The Experience is the Brand

In Praise of Long Copy

It is fashionable, of late, to conclude that brief, punchy copy is best suited towards capturing the attention of people online, because as everyone knows, nobody reads anything online, ever.

This strikes me a conflation of a well observed and well documented browsing habit among most people, who tend to scan rather than read. And particularly in terms of lists of items (products, news headlines, categories, etc.), scanning is even more punctuated, limited to the first two words or so of each item.

Is it a contradiction, then, that some eyetracking studies have shown that the Ogilvy Layout still works? Or would it simply be a mistake to conclude that offline and offline behavior are so different that conclusions drawn from one medium simply couldn’t apply to another?

Now, in magazines, which are mostly read as a diversion, the first thing to get scanned are pictures.  We are visual creatures and pictures typically convey a lot of information (and emotion) fast, so a strong visual is almost always going to be the first thing the eye fixes on when the reader is engaging in general browsing for interest.  Please note, though, that this scanning order changes for task oriented individuals interacting with a website.  People scanning a web page redefine “worthwhile” by relevance to their task, and therefore focus on the headlines first.

- Jeff Sexton, Tests Indicate Ogilvy’s Old School Layout Still a Winner

Behavior changes with context, so it’s perfectly reasonable to assume that the context of an online experience is different enough that the “old school” rules don’t apply – users don’t read, they scan, for instance. How is it that scanning – apparently superficially – conveys enough information that users can choose intelligently among the maelstrom of options available to them? Are they so focused on their task that all distractions are ruthlessly cast aside in single minded pursuit of their goal?

This post started as the result of an item in my Twitter feed, led through three other related, linked articles (all quoted here) and will probably terminate in a half-finished thought. All the while, I should have been assembling an analytics dashboard for a client. Using oneself as a proxy for examining the behaviors of others is never a good idea, but I doubt that my experience in the last 30 minutes is drastically outside the mainstream.

We humans have remarkable brains, capable of processing scads of data coming from multiple inputs. But we are hard-wired to attend to novelty, and generally not quite so good at efficiently filtering out distractions. Why do 30-50% of DVR owners not skip the ads? Is it because so much of the advertising out there is so good, we just can’t bear to miss it? Or is it because the change of scenery is, itself, interesting?

I’m inclined to think – though I can’t yet prove it – that the tendency towards brevity, pithiness and punchy copy has not been successful in the aggregate. There may be specific advertisers who’ve benefited, but if the whole of the marketing universe has simply been sub-divided into smaller pieces and split between more voices, the overall quantity of message hasn’t improved (and I’m of the belief that more voices in a conversation tend to degrade its overall quality.)

Can a truly persuasive, high-quality argument (in the sense of a position one wishes to communicate and convince others of) be encapsulated in a 10-word snippet? In 155 characters? Can a customer be won for a lifetime of loyalty with a one-liner? A statement so inexplicably powerful that resistance is futile? (Wouldn’t that be quite an effective weapon?)

10 Reasons Why the Crowd is Not Your Friend

TwitScoop, tag clouds, crowd-sourcing, flash mobs, prediction markets… at least a couple of these things sound like they should be useful tools. But as some have pointed out, aggregated knowledge often turns out to be nonsense where the future is concerned.

1. A short-term trend is really just noise.
2. Crowds follow leaders, but:
3. The mere fact of leadership does not validate, by itself, a direction.
4. Popularity can often trump reason.
5. After a certain inflection point, trends can generate their own steam, amplifying a false signal.
6. Mobs are not, generally, known for tendency to consider unintended consequences.
7. Stripping a problem of its complexity, so that judgment can be rendered by thousands of people simultaneously, often produces an irrelevant answer.
8. Early results in prediction markets – when they would be most useful – fail to predict alternative, as yet unforeseen outcomes.
9. In order to get useful input from a large sample of market participants, the problem set has to be readily understandable (see #7.)
10. Self-awareness of their impact on the outcomes they are predicting should, eventually, muddy the predictive waters enough to make them useless.

If I were a betting man, I’d say less than 1/2 of these eventually turn out to be provably true.

