Archive for the ‘statistics’ Category

Three Ways To Wag The Long Tail Of Content

I was glancing at my blog traffic stats the other day, and noticed something that made me sit up and take notice – after three years, the most-viewed post on this site continues to be the opening post in my good communications planning series, with over 125,000 views.

What’s more, the traffic to this post is continuing to rise over time. Here’s a chart of the daily traffic to the post:

Doesn’t look much like the typical ‘long tail’ image of traffic over time, does it?

I got to wondering why this is happening. Here are my ideas:

 

1. Useful content

The 13-part series of posts I wrote on communications planning walk through the process of creating a communications plan, from start to finish. It (I hope) is useful stuff; content that people find applicable and helpful.

2. Evergreen content

These posts are as helpful today as they were when I wrote them. While best practices around plan development will, I’m sure, evolve over time, this series should remain helpful for a long time.

3. Optimize for search

As someone pointed out to me on Twitter, Google “good communications plan” and this post is the top result. “Communications plan” continues to be one of the top search terms used to reach this site. I thought-through the titles of the posts, and the cross-linking between them, when first writing them, and it worked well.

I’d love to hear your take – have you experienced this kind of effect before? What caused it then?

 

 

 

 

Lies, Damned Lies and Mis-understood Statistics

Earlier this week, All Facebook featured a post on a report by DDB and OpinionWay examining the reasons that Facebook users “unlike” brand pages.

The key findings, as reported by All Facebook (the report appears to have been taken down from SlideShare, so I can’t link to it):

  • The brand was no longer of interest to me (49 percent);
  • The information available was not interesting (46 percent);
  • Information was published too often (36 percent);
  • The brand published information I did not appreciate (27 percent)
  • Information was not published often enough (14 percent).

Interesting, useful data.

This follow-up chart in the post, however, is next to useless.

Why is this chart useless?

Because the sample size is too small for this kind of segmentation.

The original data is useful because the analysis is conducted at an aggregate level, over 630 respondents. At that sample size, we’re looking at a 3.9% margin of error at a 95% confidence level. That means, while there may be some variation among the top results, they’re useful at a high level.

Dig down to a country level like the chart above, though, and things start to fall apart. With a sample size of 78, given the number of Facebook users in the United States (155,746,780 according to Facebook), the margin of error for the US numbers is over 11% at that same confidence level. It’s not just the US, either – the margin of error for the France numbers is over 8.5%. Despite this, there’s no mention of these details on the post or comments; just an assumption that the numbers are correct.

This is a great example of why I think math is a critical skill for PR professionals.

PR pros need to understand the difference between valid statistics and invalid ones, so they can take advantage of useful information (like that at the top of this post) and disregard the non-useful stuff (like the regional breakdown above). What’s more, they need to know what’s news and what’s non-news too, so they can make an informed decision on what to pitch as the former and what to advise their clients to pass on promoting.

Are you comfortable reading between the lines when it comes to statistics? If not, it might be time to brush up.

Correlation/Causality Confusion

As the social media space slowly matures, we’re starting to see more and more reports released by companies offering insight into the ebbs and flows of peoples’ behaviour online. Not surprisingly, these reports are pretty popular – they possess two desirable attributes:

  • They offer easy-to-digest, soundbite-sized nuggets of information – perfect for short-form media such as Twitter
  • They offer the potential for increased understanding of the latest trends

However, these reports can sometimes unintentionally misrepresent the data, through a simple statistical error:

Correlation does not dictate causality.

What does this mean?

It means that just because there is a link between two things, it does not mean that one causes the other. So, if there is a correlation between A and B:

  • A could cause B
  • B could cause A
  • C (and/or D, E, F etc) could affect both A and B

As Bob Hoffman explains it over at The Ad Contrarian:

If you were to study people who are hard-of-hearing you would probably also find that they have a much higher likelihood of being bald. Does this mean that bad hearing causes baldness? Of course not. It occurs because old age causes both hearing and hair loss. So there is a correlation between deafness and baldness, but there is nocausality. One does not cause the other.

One more time: correlation does not mean causality. Just a quick thought, but an important one. Remember this next time you read a snippet about a new study, and make sure you always read closer before believing the headlines.

(Image: Shutterstock)

Think PR People Don’t Need Math? Think Again

Public relations folks aren’t generally very good at math, according to their reputation, anyway. They’re creative people, you see? They work magic with words; they build relationships with people; they persuade people. So what if they’re no good with numbers?

I argue: it matters. A lot.

