We're over a month into Armageddon: Quarantine Edition, and some things have become clear. First, few people seem to know how far apart six feet is, so here's a hint: it's more than you think. Second, Saturday Night Live is the same amount of funny whether done in front of a studio audience or via a bunch of Zoom video chats. I'll leave it to the reader to decide exactly how funny that really is. And third, the amount of time it takes for a new word to be introduced, picked up by the general populace, and then overused to the point where it'd be preferable to scoop out one's eardrums rather than hear it again is about three months -- at least if my experience with the terms "coronavirus" and/or "COVID-19" is any indication. So while a few weeks ago I wrote about the notion of The Big Meme, and how when a single thought can quickly capture an enormous amound of mindshare , now I'd like to address the next phase of any such thing, namely deciding when it's time to let the Big Meme go.
The allure of using an already-propagated meme is easy to understand: with a few words it becomes possible to capture attention and inspire imagination easier than if you had to rely on your own creative assets and messages. It's why mascots and jingles dominated the first near-century of modern advertising. These days the study of meme propagation and how and why things go viral is a big deal, as a marketer that can make their messages consistently go viral might literally be worth their weight in gold. Research conclusions on the topic vary from "it's impossible" to "it's nearly impossible," with folks in the latter group mostly admitting that there just isn't much rhyme or reason to what makes one meme go viral while another doesn't. Approaches to studying memes vary to highly academic and math-heavy analyses of propagation and timing, like this meme study from Stanford, to the decidedly less formal (though perhaps more intense) approach taken by the armchair statisticians of the popular /Memeconomy community over at Reddit. And in fact, the best explanation of a message's "lifecycle" that I've come across comes from a Memeconomy Bogmire, who offered up this diagram:
In short, any sufficiently viral-worthy new content starts out unremarkable. Like watching a nuclear reaction in slow-motion, there comes a point where the message reaches critical mass. It instantly becomes fantastically energetic -- at least until its supply of "mental fuel" runs out, at which point it suffers a sudden, sharp drop in usage before either fading away into obscurity or showing up on ironic T-shirts at college campuses.
The world's talent, time and treasure are currently focused on slowing the spread of the novel coronavirus and limiting the pain and suffering of those who contract it, and while it's too early to say for sure, the current plans seem to be working. People are self-isolating, kids are going to school from home, and while my local supermarket still occasionally looks like something out of I am Legend, it's better than a few weeks ago when it looked like something out of The Purge. Indeed, Google's trend tools suggest that we may be flattening the "coronavirus" search curve as much as we're flattening the viral contaigen curve.
All this begs the question -- when will marketers stop using the coronavirus to hawk their wares?
While jumping on the Big Meme bandwagon is usually an easy decision (the upside almost always outweighs the downside), the same can't be said for leaving it. As the first chart above indicates, once past that moment of critical mass the meme will grow and expand while maintaining relevance. Cross the precipice, though, and there's a sharp drop into obscurity. While it's unlikely that the coronavirus and our time in self-quarantine will fade from memory anytime soon, marketers invoking it as a shared thought once it's past its prime will run the risk of tapping into our collective fatigue -- and its siblings, irritation, frustration and paranoia -- of all things COVID-19 related.
As for me, I'll be tapping out now.
I'd like to start out by saying that I fully understand the irony of this post, having just written it.
It's 2020, years since I last wrote a serious blog post for WireSpring, partly due to a business pivot but mostly due to the fact that the nature of business writing just started to feel... obsolete to me. When I first started blogging -- back before that was even a word -- things in the digital signage industry were new, fresh and exciting. Hell, things in the online business writing field were new, fresh and exciting as well. Clever writers developed content not just because it would be indexed by Google and used to provide some minuscule bump in search rankings for whatever terms it was optimized for, as is all to often the case today, but because they felt they could contribute knowledge that wasn't yet widely available or commonly known. Of course they took opportunities to try and establish thought leadership and brand recognition (and the bump in Google search rankings certainly didn't hurt), but the signal-to-noise ratio was pretty high: for every decent, educational post there was a manageable amount of garbage.
Fast forward to today. Social media offers vast echo chambers capable of turning a single modest (and frequently incorrect) thought into a barrage of memes, tweets, reposts and comments. More often than not, long-form content gets developed and then immediately redeveloped into slate of shorter episodes, each with their own clickbait-y title. Meanwhile, outsourcing groups in low-wage countries backfill with new content more focused on word count and search engine optimization than insight, further amplifying the noise. And I have no idea what's going to happen once the AIs get good enough at writing to pass for human.
