How Do You Fix A Broken Record?

I recently gave a speech in Amsterdam at Radiodays Europe, a significant gathering of broadcasters from throughout the Continent (and indeed, the world) and a fine conference. One of the topics I discussed was how radio, until very recently the top source for music discovery, was gradually losing that image to YouTube, Spotify, and other Internet-delivered services. Of particular concern to the radio industry is this graph, showing the top choice for keeping up-to-date with music for 12-24 year-olds:

My Edison Research colleague Sean Ross tackled this issue bravely and well from radio's perspective here, but I want to take a bit of a broader view--one that is relevant to anyone in media, no matter the channel.

The question I was most often asked in Amsterdam was this: "How does radio get the music discovery image back?" I submit that this may be the wrong question. Consider the following recent developments:

• Garth Brooks has his own channel on SiriusXM, on which he admittedly does not play any new music (yet.) Beyoncé, Kanye, and Rihanna have all released new albums on Tidal--and nowhere else. Artists are "balkanizing" like never before.
• The rapper Future's new album, HNDRXX, just debuted at number one on the Billboard charts. The album he replaced was the self-titled album from...Future. He replaced his own album at number one. This has never happened in the history of Billboard. I've listened to both albums. They are good. But it wasn't exactly Rumours replacing Hotel California in 1977. Just sayin'. 
• The top 16 songs on Spotify's UK and Ireland charts are the 16 songs from Ed Sheeran's new album. Those 16 songs are also all in Spotify's global top 50. Every song. Again--good album, but not Thriller-good, surely.

The fact that Future replaced himself begs the question: was there no other good music between Future releases? And how can Ed Sheeran take up the entire British Top Ten on Spotify? The answer is partially that both artists have significant followings, of course, but also that nature abhors a vacuum. And that vacuum, my friends, is the lack of a variety of new hits. The problem isn't with us. It's with an ecosystem that can produce these increasingly bizarre phenomena.

I was quoted a few years ago in Charles Duhigg's recent book, The Power of Habit, about how radio "builds" hits, protecting brand new music (which nobody loves immediately) by surrounding it with familiar hits and promoting it with enthusiastic DJs. Today, music listeners are exposed to fewer unfamiliar songs than ever before, thanks to on-demand streaming, searches on YouTube, and forgoing albums in favor of 99-cent singles on iTunes. 

Here's the thing: that's what consumers want. The genie doesn't go back into the bottle. So lamenting the continuing erosion of radio's music discovery image by blaming listeners or "the system" isn't productive. The market is speaking, loudly, and it is choosing familiarity over serendipity. And as a result, music discovery as we have known it is broken.

So, to return to Amsterdam, the question for radio executives isn't "How do we regain our music discovery image?" Rather, it's "How do we fix music discovery?" How can radio innovate, and disrupt the disruptors?

My good friend and podcasting partner Mark Schaefer is fond of telling the story of the Rand McNally corporation, known for years as a leader in printed maps and road atlases. It's not a brand we talk about anymore, because we don't use road atlases anymore, thanks to modern GPS navigation. Rand McNally is still around, but their failure to develop a strategy to deal with digital upstarts like Mapquest and TomTom led to a Chapter 11 filing in 2003, and they lost their leadership position in navigation.

I'm sure that in the boardrooms of Rand McNally there were dozens of heated strategy sessions focused on regaining that position. But often, the solution to being knocked off the top rung of a ladder isn't always to keep scrambling to regain your rung. It's to build a better ladder. While taking a canal cruise through Amsterdam last week I happened to see TomTom's HQ building (they are a Dutch company) and I was struck by the same notion Mark was in his Rand McNally story. Just as TomTom knocked Rand McNally off the navigation ladder, TomTom has itself been knocked down a rung, by Waze and Google Maps. Waze was developed not to be a better map, or even a better navigation tool. It disrupted both spaces by focusing on getting people from point A to point B faster. Even if that means winding through housing developments instead of taking the Interstate.

