Advertising performance: What do the numbers really say?
In our article back in October about hedging traditional media buys with ad placements on digital signs, I asked whether a spot on a digital signage network will "do" more than a spot running on TV, and if so, whether that additional work would justify more experimentation by brands and their media buyers. Mary Anne Fleisher (from aim digital visions, if the link is anything to go by) noted that:
[The] economy today has set our revenue back a bit. But it has really set back revenue for radio, tv and newspaper. For the first time in history radio revenue is going backward. Layoffs are huge and their big ego's are shrinking. We are doing a great job taking a share and as you pointed out even a 10% share can be big even with local advertising. Its a hard way to go the revenue generating digital signage, but testing the results is very exciting. The best thing going for us is 80% retention.
Image credit: ArtemFinland
I have yet to find few really good case studies that we can show our advertisers the merits of our medium/industry. You mentioned that from your own experience you think the return is 7-8% with sometimes even 300-400% - could you provide some details on these cases?As I noted in my response, a big part of our problem is that ad performance is very inconsistent, partly because the networks themselves are inconsistent, and partly because much of the content being produced today is still pretty bad. (TV advertisers have had 50 years to perfect their skills. We haven't... yet.)
The 400% example I cited in the article comes from an old client of ours, Bass Pro Shops. Bass Pro is a chain of sporting goods stores with a very loyal customer base. In the example I mentioned, they were advertising a type of fishing line, which was on the shelf with approximately 30 other similarly-packaged items. When advertised on the in-store network, sales of the product jumped 400% versus other brands of fishing line, none of which were advertised on the network. The ad ran in 22 stores, and was not featured in another 25, yielding a statistically significant result.
Was it typical? No, of course not. Most ads in that chain had nowhere near that kind of impact. But it does show what's possible.
In-store privacy: Are consumers really at risk, or am I just whining?
I got a pretty strong reaction to our recent "Uncanny Valley" article, which talked about the dangers of trading the privacy of digital signage viewer data in exchange for added convenience. While most of the folks who either commented or emailed expressed a view similar to my own (that a tight policy with strictly-enforced opt-in and opt-out provisions could yield a working system for trading privacy for personalized service), one commenter, only known as Anonymous, clearly felt I was off the mark:
Bill, I disagree with your chart and your theory behind consumer squeamishness.I don't even know where to start with this one. Obviously everyone's entitled to their opinion. That includes our anonymous commenter. But it also includes me. The chart I made reflects something close to my own opinion on the privacy-personalization exchange, so of course it's going to be different for others. My main point, though, is that I think everyone ought to be able to decide this for him or herself. What we have right now is a wild west-type scenario where people are gathering data, often without consumer consent or even knowledge, and are then using that data in ways that might not be in the consumer's best interests. I agree wholeheartedly with the anonymous commenter's last point -- we do need to establish a set of standards and stick to them. I know it's something that's being worked on by at least one major industry group, so I'm hopeful we'll hear something about it soon.
First, your chart only places face/iris identification below the X axis, without explaining why the other applications you list don't belong there as well.
In fact, the application in which the kiosk is activated by RFID in a consumer's loyalty card is actually a much greater invasion of privacy than iris/facial recognition.
The loyalty card not only identifies individual consumers, but links this identification with their shopping histories.
This application belongs below the X axis of your chart even if it has yet to gain as much media attention as facial recognition.
The other loyalty card application is less privacy invasive because it suggests tea and crumpet information is analyzed in aggregate, not at the individual level; so this may not belong in the valley.
Second, I think consumers' distaste for facial recognition in digital signage has very little to do with the existential paradox of human-like robots; that theory trivializes some very concrete privacy issues.
On one level, it's more simple than that: people instinctively don't like being watched or scrutinized without their consent and especially without their notice.
Digital signage with identification technologies represent a new front in mass surveillance; in this case the surveillance is used for marketing and not security, which makes the privacy encroachment more offensive because it is surveillance for profit and not for safety (and consumers are already bombarded with ads to begin with).
It's naive to think that digital signage will not evolve to routinely identify individuals, because it will be profitable to do so once the technology is less costly.
Similarly, it's against the trend of history to believe that the data digital signage firms collect on individual consumers will never be shared with other parties or used for purposes other than marketing; law enforcement is one good example: remember that any records kept by a digital signage firm are available via subpoena or court order.
Digital signage companies, trade associations, their partners and their affiliates must commit to consumer anonymity when using facial recognition cameras, and notify consumers of when such cameras are in use; RFID and mobile applications should operate strictly on an opt-in basis.
It is past time for the industry to establish concrete consumer privacy standards, both technical and policy-based.
Scrolling tickers: How should you display two tickers at once?
That's the question asked by Jessica, on our article about using motion and silhouettes to improve your digital signage content. My simple answer to "how to use two tickers?" Don't do it. Seriously, unless you have a very unique situation, two tickers is almost certainly a terrible idea. In fact, unless you really know what you're doing, even one ticker often doesn't make sense. It all has to do with how we absorb information from the screen, and while this will be the subject of an upcoming blog article (and I went over it during my presentation at the DSE last month), suffice it to say that using a ticker is one of the least efficient ways to transmit information to your viewing audience. Heck, even CNN figured that out and abandoned their ticker late last year.
So that wraps up our first collection of reader questions and comments. As you can see, there are some pretty interesting conversations taking place throughout these five years' worth of articles. And sometimes (OK, often) the commenters prove more insightful and perceptive than me. So the next time you come across something I've said and you want to call me out on it, please be my guest. My loss will certainly be the community's gain -- but if you want to agree with me, that's OK too :)
Did you enjoy the format of this article? Leave a comment and let us know.