None of this is news to anyone who visits websites or uses social media. The competition for your attention and personal data is more fierce than ever, even as cheap aggregation services and a never-ending stream of major security breaches makes our information more available and accessible than ever. Which leads me to the true subject of this post: personalized messaging versus the Big Meme™.
A long, long, long time ago I wrote a blog post about the "Uncanny Valley" of messaging -- when personalized marketing starts to feel creepy. It's still a fun (and relevant) read, partly because it contains this silly chart:
In the years since I wrote that, a strange thing happened. Marketing automation services enabled the cheap production of highly personalized messaging. Legit mail marketers and spammers alike no longer had to be limited to simple template variables, and could instead make up complex algorithms to fill whole paragraphs with highly specific text based on troves of personal information sold for pennies a person. And sure enough, these unnaturally crafted tomes did really start taking us toward the uncanny valley of text -- messages that seemed to know much too much about us, but clearly got something fundamentally wrong.
But the thing is, we never really crossed over to the other side of the valley. We may still in the future (see the aforementioned note/fear about AI-generated content), but for now we're firmly stuck on the left. Marketers, seeming to realize this (or, more likely, having realized that response rates started falling off, or that the effort of automating all of that customization just wasn't worth it), seem to have been stepping back from the brink and simplifying their targeting approaches. For various reasons conventional digital signage and OOH media in general never really achieved the same level of hyper-targeted customization, so there's been a shorter distance to fall back. That said, though, the content I've seen running around town certainly looks a bit simpler -- on average -- than it did a few years ago. Whether or not it's being generated by some super-sophisticated algorithm on the back-end I have no idea, but I kinda have my doubts about that.
So what does this have with COVID-19 and the "Great Hunkering Down of 2020?" Simply put, for the first time in... well, maybe ever, there's a thought or idea that virtually every person on Earth has an interest in, and that has allowed marketers to shift their content way over to the left side of that chart. The coronavirus is the biggest, baddest meme around, and it will be top-of-mind for virtually every consumer for not just the next month or two, but long after. Behaviors will change, shopping and travel patterns will upended, and people will have the opportunity to find new and (potentially) better ways to do things that they've been doing for their whole lives, whether it's going to school, going to work, or going to the store. Consequently, every single marketing message in your inbox, on your phone, or shown during your commercial break can safely toss in a hook they know you'll at least make some vague subconscious connection with. I saw two ads on the digital signs in my local pizza place -- one for a personal injury lawyer and one for a real estate broker -- that both managed to shove in a reference to the novel virus in the time it took me to quickly grab my pizza, throw some money on the counter, and run out the back door (I can only hold my breath for so long, after all). These services had absolutely nothing to do with healthcare, social distancing, or any other relevant topic. Yet because they latched on to this giant, global meme -- albeit clumsily -- they're positively going to catch more eyeballs, and I wouldn't be surprised if they converted better as well. We see similar effects on a smaller scale all the time. Superbowl Sunday is preceded by weeks of sales announcements and special deals. The furniture industry has mastered the art of leaping from holiday period to period in an attempt to build excitement around limited time offers that are only limited if you happen to be foolish enough to be shopping during the two or three days between them.
If the spam in my inbox is any judge, marketers are having a hard time figuring out the right ratio of deep customization to Big Meme. At the moment I'd say most are ditching highly targeted content in favor of heavy-handed references to coronavirus. In essence, Big Meme is winning. If that was effective it'd actually be good news for digital signage an OOH, since it's far easier to develop content for public display that doesn't need highly specific targeting to work. Except that nobody's supposed to be going out for the next month, which is pretty bad for an industry that's literally called "out of home." And the persuasive power of this big meme will eventually start to wane -- and there might even be some backlash -- as more people become literally and figuratively exhausted by it.
So that's my take on the status of personalization versus Big Meme marketing, having now witnessed maybe a hundred different messages aimed at the virus-fearing public. Tapping in to the collective social consciousness can make for powerful, memorable content. But timing is everything. If the message is everywhere, viewer fatigue will eventually cancel out any benefit that latching on to the meme would have offered.
Quick note - we've updated our digital signage price estimator for the first time since its release back in 2012. More accurate pricing, better volume discount estimates, and support for the latest-and-greatest tech are available to all who might want to generate a digital signage project estimate.