To be equitable, I’ll try to come up with a few examples of how crowd-based decision making might actually be useful.

You should not follow this advice

Dustin Curtis recently did a little experiment on his blog, testing different versions of an imperative for his audience, imploring them in various ways to follow his him on twitter.

His basic premise seems to boil down to: tell people what to do, and they’ll do it.

The commenters to that post pointed out a number of problematic issues, namely:

• The non-linearity at which twitter followers grow (nothing succeeds like success) could mean that as his popularity was increasing anyway over the test period, the clickthrough rate on Dan’s site was driven by an increase in inbound clicks sourced from twitter. (Note: This isn’t true; regardless of the rate at which traffic grew, it would proportionally affect each of the variants, since they were served randomly.)
• The particular subset of the global audience which read Dan’s site would be more likely to follow him in the first place (so these results might not be generalizable)
• The phrasing is merely novel (so the wider its usage, the less effective it will be)
• It was the length of the phrase, not the phrasing itself, that made the difference.

And the responses ranged from “this is interesting, I’ll try it,” to “you’ve missed a lot of points,” to “don’t worry about why it works, just do it.”

One complicating factor that I don’t think has been adequately explored is whether the length of the wording might be assisting its noticability and readability; A variant of this line of thinking would suggest that Fitt’s Law was at work (essentially, longer links would be easier to target and click), but in each variant, the actual linked text was the same length.

The more generalized conclusion seems to be that forceful, command-oriented links might be a better way of leading people through an interface. Intuitively, this might make sense. It seems to fit within the confines of well established usability heuristics.

So, should you follow this advice? Not really.

First, you’ll have to consider your audience. If you are a highly skilled UI designer who custom-crafts the layout and design of each blog post in Photoshop, it is possible that your audience will appreciate you telling them what to do (after all, your expertise drew them in in the first place.)

If you are a business, and you are trying to gain a new customer, the hard sell might not be the way to go. Context counts.

You Don’t Always Have to Follow the Clicks

Commenting on a recent story about how the New York Times online editors manage their homepage’s content, Nicholas Carlson concludes:

But Web editors simply must pay attention to readers’ clicks for two obvious reasons.

• It’s the main way readers can show what kinds of stories they care about.
• The New York Times is a deeply-in-debt, for-profit enterprise that needs to grow its traffic online in order to survive. Web editors should not pretend that it doesn’t matter how many ad impressions the Times serves each day.

He can be forgiven for thinking this way, that competition for readership has changed every industry so drastically that no one – not even The New York Times – can afford to take their eyes off the minute-by-minute numbers and reacting to how their audience is behaving and adjust accordingly.

First, NYTimes.com isn’t, ever, going to make up for the $1 billion shortfall in the rest of the company. It’s hard enough to make a web property self-sufficient, even when it doesn’t have to bear the burden of supporting a huge contingent of staff reporters on its own.

Second, he’s assuming that popularity is a necessary component or measure of quality. It isn’t.

Note that Jim Roberts, who is NYTimes.com’s digital news editor, isn’t averse to knowing, “some broad trends” concerning readers’ interest, but the “minute to minute thing” isn’t that much of a factor.

That’s a perfectly reasonable set of governing rules to apply, especially when your job is to filter for quality. Though fans of crowdsourcing and decision markets will disagree, the crowd doesn’t always know what’s best: years of editorial experience is likely to provide for a much higher quality (and over the long term, successful) experience for readers.

Imagine the alternative: multivariate versions of NYTimes.com served randomly to different audiences, optimized in real time to promote those stories which were the most popular, the most clicked on, the most read.

We have plenty of websites that work that way, we don’t need the New York Times to do it. Crowds optimize for what is popular, not necessarily what is best.

Tommy Lee Jones said it best:

A person is smart. People are dumb, panicky dangerous animals…

Shocker of the Week: Mobile Usability is Hard

Aside from the more interesting data collection techniques used in the study, Jakob Nielsen has dropped a stunning bomb on the world of mobile web development:

Creating a useful and usable mobile experience is hard.

Of course, that’s only true if you are designing for the mobile experience, as opposed to redesigning a web experience for mobile. They’re different things.