Here are just a few of the tasks that you need basic math to accomplish:

  • Social media and traditional media audits
  • MRP analysis
  • Social/traditional media monitoring analysis
  • Situational analysis for plans
  • Any kind of statistical analysis for data-based news releases
  • Market research analysis/recommendations

Get my point? “I’m no good with numbers” just doesn’t cut it.

I’ll admit it – I’m a math nut. I’ve worked as a data analyst and I nearly took a math degree (deciding instead to shoot for business) so I have a bit of an affinity for this stuff. Regardless, if you work in communications and you break into a sweat at the first sight of a graph, you need to study-up FAST if you want to progress.

No, just being able to create a graph in Excel does not cut it. Data alone does nothing. You need to be able to analyze that data. It’s not rocket science, but at a minimum you really should be able to, for example:

  • Compare two sets of numbers and calculate the percentage difference between the two
  • Know that 100% growth is different to 100% of something
  • Conduct simple statistical analyses of data – is there a trend?

In reality, we as a profession need to raise our game beyond statistical basics. We can’t just think about the numbers after the fact – a results focus needs to feature in every aspect of our work; especially online where data is so readily available. It takes planning and forethought to cut through the mass of data and turn it into useful, actionable, relevant information.

For a great example of how we should be building analysis points into all of our campaigns, check out this excellent post on integrating bit.ly and Google Analytics in a campaign. It shows a relatively simple process for integrating basic analytics into the links we publish on different social media platforms.

Results matter. That means numbers matter, and you need to know how to handle them.

Are you up-to-speed on these skills? How have you found ways to integrate analysis throughout your plans?

(Update: Radian6 just announced new features on their platform that may make this process easier – check out the Radian6 post on web analytics integration)

Forrester’s New Technographics Data – How Do Canadians Measure Up?

Forrester recently released an updated set of data for their very useful technographics profile tool.

What’s a technographics profile?

Forrester’s tool divides consumers into six groups along a “technographics ladder.”

The tool is based on consumers’ social behaviours and on techniques detailed in Josh Bernoff and Charlene Li‘s book Groundswell (which I have not yet read, but plan to shortly).

The six groups are:

What’s the news here?

As Bernoff put it in a post on the new technographics data yesterday:

“…the big news in 2008 is that, not unexpectedly, social technology participation has grown rapidly.”

 

Critically, the proportion of people who are “inactive” is down from 44 per cent to 25 per cent. Just a quarter of US online adults are not engaged in social media in any way. Almost seventy per cent of online Americans now at least read, listen or watch some form of social media.

Enter the Canadians

Interestingly to those of us north of the border, Canadians are now featured, for the first time, in Forrester’s tool. Canadians are generally thought of as web-savvy, with high connectivity (according to the federal government, 73% of all Canadians use the Internet now) and high engagement (Canada is known as a leader in terms of Facebook use, for example).

So, how does Canada measure up against other countries in this new data? You may be surprised – this data is pretty controversial.

Let’s look at the groups one by one.

creators

Creators are the people churning out the content – bloggers; podcasters; artists.

Surprisingly, according to Forrester’s data Canada is far from a leader in the creator category. Just 13 per cent of Canada’s online adults fall into this category, compared to 21 per cent in the US, 40 per cent in China and a whopping 51 per cent in South Korea.

critics

While ‘critics’ don’t necessarily contribute their own content, they do contribute their thoughts and opinions on other peoples’ content. They’re the blog commenters, the podcast callers-in, the wiki editors.

Canada is actually at the bottom of the pile of critics according to this new data. Only Germany is close to being as low.

collectors

Canada fares a little better in terms of collectors – the people who, while they may not contribute directly, actively save, bookmark, tag, vote and otherwise store/arrange online content. Canadians are middle-of-the-road here according to Forrester.

joiners

With Canada’s reputation in the social networking arena (especially with Facebook), you would expect that there would be a lot of joiners here. For the first time so far, this is the case. Canada ranks third in this category, behind only South Korea and Australia.

spectators 

Canadians are once again near the bottom of the pack in terms of people who passively consume online media. Only Germany ranks lower in this data.

inactives

You probably know what to expect by this point: online Canadian consumers sit close to the head of the pack in terms of consumers who are inactive in social media.

Notes of caution

Three things here set-off my “don’t jump to conclusions” alarm:

  1. The survey looks only at online consumers;
  2. We don’t know the methodology;
  3. This flies in the face of other research out there.