So, if your brand loses an image, the key here is not always to fight to get it back. It's to solve a better problem. It's to imagine what will eventually kill the other brand that took it away from you, and become that thing, whether or not it has anything to do with the image you lost. The "serendipitous discovery problem" threatens to disrupt much more than the radio industry. That's the problem to solve.

But You Knew This Already

I am thoroughly enjoying The Undoing Project, by Michael Lewis, which is a story about the friendship between Daniel Kahneman and Amos Tversky, who essentially invented Behavioral Economics and whose thinking has informed everything I have ever done as a market researcher.

There was one particularly striking passage in the book that talks about the "Hindsight Bias," which refers to our innate tendency to order past events to form a kind of story, specifically one that "all makes sense" to predict a current event. For example, I learned in High School that World War One was started by the assassination of Archduke Ferdinand, but that is a neatly-arranged set of historical Legos that was put together well after the fact, because it made sense--and we all like to make sense. But the truth is, no one at the time the Archduke shuffled off 'is mortal coil would have confidently predicted The Battle of The Somme, Verdun, or Gallipoli.

I also think about this every time I read about a stock going up or down "on news of..." some event that ostensibly would affect the price. We know the stock went up or down. We know that event 'X' happened. But the linkage between the two? Our brain is wired to put those blocks in an order to craft a story that makes sense. Except the stock market is a random walk--consider all of the pundits who claimed the market would tank upon the election of Donald Trump. You'll note the following, which appears to be a depiction of a market not tanking:

There is a particular application of this bias that I, and every market researcher on earth, has heard scores of times in their career after presenting the results of a survey: "We knew that already--this doesn't tell us anything new." In my more intemperate youth, I might have enquired, "if you knew it already, why did you want me to ask it?" Today, I am more likely to point out that no one knew the thing. The thing was suspected--and there is value in confirming the thing. Often, it is uncertainty that prevents a company from allocating sufficient resources to pursuing a strategy. But knowing the thing--especially if your competitors can only suspect the thing--has enormous value.

The truth is, I very often hear "but we already knew that" from entities that, had they in fact known it beforehand, would have acted differently. We are actually crap at predictions, which is why I generally avoid them. We are, however, experts in arranging a narrative, in hindsight, that makes us seem like Nostradamus on coke. And, just as in poker, we don't remember the "bad beats." We remember the times that we nailed it, and then subsequently overestimate our prescience.

OK. I, it turns out, am also a human, and so I will stop beating humans up for this. So here is the action step: know that we all suffer from hindsight bias. So as you encounter a new piece of information in your business, resist the urge to make it fit into the puzzle you already have--to construct, in your mind, how this new fact was inevitable because of past events A, B, and C. Facts are messy, and they don't always fit together into a smooth narrative, the way our brain would like them to. Examine the information on its own merits. Roll it around, independent of what else you know. What else could explains this new fact? Under what other circumstances could this fact exist, or even matter? And, if you can confirm this fact (even if you previously suspected it), what competitive value can you extract from it?

Tversky himself said it best, as quoted by Michael Lewis: "He who sees the past as surprise-free is bound to have a future full of surprises." 

They won't all be pleasant. But you knew that already.

What I Learned From Four Of The Top Ten Posts on Convince and Convert

Jay Baer recently published the 10 posts that earned the most page views on his extremely popular blog, Convince and Convert. I was extremely honored and moderately surprised to have contributed to the two top posts from 2016:  The Most Important Podcast Statistics of 2016, and 5 Snapchat Statistics That Prove Its Power. I will also note that there were two other posts in his Top Ten that focused on statistics and metrics. 

I learned two valuable lessons from Jay's list, and specifically the fact that four posts on stats and metrics made it. The first is not very helpful to you, and that is for me to send Jay all of our data as often as I can (duh.)