Despite the prevalence of smartphones and iPhones and other mobile web devices, the reality is that, at least in the U.S., 85% of people still use a basic handset with a numeric keypad, and a 90 pixel-wide screen.

90 pixels wide: that is roughly the size of a large postage stamp. Those of you who have worked in the offline world, consider this comparison: you’ve been asked to take a full-page ad in USA Today and redesign it as a 6-line classified ad in Milwaukee Journal Sentinel. Or condense a 2 and a half hour film into 30 seconds.

These are not impossible tasks, but they are difficult. You’ve got to make some hard choices.

At The Archer Group, we’ve been hard at work on a couple of iPhone apps recently. Even on a richly interative, touch-screen platform with a screen as large as that, application development is largely a process of deciding what to leave out. Simplicity – both in terms of the UI but also at a deeper level of how information and functionality are structured, is essential.

On a postage-stamped sized screen, where input is limited to a 12-button keypad and semi-usable, at best, predictive-typing, there’s only one rule: if in doubt, leave it out.

If you find yourself overseeing the development of a mobile app, or the mobile view into your company’s existing online offering, consider this:

• The most popular feature among mobile phone users is voice. 100% of handsets can make phone calls.
• Assume that your audience is on a slow data network; graphics still take significant time to transmit, text is faster
• The mobile web is the Mathalon, not the Miss America Paegent; it doesn’t matter if you’re pretty, you have to be right.
• Of everything you offer your customers online, pick ONE THING to do on mobile. Do it extremely well.
  • After that, you can think about expanding your feature set.
  • Once you think about expanding your feature set, STOP, and DON’T DO IT.
• Go back and look at the one feature you built: make it work better and faster before you try to make it do more.

See, mobile usability isn’t that hard after all.

Twitter as platform

I’m admittedly late to the bandwagon. Now that I’m finally on twitter, I’m finding more and more uses for it everyday. Still, I have an aversion to publishing everything I’m doing, every minute of the day, and have so far mostly refrained from re-tweeting (“me too,” never adds anything to the conversation) or following others just because they’re following me (friendship isn’t something you accumulate, it’s something you nurture.)

Still, I’m impressed by the number of applications being developed on the Twitter Platform. By which I mean, applications which use Twitter as a mechanism for data transmission and manipulation, rather than an end unto itself:

• Collecting diary entries from participants in a usability study
  As fascinating as the study results are themselves, I’m more intrigued by the fact that twitter was used to facilitate the gathering of study data (albeit only an initial pass. End-of-day email surveys were used to augment the data.)
• Using twitter to self-monitor behavior
  Tweet any behavior (drive, eat, sleep, punch, meditate, etc.) and any quantity (25 miles, 6 lbs of spaghetti, 4 hours, 3 kittens, 20 minutes) and your.flowingdata.com will capture that information and chart it for you, and let you interact with that data so you can identify patterns (too many carbs + too little sleep = angry at cats) and change behaviors.

This is obviously the key to twitters future (if ever) profitability. The catch is that twitter succeeds as a platform because it’s free; any monetary component is going to have to do more than merely shift the burden of profitability to developers in order to succeed in the long run.

Website Hours: M-F; 9am-5pm

Silly me, I thought the web was an always-on medium. Here I’ve been, pretending all these years that the self-service nature of a website meant that it would be available when I was ready, on my terms.

Wasn’t that the promise way back in, oh, 1998: that I’d be able to buy dog food in my underwear at 3am on a Tuesday?

There are certainly many more pressing problems within our nation’s health care system than the unavailability of UnitedHealthcare’s website. But anyone who worries about “government rationing” of health care under a single-payer system should at least take a moment to consider this: here we have a private enterprise ostensibly in business to serve their customers, a business theoretically driven by some profit motive, a business which relies on cost-cutting efficiencies like a self-service website where members can check claims, balances, EOBs, etc.,and their website is down for scheduled maintenance every evening from 8pm to midnight, and most of every weekend.

Just to clarify: the website isn’t occasionally down because of a technical glitch, they purposefully and with some apparent forethought take every essential system offline in the evenings and on weekends for routine maintenance.

