Advertising, and in particular sponsorship, used to be a simple process. A brand would buy the rights to put their logo on a team and would enjoy the returns of being associated with a sport franchise, mostly via TV air time. However, with more than a billion photos uploaded a day, along with videos, the metrics have changed dramatically. To the point that sometimes, exposure on social media far exceeds any TV broadcast. With far richer engagements and a life span that extends sometimes weeks after the event, it is a gold mine that has yet to be exploited. Until now: Newly launched SnapRapid, using deep learning technology, can now extract the real value of sponsorship by analyzing a multitude of channels, giving sport franchises and brands a complete picture of their marketing efforts. We sat down with ceo and founder Russell Glenister to learn more:
– A little about you, what is your background ?
I’ve spent best part of last 30 years in the content space. My initial foray into the sector was as Art Director at Tony Stone Images, which was the first company that Getty Images bought, so you could say I was at the forefront of a major transition for stock content, one that took a sleepy cottage industry to what it is now a multi $billion one.
– What does SnapRapid solve? How does it work ?
We use deep learning technology to track and analyze images and videos for sports rights holders and sponsors, putting a value on the exposure they get through digital channels – much as they would have valued and still do value their TV exposure. The technology that has been used for TV analysis is just about OK for that purpose, but when it comes to tracking and analyzing thousands of images and videos across multiple channels, something quicker, more accurate and involving minimal human input is required. For example, when analyzing a Hockey game, it’s around 60 minutes of controlled footage for TV that needs analyzing. When we track same through social, we firstly don’t know where it is going appear, so we have to find it, we know it’s going to crop up in numerous locations in multiple edits, but we don’t know where so we use the system to find it and then analyze it. There is no room for any manual process in this, as there is with tracking TV.
– Do you analyze anything besides logos ? For example, if it’s a woman wearing a t-shirt of if the photo was taken on a beach ?
We can and do analyze all sorts of objects and elements, the system can be trained to recognize anything, but our system was set up to track logos for analyzing sponsor value, which was a market that was not covered. We’re now taking it into all sorts of markets, product placement and to help brands understand the negative, as well as, positive footprint their brands leave, such as discarded polystyrene cups.
– How do you estimate the value of the exposure you find?
Before we came along, estimation was the name of the game in social, now we calculate. We analyze the reach by valuing only on engagement, so a view, tweet, retweet, like etc. We use media value per platform to calculate a value which we created over a nine month period with the help of some global media companies and sports rights holders. We believe our valuation methodology is much more accurate than that used to value TV exposure, which is far more likely be based on guesstimated figures.
– Who are your typical customers, brands or sports clubs ?
We work with rights holders, sports federations, brands, and sponsors. They all need to know the value of sponsorship, that is where it is important to have a valuation methodology that is accepted. It’s not possible to say to a rights holder, you have X value and their sponsor, it’s worth Y, they have both be happy with the figures that are being put in front of them.
– Can brands use/license the images they discover or is it purely statical?
What we primarily do is value exposure and then work with the rights holder or sponsor to better understand how they can maximize return going forward, so although it’s something we have yet do, it is something that will enter the mix going forward, I have no doubt. Can we do it? Sure.
– What interesting patterns have you seen emerge ?
I have to be careful about what I say, but one thing that has struck us is how unprepared for digital most brands and sponsors are. The sponsors, quite rightly, as they are paying a lot of money to the rights holders, want more granular data, especially around digital, but they are largely not getting it and have until recently accepted that. They are now getting more techie, when told it can’t be done, because they know it can.
Instagram is bringing a lot of value, which is growing very quickly. It’s a platform where engagement is incredibly high and shows no sign of slowing. UEFA Champions League and NBA playoffs have been the two standout tournaments for engagement, of all those that we have tracked. From what we have seen, the vest patches are going bring incredible exposure for their sponsors – interestingly, we have the only data on this through social media exposure, so any sales of sponsor patches without talking with us, is likely going lead to mis-selling of the patch.
The Champions League brings enormous value to its sponsors through social media, even though they still get the content removed as quickly as they can, to help protect their TV broadcasters – they work with YouTube, who help find the material and then they take it down, but not before we have found it and valued it. It’s a bit cat and mouse. We both chase the mice, when we get to them we value them, when YouTube find them, they eat them!
The amount of value that comes from unofficial channels versus official channels is remarkably high. It does depend on the event and the activation plans the rights holder have, but generally no more than 20% of value will come from official channels – 9% in case of Champions League final, for example.
– Does Snaprapid work outside of sports sponsorship ? Can brand use it to monitor social conversations for example ?
We do work outside of sport, we work with product placement, the film industry and the broader broadcast industry helping them get their content online – we have a tool that enables brands to get their live content to social media platforms, called Live2Social – track, value and analyze it.
We set the system up to value visual exposure, but at the same time we analyze word sentiment, so for us to track and monitor social conversations it’s part of what we do anyway. Just as TV sponsor valuation has been covered, listening to social has been to, but when, as we do, TV, digital and social media sponsor coverage can be analyzed through one platform – 3Play, as we call it – that is an enticing proposition. As with social listening, add in look and you have complete social coverage.
– Let’s talk technology. Obviously, content recognition is involved but what else did you need to develop ?
Our technology came out of technology that was used to help build driverless cars. When you think about it, what we do and what driverless cars do is much the same, they have to recognize road signs, objects, markings, it’s much the same as logos, brands and products.
– How many images does Snaprapid analyze a day and on what platforms ( Instagram, Tumblr, Pinterest, Snapchat, Flickr,?)
We look at millions of images and videos over any week, the volume depends on what we are tracking and the amount of engagement there is likely to be with the event. We have to spread our net wide to ensure we get to all the content, so on average only 4% of the content we look at has what we are tracking in it, that’s roughly the same for images and videos.
We analyze 11 social media platforms: Twitter; Facebook; Instagram; YouTube: Weibo; Vine; DailyMotion; Tumblr; Google+; Reddit and Flickr. We analyze millions of images and many thousands of videos per day, volume largely depends on the sponsor or product we are tracking and the event.
– What would you like to see SnapRapid offer that technology cannot yet deliver ?
There isn’t anything I can think of, we are close to having our face recognition system ready, that is the final piece in our technology solutions puzzle. We are testing it now, at the Olympics, hope to have it launched through September.
Photo by Jan S0L0
Author: Paul Melcher
Paul Melcher is the founder of Kaptur. He is an entrepreneur, advisor, and consultant with a rich background in visual tech, content licensing, business strategy, and technology with more than 20 years experience in developing world-renowned photo based companies with already two successful exits.