The second, however, is hopefully less self-serving and a more exportable insight--and that is to think about why posts about statistics are such "catnip," as Jay describes them. The Scottish writer Andrew Lang once noted that Politicians use statistics "as a drunkard uses lamp posts--for support rather than illumination." Of course, in my field, we rely on them for illumination, and nothing gives me more satisfaction in my career than to have contributed new facts to the world that did not exist prior to our asking the questions.

But, let's consider the unjustly-maligned drunkard in Lang's quote. Yes, it is true that many times people look for statistics that prove or support the opinion they already held--indeed, it's a cognitive bias that we often can't overcome.  Now, I would never counsel using research to prove yourself right--if that is your goal, you always will. The true gift you can give yourself as a marketer is to always seek to prove yourself wrong.  But putting motive aside, why were so many people looking for statistics on podcasting, Snapchat, and the myriad other things we and others have researched over the past year?

I suspect these articles were used more for support than for illumination, but I don't think that is a bad thing, really--especially as we are about to enter a period of extreme upheaval in the discipline of marketing. There is a lot of disruption ahead in 2017, as the chasm between the haves and the have-nots in digital and content marketing widens, and as we are forced to learn what really works today when the things that used to work start breaking. Digital ad spending is up 20% over last year, for example, but if you look closer, that increase is mostly all going to Facebook and Google--and largely because they have built a strategic moat around their targeting capabilities that other vehicles are struggling to match. The incredible rise in subscription media, podcasts, and other ad-light media means that reaching an addressable audience is tougher than ever. And the lifespan of the average CMO is not increasing, that's for sure.

In short, to quote noted marketing expert, Big Daddy Kane, pimpin' ain't easy--and it's getting tougher. So maybe those people using the lamp post of statistics for support are doing it because, frankly, they need something to hang on to. I don't know what is going to happen to the global economy in 2017, but that is kinda the point--this is a period of great uncertainty, and any lamp posts and life rings we can throw people are likely to be appreciated because they are needed.

In uncertain times, a flight to safety often ensues. That's why people buy bonds in a bad economy--we have an almost animal-like instinct to seek shelter in a storm. For many marketers, that storm may be coming very soon. So, while there will always be an audience for content on cutting edge strategies, the newest tactics, or experimental techniques that might provide a competitive advantage, no one can build something lasting on a weak foundation. 

So, my unsolicited advice to you, dear marketers, is to invest your time and treasure in helping your prospective clients reduce their uncertainty. The answer is not your product (what a boring conversation that is!)--the answer is to find out what they are most nervous about (the knowable and unknowable unknowns, to quote the great American poet Donald Rumsfeld) and to offer them the certainty of a lamp post by doing what you can to help them know.

I'm often shocked by how little some companies really know about their target customers. I bet that is true of many of the companies to whom you are marketing your goods and services (and possibly true for you, as well.) Consider what piece(s) of information would, if known, reduce their uncertainty. When companies feel the pressure of an unstable economic environment, they don't increase their spending, that's for sure. So in a disruptive environment, it isn't enough just to market your stuff--especially if there is a flight to safety occurring in your industry. You need to find the levers to create the mindspace for your prospective buyers to even consider something new.

Chris Brogan once asked me what I sell. The census code answer is "market research," while the highfalutin answer is "Decision Support." But ultimately what I told him was this: I sell comfort. As you craft the bits and baubles of your marketing plan for 2017, think about the ways in which you can also provide comfort to your prospects, as they bob and weave through uncertain times like drunkards trying to get home from the pub. Give them ways to remove that uncertainty by helping them know some of their unknowns.

Build a few lamp posts.

On Fake News

There has been a lot of news on...news, lately. Specifically, fake news. Facebook is tackling it. Facebook's employees are tackling it. Google is piling on. All because so many were surprised by the results of the U.S. Presidential Election, and people are looking for something--anything--to explain how they could have been so blindsided.

I'm not going to claim that fake news had no impact on the election. I am sure there is some marginal correlation to be made. But the belief that fake news had a material impact on the outcome is simply not borne out by decades of election research. 