Imagine if ATMs didn’t work on evenings and weekends: what would be the bloody point of having ATMs in the first place? They’re automated, for heaven’s sake.

I know just enough about systems architecture and complex development to know this much: there is no legitimate reason for this to be so. I’ve worked on systems that pulled stock trading settlement data from fixed-length data files retrieved from a tape memory system over a 28k modem, and even those were down for maintenance less than 2 minutes a month.

I can surmise that UnitedHealthcare treats their customers this way because they know that most of their customers have no choice but to put up with it. I’m in an employer-sponsored health plan (and lucky to be so, at that), so I can’t really take my business elsewhere. And I’m not about to get on the phone at 10pm at night to talk to a customer service rep (not that their call center is open that late, regardless.)

Still, strategically it makes no sense to neuter a self-service system this way, and force your customers into a more costly service channel.

Getting my act together

I’ve made a number of, frankly lame, attempts over the past 9 years to find a good home for my blog. It’s not that I write that often, but I do make sporadic attempts to commit some thoughts to paper, as it were, and the pace of that seems to be accelarating.

There is nothing worse than knowing, somewhere in the back of your mind, that you had the kernel of an interesting idea to explore, and not being able to figure out where you left it.

While I troubleshoot some WordPress importing issues (most of the setup of which has been an absolute pleasure), the full 9 years of archives will be reposted here. It’s nothing to be excited about, there were multi-year gaps in between some posts. But for the sake of continuity, they should all be in one place.

I’d invite you to check back soon, but who does that kind of thing anymore?

A Million Little TV Spots

When I hear that the ad agencies and ad networks are starting to play nice together, I’m struck by the fact that the conversation keeps focusing on how large media clients are going to better utilize the new technologies afforded to them by microtargeting, mass personalization, and segmentation.

It’s as though, having shown up for a rugby match dressed for badminton, Publicis and WPP have decided to put on a brave face and hope for the best.

There has been, is now and will always be a place for broad campaigns aimed at driving brand awareness, and herding a large group of people towards satisfying the desires of a single, huge, business.

But there is a larger need for a mechanism that allows a hundred million small enterprises to market their goods, services and themselves to a similarly small and scattered audience – and audience measured in dozens or hundreds, not millions.

The big agencies continue to try and figure out how to play the “new” media by the old rules. They ask, “how can I build a viral campaign that spreads my message far and wide?” “How can I use social media to get people to talk about how much they love my brand?”

For decades, global business – brands and the companies that support them – have lurched towards ever greater conglomeration, going for scale and scope.

But the digital advertising landscape is built for a different kind of scope: it is engineered for point-to-point communication which, although usable in a large broadcast model, isn’t optimized for one-way message delivery. Why do brand managers and their agency partners fret over how to handle negative feedback online? Why do they try to fine-tune their placements so that video overlay ads show up only on content they deem “brand-appropriate?” Why do they struggle with content- and keyword-targeted campaigns, throwing armloads full of cash at a medium where the best results are achieved not with a broad-based “messaging” campaign, but with a hundred thousand individual conversations.

Why are WPP, Publicis and now Google and Microsoft working together to develop new ad formats for digital media?

What happens when Unilever’s ads for Dove soap get elbowed out of their “prime” advertising space by 1,000 artisans selling hand-made goat-milk soap, who join together in a loose social group and collectively purchase $20 million worth of in-video ads on YouTube?

Who’s going to knit that group together? Who’s going to provide them with the technology and creative backing to develop that campaign? Who’s going to place it, bid for the space, monitor and fine-tune performance?

Does the world inexorably move towards larger and larger enterprises selling more and more of the same stuff to everyone, or is there a point at which smaller becomes better?

This is going to be big: Virtual DVR gets the green light; Digital marketers to become the new “traditional”

To date, the convergence of digital media and television (a distinction which, post-6/19 is largely semantic) has been stymied by one fundamental difference between these media: time sequencing.

Despite TiVo and OnDemand and Slingbox and others, television is still largely – for most people – a time-bound medium. Even though many cable, sattelite and over-the-air viewers are able to time-shift their TV schedules, most television viewing is still conducting in “real time.”

