Every so often I like to take a bit of time and look at some data about how people interact with the web. There are a ton of opportunities to study user behavior online.

Pomplamoose: covers vs. originals

After looking through a few of the videos from Pomplamoose on YouTube, I began to notice a few patterns. The most striking trend was that the duo’s covers of famous popular songs – by far their videos with the highest number of views – tended to have a much higher ratio of dislikes to likes than their original songs. I was curious whether this was due to the videos’ popularity or whether it was simply because users liked their original songs more than the covers.

I grabbed the statistics for Pomplamoose’s videos from YouTube, plotted them, and made a couple of interesting observations.

Pomplamoose’s three most popular videos are all cover versions of popular songs (“Single Ladies” by Beyoncé, “Telephone” by Lady Gaga, and “Beat It” by Michael Jackson). Those videos also happen to be three of the four least-favorably rated videos by viewers (the fourth another cover of a popular tune, “I Don’t Wanna Miss a Thing“). The four videos range between 88.6 to 91.5 percent of raters that chose “like” as opposed to “dislike.” No other video by Pomplamoose rates under 95.4%.

In short, Pomplamoose’s four videos that have a less-than-95% favorability rating are all well-known covers, and none of their three videos that have crested three million views exceed the 95% mark.

It’s important to note, though, that Pomplamoose’s cover videos don’t always attract the lowest ratings on their channel. Their most-liked video happens to be a cover of Mark Owen’s “Makin’ Out.”

Another view: Julia Nunes

Pomplamoose is a bit of a unique case. They often produce cover versions that vary quite a bit from the original version of the song, and that difference may attract lower ratings from fans of the original artist and arrangement. To provide a different view of the covers versus originals question, I pulled in some data from well-known YouTuber Julia Nunes.

Julia’s videos are all over the map. Her two most-watched cover videos – “Build Me Up Buttercup” and the Counting Crows’ “Accidentally in Love” – are actually rated quite high, both around 97%. Her two lowest-rated videos are two original songs, both of which have received over a million and a half views.

Final thoughts

Using Pomplamoose and Julia Nunes as examples, it seems as though it’s difficult (but possible) to make videos that are both extremely popular and near-universally liked. In both cases the most popular videos also have relatively low view counts. There doesn’t seem to be a clear association between covers, originals, and liking, but the majority of original songs tend to be less-watched but more-liked. That may speak to the value of devoted fans that watch a greater variety of videos on a channel as opposed to people who stop to watch a cover of a song they already know well.

Before making any generalizations about popularity and favorability, remember that this is all data from YouTube. Julia Nunes’s most-liked video is a clip of her singing as the Cat in the Hat in a school play. It’s a weird crowd on the web.

On Tuesday the editor of three top Maine newspapers (the Portland Press Herald, the Kennebec Journal, and the Waterville Sentinel) announced that they’d no longer include the ability to comment on news stories due to the fact that comments often ranged from “insensitive” to “vicious.” The decision was heralded by many readers who disapproved of the content of comments, but some criticized the move and suggested that it’s valuable for media outlets to collect reactions to a story.

Let’s take a look at a couple of the larger issues at play: the current problems with online discussion systems and the value of providing an outlet for reaction.

The current (and ideal) online commenting atmosphere

The intent behind comment sections on articles and posts is great: engage readers, allow feedback for your work, and build a community. But as readers of popular news sites and media outlets know well, these areas often turn quite vile and, in the end, can detract from the main story. The tone isn’t (entirely) the fault of the news source; it’s largely due to the fact that comment systems aren’t adequate for the kinds of interactions people have online. An ideal discussion system must have a way to implement the concepts of value and reputation.

If an article has five hundred comments, it’s often difficult to find which really add to the conversation. That’s where value comes into play. Users need a way to mark whether a comment adds value to the conversation or whether it detracts from it. (Preferably, there should be an option to flag a comment as inappropriate or offensive as well.) While it’s an important function, it should be very simple for users to “vote” on a comment. You’ve probably actually already seen a commenting system with value built-in: YouTube’s “thumbs up” and “thumbs down.” This also allows YouTube to place the highest-rated comments right underneath the video.

Reputation is a bit trickier; it requires value to be aggregated across all of the contributions from a particular user. eBay is a simple and well-known example of reputation, with users’ scores being a good first indication of trustworthiness. By combining vote totals, both positive and negative, across all of the comments from a particular user, a system could generate a fairly accurate idea of the user’s reputation on the site. Of course, the site would have to decide whether a user’s quantity (number of comments) or quality (value of each comment) weighs more into users’ reputation values. (Cross-site aggregation systems like Disqus could theoretically implement this on a larger level, providing a “global” reputation score for a user.)

