church, founder, and CEO of WatchingThat, a video intelligence platform based out of the UK about video marketing and understanding video performance. Church has a solid history in the video martech and consultancy market. He was on the founding UK team for Brightcove for four years before going to Rightster and then starting his own video marketing consultancy. Watching That, founded two years ago, is Church’s foray into video analytics technology.
Here’s the big challenge with video - it’s really hard to know if it’s working, not just from an engagement perspective, but also advertising performance. Which is why we use video analytics. The problem is, most analytics focus on basic metrics that give you a high-level understanding of how popular your videos are on a single channel. It’s hard to get that bigger picture of performance across all channels.
In a recent survey by Wibbitz on video content marketing, “Navigating the Video Content Marketing Landscape”, 43% marketers run video ads on social platforms, 31% invest in online banner ads, 26% pay for search ads, 18% publish video on paid content placement on external sites, and 15% include video in their press releases. This is all in addition to the use of video on owned platforms like the website, emails, newsletter and so on:
From Wibbitz survey, Navigating the Video Content Marketing Landscape
One of the biggest challenges businesses come to Church with is getting their data to match up across all these platforms so they can get that holistic view of video performance.
Watching That is a business intelligence platform for videos. Church described it as the intersection of big data trends and the use of video as an engaging, rich format. You can collect rich contextual data around when and how a video is viewed on both your owned properties and partner channels and pull all that data into a unified space to analyze it.
A plug-in connects to many popular video player platforms (Brightcove, JWPlayer are two) and an API connects to other platforms including social media, ad servers, SSPs and so on. A wide range of data is collected including player event data, video content data, page context data, ad delivery data and ad playback data and all the data from the other systems. Good to point out this data is contextual; it doesn’t identify any particular user, so no privacy concerns to deal with here.
The data from all these systems are stitched together to give you a full view of how your videos are performing. Some example metrics that Church provided included diagnostic data such as ad fill rate, device type used, geo-location and time of day, but also which ad systems are buying which ads and where errors are coming from and what caused them.
Is there a difference in how brands leverage video and video intelligence software compared to media companies? Church said the technology they need is no different, especially when we see many brands becoming media companies themselves. But there is a difference in how they look at the data and apply it. That is something Watching That is working on now.
The maturity curve for video
Church has companies at all different stages of video use using the Watching That platform. He described a phased growth starting with experimentation - creating content and testing.
Then a company moves to build volume. At this stage, the focus is on strategic monitoring. The third stage of maturity is about achieving significant scale and growing revenue.
Most companies sit somewhere along the middle of this curve trying to figure out what is working and how to do more of what is working.
In the Wibbitz video survey, 75% of respondents spend $15k or less on video creation in 2018. Also, 75% of these marketers produce 31 or fewer videos a month (18% create 32-100 videos a month, and 7% over a 100). That’s not a lot of money to create great videos, and it’s certainly not enough to add in any type video analytics tools to measure performance.
Fitting video into the overall story
Video is not self-paced like text. A person can’t scan a video to find the information they need. That makes creating great videos challenging. Church told me that the balance is in telling the right story and figuring out the best medium to share it.
The key to improving video performance (aside from creating great videos) is guiding the visitor to it. For example, your video is not the only asset on a web page. It’s usually surrounded by text. If all you are doing is writing copy and showing a video, people may think it’s an ad, and they won’t watch it.
Instead, put the video in the copy. Guide the visitor into the video and guide them out. Church said this is a very effective approach to leveraging video because it’s a natural flow; you aren’t making the visitor do the work to figure out what the video for.
I like Watching That’s approach to video intelligence because it merges all the different channels video is used to provide a single view of performance. It’s like having a single view of all your content on digital channels using an analytics provider (of course not every organization does that).
But what I’d like to see is an intelligence solution that merges video intelligence with text-based intelligence to give that full picture of your content overall. Web pages use a mix of text and video and imagery to tell a story, engage an audience; you need to understand how it all works together.
How do you know what pieces of that web page are working and what aren’t holistically? How do you know it’s the video that’s the problem and not the text? Or vice versa? Church’s guidance of leading the visitor into the video makes a lot of sense, but still, tell me in the analytics if I’ve done that right (can analytics do that?).