I spent Election Day 2016 the same way I've spent every U.S. election since 2004--looking at precinct-level data from both our exit polling and the actual vote count. And here is what I can tell you--our "50/50" country is not really evenly divided. It's the average of precinct-after-precinct of homogeneous neighborhoods that mainly vote the same way. I saw precincts that went 91% for Trump, and some that went 93% for Clinton. And this election is no different than the last several have been in that regard. In 2012, for example, there were 9 precincts in Cuyahoga County, Ohio (Cleveland) that cast zero votes for Romney. And more than fifty in Philadelphia County, Pennsylvania, that did the same!

We are surrounded by people who voted the same way that we did--not just in our Facebook feed, but at Whole Foods, and our local Cracker Barrel. That's the cold hard truth of the American electorate. That 50/50 coin flip finish is the average of thousands of precincts that aren't close at all. When you gather in social settings with friends, it is very, very likely that you aren't speaking with people who voted differently than you did, feel differently about the issues than you, or shared news items on their Facebook feed that fairly represented the opposing view.

In short, we might have shared a lot of fake news, but we shared it in support of an opinion we already strongly held, to an audience of friends who also shared that opinion. "Fake News" was less about swaying hearts and minds than it was about providing post-decision rationalization to hearts and minds already swayed.

I think you know this to be true.

The issue is not how much fake news was shared. The issue is how readily we wanted to believe that news, and why. Facebook can't fix that. Only you can.

The Expert's Guide to Content Marketing with Research

One aspect of my day job that is booming right now is conducting research studies to help companies gain exposure through thought leadership, like this wonderful, two-year collaboration we've had with Oracle on Holiday Shopping. It's something we know a little bit about, since that's exactly how Edison itself has built its own business over the last 20+ years. We've also been the sole providers of U.S. Election Exit Polling for the past twelve years, which is the largest research project for content marketing in the world, among other things.

Having your company or brand attached to a significant piece of research that actually reveals new, useful information is one of the best ways to show your prospects that you are not just trying to sell your stuff, but also trying to contribute to the field. These days, however, it can be a bit tricky to get your studies and research findings out there, since the Internets are lousy with quick stats, infographics, and other ephemera. When everyone is doing the same thing, and no one is standing out, there is only one thing to do--you need to do what others will not do. Here are five ways to do just that.

1. Answer the Question on Everyone's Lips

Often, the best way to determine what your brand could study is simply to listen to your customers, or your competitors' customers, to see the most common types of inquiries. What you are looking for are what the poet Rumsfeld might call "Knowable Unknowns": things we do not know, but are knowable through proper study. Last year, for instance, Netbase contacted us to get to the bottom of the true impact of social media on fashion buying decisions. Anyone with web analytics facility can tell you that a link or impression led to an online purchase, but what the vertical they were trying to serve really wondered was this: how much does what your friends say on Facebook or show on Instagram affect your decisions when you buy something offline? (We love researching the things you can't click.)

This was a "knowable unknown," and led to a very successful series of white papers and mainstream trade press placements for Netbase, and taught me a little bit more about shoes than I wanted to know.

Questions like this are everywhere, by the way. On a past episode of The Beancast (Bob Knorpp's wonderful and well-produced marketing podcast) we discussed the fact that most people who view YouTube ads skip the ads just as soon as they can, five seconds in. Clickstream analytics can tell you that a potentially appallingly low number of ads are watched in their entirety, but they can't tell you want the impact of those five seconds are in terms of branding. Good question, eh?

2. Answer the Hard Questions

For the past few years, Edison has put out a study that not only met criterion #1 in spades, but also had an added bonus: it wasn't an easy question to answer! We have put out research on the online radio space for years, but one question kept popping up from media buyers, agencies, and investment analysts alike: how much of the total time spent with audio goes to online radio, compared to terrestrial? How much time is spent listening to "owned music" (your own CDs or MP3 files) versus podcasts, or satellite radio?