Now, some programming simply won’t be time-shifted; sporting events and news, for example, lose their value over time. Can you imagine tuning in for the Superbowl three days after everyone else? Do you really want tomorrow’s weather report two days from now? (The exception might be for sports fanatics, who want to watch 15 football games from over the weekend but can’t actually do it all at once; so they time-shift some of the games into subsequent days of the week.)

“Water-cooler” shows are another exception: no point in watching Survivor after everyone has already talked about it the next day. (Actually, I don’t know that there’s ever a point in watching Survivor, but maybe I’m the exception.)

But with 500-channel cable and sattelite lineups routinely accessible, time-shifting becomes less of a luxury and more of a “necessity”, at least in terms of people maximizing their leisure time. With so many choices, it’s unlikely that everything you’d care to watch is on in precisely the right order on precisely those days of the week when you’d want to watch them. So you set your Digital Video Recorder to tape the shows you want, and they’re ready for you, when you’re ready.

There are a couple of problems here, that keep this from being practical for everyone:

1. You likely don’t have space on your DVR to record everything you’d want
2. You can’t record something in the past (say, the first three episodes of Big Love that you missed before all your friends were talking about it.)
3. You can’t record something you don’t already know about (this is a variation of number 2.)

There are other issues, too, in terms of the usability and practicality of some DVR’s interfaces and ability to put the show you want to watch where you want to watch it. (For instance, to play an episode of Jon & Kate Plus 8 minus their marraige on your iPhone while waiting to clear customs at Gatwick.) I’ll cover those later.

If I haven’t already bored you to tears, let me get to the point.

This week, the U.S. Supreme Court cleared the way for Cablevision to start offering Virtual DVRs. This is important for marketers because Cabelvision and other cable and wireless network operators can combine virtual DVRs with a couple of other technologies, and fundamentally change the landscape of entertainment.

With a virtual DVR, the recording resides not on a box on your entertainment system, but at the cable company. Your recording is shared with everyone else – so rather than store 100,000 copies of Most Dangerous Catch, Cablevision stores just one copy and anyone can start watching it whenever they want.

• This eliminates most of the storage capacity problems, not only by centralizing storage of programming, but also by eliminating 99% of the redundancy.
• If the recording resides at the cable company, it can be instantly delivered anywhere there is sufficient bandwidth – to your TV, your phone, or a flat screen in a hotel room.

So, you can watch anything you want, anywhere you are, anytime you like. Aside from being kind of neat, why does that matter?

It matters because all of the world’s entertainment content is now digitized and delivered on-demand. And for it to be delivered to you on demand, the cable company needs to know where to send it when you ask.

Which means that they are sending you a video stream, digitized, to an addressable device registered in your name. Sound familiar?

It’s called YouTube.

Now, if only there were a platform out there which could embed, in real-time, advertising that was targeting specifically to each viewer, either as part of that video stream or as a companion piece of data displayed along side of it.

That’s called AdWords.

Even though the first itteration of addressable tv commercials has been scrapped, a new launch date is set for the end of 2009. Still, the technology already exists to deploy this all pretty much right now. (And as much of a fan of Google’s tools as I am, they’re not the only player – there are other ad delivery platforms and video streaming services, and they just need to be linked up; Google just happens to be there already.)

This isn’t about watching YouTube on your 56 inch LCD screen in your living room, or having to endure text ads running alongside NBC Nightly News. When technologies converge, it’s rarely in the form of one technology simply expanding into another technology’s space. Rather, the best elements from two technologies are combined to create something new.

So let’s review:

• Virtual DVR brings the public an unlimited quantity of high quality programming that they currently want to watch (and don’t have to find), delivered digitally in a personal, and personalizable stream.
• Online advertising gives marketers a platform to develop, target, and deploy advertising in a highly personal way; content producers have a platform on which to sell ad space (either directly negotiated or in an auction model.)

For marketers who have spent a lot of time building large branding campaigns predicated on a broadcast model, I have some bad news: the world is going to get unmanageably complex, and that model is going to break. It may not happen right away, but it will happen.

For marketers who have been deciphering the complexities of micro-targeted online advertising, the landscape is about to open up.