There are three big upgrades in this version of the price calculator: first, we added an option for 4k support. This basically just bumps up the cost of the screens and media players for those people who simply must have the newest, shiniest objects. Are there cases where using 4k content makes sense? Yeah, a couple. High-end video walls and retail installations might benefit from some really eye-popping ultra high def content, but the vast majority of the time it adds nothing. You remember this graphic, right?
Even in the best case -- and using a huge screen -- the visual advantage of 4k content basically disappears before the viewer is even 10 feet away. Buuut, people want to know how much it costs, so we stuck it in the calculator.
Next, we added support for speccing touch screens. We see a lot of projects these days that make use of a smaller touch screen and larger standard screen together, or else touch screens that are integrated into some larger device. Speaking of "larger" we also added pricing guesstimates for a larger class of screen -- 60-70 inch -- but honestly there's so much of a price difference between a 60" and 70" screen right now that our price estimate for that not going to be very accurate.
Finally, we considerably improved the way volume discounts are calculated, though honestly no online estimator is really going to do a good job estimating both a 10-screen pilot and a 1,000-screen rollout. But at least now we're making the attempt!
I'd love one day to be able to put a content estimator-guideline-thing up to give people a better idea how how the initial capex of installing a network pales in comparison to feeding it content forever, but that's a tall order and perhaps of limited value, since most people who are serious about their projects will so some internal calculations and everyone else... well... they probably won't be interested in a lot of math no matter who's doing it.
At one point we also considered something along the lines of an organization estimator which would attempt to quantify the number of people needed to manage the network, create the content, etc. (based on the estimates from Digital signage staffing analysis), but honestly that's hopelessly complicated and would be so extremely error-prone.
There are also some additional fixes that we need to make, in particular the handling of very low-cost Android-based players, HDMI sticks and the like. While many commercial-grade systems are still in the approximate ballpark of a lower-end Intel-based player, there is definitely still a good deal of downward pricing pressure, and we haven't fully captured all of that, so it goes to the top of the to-do list.
More than anything we'd like some feedback on whether this new and improved estimator is useful, and whether the numbers it's spitting out mirror people's experiences in the real world. Let me know what you think!
It's been a while since I wrote about making effective digital signage content, and in fact these days I don't get the opportunity to focus in on content strategy with clients nearly as often as I'd like. And that's a shame, since even after a decade of considerable growth (I think it's fair to say that digital signage in nearly all forms is solidly in the "mature" section of the technology catalog) I frequently come across signs in serious need of improvement. In just the past few weeks I've noticed a half dozen offenders, ranging from QSR menu boards to corporate lobby screens. But the real catalyst for this post came from research house eMarketer, who noted that even in 2016 lowly email marketing outperformed the more fashionable social media and paid search industries. Perhaps most interestingly of all, the humble one-size-fits-all email newsletter was still found to be the workhorse of the industry, offering a greater return per dollar than far more elaborate hypertargeted mailings. In light of the fact that marketing automation platforms these days will slice and dice content into dozens or hundreds of different forms in an attempt to best fit each recipient, that's pretty telling, especially as it flies in the face of the conventional wisdom in the digital signage arena.
The drive toward automating content production on digital signage networks is nothing new. I remember listening to Jeff Porter (a longtime VP at Scala) give a presentation on the subject well over a decade ago, and since then the barrier to entry for auto-generated content has fallen shortly. For large networks the need to automatically customize content playback can be particularly important, since it quickly becomes unwieldy (and expensive) to create, catalog, organize and deploy multiple variations of even a simple piece of content. Tech companies were quick to answer the call, and today there are hundreds of solutions on the market that will ingest your data inputs and spit out as many different output variations as you care to make. This kind of wrote work is exactly the type that benefits from automation, after all.
I continue to read about new tech developments in face recognition, beaconing, etc. that put a lot of stock in being able to show a highly-targeted spot to a specific person as they make their way about in the real world. At one point in time, this seemed like a pretty good idea to me. Now, though, I'm in the same boat as the email marketers mentioned earlier on. If I had to, I'd bet that quality content that follows many of our best practices for digital signage content and has high production value will work better (and be more cost effective) than content that bends over backwards to hypertarget to a very small group of people (or even just an individual). I don't expect that we'll ever see a well-controlled test of this theory given how ridiculously expensive it would be (and paradoxically how spendthrift networks can be when it comes to making content), but if the companies like P&G have difficulty getting hypertargeting to work online, I have to imagine that the vast majority of brands advertising on digital signage networks will suffer the same fate in their much less forgiving real-world settings.