We were asked this question enough to know that it was worth finding the answer. Turns out, it wasn't an easy question to answer, but it was, in fact, a "knowable unknown." We saw the fact that it was a hard question as our opportunity to answer it--again, to do what others will not, in order to cut through the din. So we answered it, in what has become a very successful research series for us called Share of Ear® (covered here in Billboard.)

3. Know How to Reach your Audience

For us, getting a gazillion hits on Buzzfeed is gratifying, but ultimately does not put food on our table, in terms of our specific prospects. For example, part of our business is election research, so it is far more important for us to produce the kind of quality work that will get covered by Huffington Post's Pollster column or Politico than to get hundreds of irrelevant placements. And we love getting picked up by the print editions of things like The New York Times, or The Wall Street Journal. Not only do our prospective clients read those publications, they also tend to respect and acknowledge the level of scrutiny those resources place on research they cover.

So, when we are working with clients to produce research studies that will motivate action with a desired target, we learn as much about that target as we possibly can (yes, by doing some research--we drink our own champagne here at Edison). This helps us to work with a brand's PR agency or their internal communications department to ensure that our research isn't just "interesting" (the damnable faint praise of the Internet) but useful and well-targeted to the kinds of placements that matter.

4. Don't be Known. Be Known for Something.

Here is a little secret: getting your research-based content marketing shared a million times or featured on Mashable and Techcrunch could be wonderful on the surface, but devastating to your brand if you didn't do the work right. The initial venues that post your work will do so if it sounds interesting, or has a slick, well-designed infographic, or addresses a hot topic. This will put your pie charts in front of a lot of eyeballs, and your vanity metrics will soar. This is good, if you get paid with eyeballs. If, however, you get paid with cash, there is a very real danger here: if your work won't hold up to scrutiny, was shoddily conducted, mis-reported or otherwise will not stand the test of time or numeracy, all you have done is get famous amongst those who will not buy, and infamous to those who might.

I see this all the time--a prospect will reach out to me asking me if I have seen (famously shared study X), and then proceed to tell me how awful it is. Essentially, you've had a grand opening for your store, but left the shelves dirty and bare. Doing your study quickly and cheaply is the best way to maximize your eyeballs to dollars ratio. But doing it right is the best way to be known for doing things right, and for the quiet minority--the brand managers, the B2B customers, the Agencies--to recognize that your company cares about actually advancing the field. People notice.

5. Illustrate your Findings Simply, Clearly and Accurately

This is as much as I will say about actually presenting and illustrating your data: just be clear. We don't use 3D graphs, pie charts with dozens of slices, or tiny-fonted footnotes about the sample. We want our audience (and your audience) to see exactly what the point is of even displaying the data in a graph in the first place. Ultimately, we want our data to live, both on and off the screen. What we have found, over and over, is that if we do the work right (#4, again) and present it in the clearest, simplest way possible, our data gets more uptake to reputable sources and curators of data. Period.

For us, this means we don't have to have the shiniest graphs, or the most vividly illustrated infographics. If we have done our job right, it is the data, and its usefulness, that will live on. Indeed, it's the only thing that truly does. We LOVE it when reputable sources for research like Statista (below) and eMarketer not only cover our data, but regraph it (with attribution, of course :) ) Again, that gets it not only shared more widely, but also lends more credence to the work. If a sharp, clear graph like the one below from Statista can't be drawn from your data, you're doing it wrong.

Final Thoughts

There has been a lot of talk about the flood of content that overwhelms us (my friend Mark Schaefer calls it content shock) and there is no question that it is harder than ever to stand out with research studies or any other kind of content. Certainly one way to do so is to establish genuine expertise, and to be known for not just quantity, but unimpeachable quality. When 80% of the players at the poker table are similarly skilled, the only way to win even marginal gains is to do what others will not. In the case of fielding and publishing research data for the purposes of content marketing, we've been holding to the last full measure of research devotion for two decades now. Take these five principles to heart, do the work, and you'll do more than create content, you'll create value.

(This is an updated version of a post I first published on the Edison Research blog.)