It should be noted that the ongoing discussion between using real names and pseudonyms on sites doesn’t really play a part in this system. While real names do lend some degree of credibility, commenters sometimes have valid reasons for posting under a pseudonym (privacy, security, etc.). As long as comments are linked with an individual score, reputation can be calculated either way.¹

Many papers and readers have suggested that moderation is the key to a successful environment for comments. By moderating discussion, though, the site’s readers often cry of censorship. Moderation is an important last resort, but the task of managing comments is best left to the site’s community.

Providing a place for discussion

Allowing comments on a site also provides another benefit: a centralized and somewhat controlled place for reaction to an article. It’s a benefit, no doubt, but I’d argue that a comment section isn’t the only (or the best) way. It’s important to provide a way for people to react, but this could be done through a number of ways:

• A way to provide a URL for a reaction post
• A Twitter feed of tweets that mention the story’s URL in a short link
• The ability to use trackbacks from blog posts about the article

If a site is wary about encouraging discussion on the same page as the original article, a more link-based approach can provide a bit of a compromise.

Comments on the web – especially on media sites – hold a lot of promise, but right now they seem to detract from an article as often as they add to one. Most systems for commenting aren’t at a point where they can handle the complexity of discussion with concepts like value and reputation, both of which encourage users to be good citizens on the site. Hopefully we’ll see more sites implement better systems for commenting as the web evolves, and hopefully the discussion will improve along with them.

¹ Psudonymity and anonymity aren’t the same. I’ve yet to find an example of discussion in which complete anonymity adds any value at all.

UPDATE: I’m happy to see that the Press Herald has re-enabled comments using Automattic’s Intense Debate. I hadn’t seen it used in the past, but it looks as though it’s a robust system that, indeed, incorporates value and reputation. I’m anxious to see how the discussion evolves with the new system in place.

Digital marketing is still a relatively new area of the advertising world, and many companies seem to stumble while wading (or diving head-first) into the waters. It’s actually quite rare to find a company that does it right.

Old Spice has shown that it’s one of those companies that understands the digital space. A few months ago, the company aired what became a wildly successful ad featuring Isaiah Mustafa as an over-the-top yet deadpan “manly man” to advertise its products. The quirky campaign returned in June with a follow-up ad.

While the traditional ads stand well by themselves, Old Spice’s online integration really makes the campaign shine. They’re using Twitter’s Promoted Tweets program to reach out to the social networking community. They’re sponsoring a free shipping deal over at BustedTees, a company which (perhaps not coincidentally) sells a shirt based on the Old Spice ads. But their sponsorships and paid advertising pale in comparison to the truly innovative component of the campaign.

Over the course of the last day, Mustafa has starred in a series of over a hundred reply videos on Old Spice’s YouTube channel. Each video directly addresses a comment from a user on Twitter, Facebook, YouTube, or Yahoo Answers, and many were shot and uploaded in just a couple of hours after the original post. The replies retain the voice and humor of the traditional ads, but the content of each is conversational; the replies cover everything from marriage proposals to commercial reenactments. (In some, the Old Spice Man uses other tweets from the commenter as fodder for his reply.)

This is digital advertising at its finest. It’s funny. It doesn’t push the product in every video (some don’t even mention Old Spice at all). It bridges the gap between traditional and digital advertising by using the same actor and character in both media. It treats both popular and lesser-known social networking users equally. And above all, it starts a conversation and generates buzz.

This is where we’re heading.

UPDATE: He/they sent roses to Alyssa Milano, playing off an exchange they’d had earlier in the day. Awesome.

With last week’s iPad announcement, Apple hopes to start a technology revolution. Again. They nudged the smartphone market into the future with the release of the iPhone. Just a couple of years later, many of the over 140,000 iPhone applications available for download or purchase have pushed mobile development to new heights.

Now developers have a new platform to test their experiments. But while the iPhone ushered in a wave of applications geared around portability and location, the iPad will likely drive developers to create completely new user experiences.

The iPad offers developers a blank 9.7-inch canvas on which they can develop any interface they choose – with the added benefit of multi-touch, positioning sensors, and more. While tablets are not a new concept, they’ve always been built atop the recognized interface and conventions of a traditional computer. Users expect a certain experience with these sorts of applications; if it doesn’t feel familiar, there’s often resistance.

On a new form factor like the iPad, the resistance to change almost completely disappears. The iPad is a new experience from the start and lacks the complex interface of normal computers, allowing developers to spend less time focusing on what people expect and more time focusing on what is natural.

Developers around the world have a new, intuitive interface for applications. Expect to see some true innovation for the iPad over the next few months.