Online consumers only

From what I can tell, this data only looks at the breakdown of online consumers. It does not, I believe, consider those people who are not online.

While this makes sense as we’re looking at peoples’ behaviours online, it does not consider the proportion of the total population that these consumers make up. Canada is highly connected, while others may have a much lower percentage of their population online.

For this reason, Canada may be short-changed somewhat in this analysis if you look at the broader populations of each country.

We don’t know the methodology

To be more accurate, I don’t have the $995 needed to buy the complete Forrester report detailing the methodology for this study. That’s not Forrester’s problem – it’s mine – but it sows the seed of doubt in my mind over the methodology behind the study and whether it is consistent with that used in the other countries listed.

What about the other research?

As Ed Lee and Sean Moffitt (among others, I’m sure) have pointed out, there’s a host of other pieces of research out there that suggest Canadians are far more engaged in social media than this data indicates.

Back in June, in fact, Ed highlighted comScore research showing that more than 84 per cent of Canadians are active on social networks and that 89 per cent of Canadians watched online video. These numbers seem a bit extreme on the other side to me, but it goes to show that there are lots of different data sets out there.

These three factors may go some way to explaining the difference between the picture of Canadians painted by this data (below) and what I expected to see.

canada

Furthermore, this is raw data that leaves a lot of room for interpretation.

Personally, I’m taking this new data with a pinch of salt until I have more information.

What’s your take on this?

(Image credit for the first two images: Josh Bernoff. All remaining charts created by me, based on Forrester data.)

Twitter Packs, Tweetmeme and Releasing Social Media Metrics

Analyzing Twitter Usage

I came across an interesting tool the other day – Twitter Stats – provided by Brad Kellett.

The tool is a web-based adaptation of code by Damon Cortesi, who produced a tool to grab stats from a user’s Twitter feed and aggregate them to show overall stats.

The tool looks at:

  • Tweets per hour of the day
  • Total Tweets per day
  • Total Tweets per month
  • Top @replies
  • Top overall @s

Here’s a quick analysis of the stats from my Twitter account.

Total Tweets Per Month

If you’ve read any of my recent posts on my social media tool usage, this won’t surprise you.

As with many tools, my usage of Twitter has rocketed since I started using it in September. There are two good reasons for this:

  1. As I follow more people (I follow roughly 300 right now), I find more interesting conversations that I want to participate in
  2. As more people follow me, Twitter becomes a more powerful tool for soliciting feedback and opinions

Of course, December covered the Christmas period when I wasn’t at work and was generally freer to engage with people.

twitterstats1

Total Tweets Per Day

I’m not sure what to make of this one. My usage drops off on Friday, perhaps because I’m less available in the evenings on Fridays.

I don’t, however, have an explanation for why Tuesdays and Wednesdays are my busiest days. Anyone have any ideas?

twitterstats2

Tweets Per Hour Of The Day

This makes sense. I’m most available at two times of day:

  1. When I first get to work, drink my coffee and catch up on what’s going on
  2. During my lunch hour

There’s a drop between 6 and 7, when I’m usually out running, but I’m usually online around 8-9pm, when I tend to write my blog posts.

I’m a little surprised that I apparently do some tweeting between 4 and 6am. Bizarre.

twitterstats3

Top @ Replies

These are apparently the top ten people I send messages to:

  1. Ed Lee (edlee)
  2. Chris Brogan (chrisbrogan)
  3. David Jones (doctor_jones)
  4. Todd Defren (tdefren)
  5. Michael Allison (michaelallison)
  6. Mitch Joel (mitchjoel)
  7. Neville Hobson (jangles)
  8. Joseph Thornley (thornley)
  9. Bryan Persons (bryper)
  10. CC Chapman (cc_chapman)

Makes sense to me – all ten of these guys (yep, all guys) are smart and interesting. Seven of them produce podcasts that I subscribe to (if you include Todd on the New Media Release Cast).

If you’re on Twitter, into social media and you’re not following these guys, you’re missing out.

twitterstats4

Top Overall @s

I’m not sure what this one is all about. Perhaps the people I message out of the blue, as opposed to reply to? Enlighten me.

A few extra names here – more people that I suggest you follow if you’re into this space:

  1. Shel Holtz (shel)
  2. Joseph Jaffe (jaffejuice)
  3. Chris Clarke (clarkey) – blogs at StudentPR.com, which appears to be down at the moment
  4. Brian Solis (briansolis)
  5. Tod Maffin (todmaffin)

twitterstats5

What do your stats look like? Do they resemble mine? What patterns do you see?