One of the latest pieces of conversation around the Twittersphere and blogosphere has been the acquisition of online footwear (, etc.) merchant Zappos by online behemoth Amazon.com. Zappos, known for their unique culture and unrelenting devotion to their customers, has understandably faced a pile of scrutiny and doubt over the future of the Zappos way of business. Even if Zappos remains an independently-run entity from Amazon, how could they possibly maintain the culture they’ve worked so hard to achieve?

Instead, I’d like to propose a different theory. Amazon’s already attempted to venture into the world of customer-focused service – in the shoe business, no less – with their own Endless.com. Although Amazon’s customer service reputation has generally improved as the company matures, they don’t hold a candle (or throw a shoe at, if you will) the CS credibility held by Zappos.  What if Amazon intends to use Zappos as a model for a new, company-wide customer service philosophy?

What do you think? Could Amazon pull off the Zappos culture? Is that business philosophy even feasible in a company of Amazon’s diversity and size? How could they make that happen (and make it work)?

Remember viral campaigns? A year or two ago, viral media – videos, sites, or entire campaigns that were so quirky or shocking that friends forwarded them through networks in a viral fashion – was the Holy Grail of the tech marketing world. Now that the viral world has gone mainstream (with some marketing firms even “guaranteeing” a viral smash), the search is on for the next Grail on the web.

Now we’ve found it: the trending topics list on Twitter. Nestled among breaking news topics are a spectrum of memes, games started by celebrity Twitterers, and – you guessed it – ads in the form of giveaways. The most recent participant in the race of trending visibility is an ancient content management company that is giving away free MacBooks to random Twitter users that choose to retweet their company’s hashtag… and they’re putting Twitter’s paid sponsors to shame. TechCrunch reports that the latest stunt is using 2.5% of all Twitter traffic.

Twitter trending is the next wave of guerilla brand promotion online, but its lifetime is sure to be short-lived. If other companies hop on the trending bandwagon, there’s likely to be a competition for the limited space on the trending list; that means that there could be more promotions for bigger prizes down the road. But if the 2.5% spirals up with a flurry of new promotions, Twitter users are bound to tire of seeing sanctioned ads for random companies in the streams of the users they follow.

Where’s the tipping point between entering yourself in a hashtag promotion and becoming a burden for spamming a stream?

Web service developers today have a relatively easy job to do if they so choose. Don’t want to spend a lot of time doing statistics for your site? Use Google Analytics. Need event listings or reviews? Pull them from an API. Storage space or bandwidth concerns? There’s Amazon Web Services for that.

One of the key tenets of the connected and shared Web world of today is to let best-in-breed service providers handle the complexities of your Web environment. How valid does that philosophy turn out to be when one of the specialized providers has service interruptions?

Services around the Web felt the pinch of service outsourcing this afternoon due to an internal communication problem with Amazon’s S3 and SQS services. The effects of the outage for some were purely cosmetic; Twitter failed to load user avatars for a period starting around noon Eastern time. For others like Basecamp and other 37signals products, the downtime caused features to be disabled temporarily.

The hardest hit this afternoon, though, were services whose entire system revolved around Amazon’s service. Users of SmugMug were greeted with a picture of the service logo watering a garden of servers and a brief message explaining the current situation. The company’s reaction? “We’re not happy about it, of course…”

Amazon does provide a service level agreement (SLA) for their AWS suite; any uptime of less than 99.9% of any given month results in either a 10% or 25% credit of that month’s service cost. (A 6-hour isolated outage on a 31-day month would result in a monthly uptime of 99.2%.) With the number of services that fundamentally depend on external providers for the entirety of their business, is a simple SLA valid?

As SmugMug CEO Don MacAskill wrote this afternoon on the SmugMug Status Updates blog:

Since problems in this industry are inevitable, and Amazon’s performance over the last two years has been so exceptional, we’ve been afraid an outage like this. I’m sure there will be more over the next few years, too.

While MacAskill’s point may sound pessimistic, it’s historically valid. While using best-of-breed services and APIs provides a service with a host of tangible benefits, it comes with the risk of multiple points of failure for a connected Web business.

(In 2006, MacAskill wrote a post detailing why S3 is a good choice for his business.)

There’s been quite a bit of discussion over Apple’s decision to ban iPhone SDK applications from running in the background.  I found a link to a great argument by Hank Williams in favor of background processes on a post (“The Flip Side of the Multitasking Argument“) over at John Gruber’s Daring Fireball.  In it, Williams mentions that “with location aware devices we can broadcast not just ‘presence’ but location.”

With a device so personal, would you want to allow background applications to broadcast your exact location or status? For me that’s an action I would want to explicitly allow, and that permission would most likely come in the form of actively running the application in the foreground.

I realize the argument may venture be on the borderline of Orwellian paranoia, but there are new considerations that have to be made for such smart mobile devices.