The content’s there but the data often isn’t

Neil Heyside’s Oct. 18 story on PBS MediaShift about how newspapers should analyze their content by source type – staff-produced, wire/syndicated or free from citizen journalists – got me thinking about other ways content should be analyzed to craft audience-driven hyper-local and paid-content strategies.

 

Picture 1 Most news sites have navigation that mimics traditional media products – News, Sports, Business, Opinion, and Entertainment.   However, those types of broad titles don’t work well with digital content because people consume news and information in bits and pieces rather than in nicely packaged 56-page print products and 30-minute TV programs.

Picture 2

Each chunk, each piece of content – story, photo, video, audio, whatever – should be tagged or classified with a geographic area and a niche topic so a news org can determine how much content it has for each of its highest priority audience segments – and how much traffic each type of content is getting.

By geographic area I mean hyper-local.   East Cyberton, not Cyberton.  Maybe even more hyper – East Cyberton north of Main Street, for example, or wherever there’s a distinct audience segment that has different characteristics and thus different news needs and news consumption patterns.

 

Similarly, news orgs need hyper-topic codes, especially for hyper-local topics.  The Cyberton community orchestra – not Classical Music, Music, or Entertainment.  If a news org is looking at web traffic data for “Music” it should know whether that traffic is for rock music or classical, and whether the content was about a local, regional, national or international group.

 

Picture 3 Oh, and there’s one more aspect to this hyper-coding.  Content should be coded across the site.  Ruthlessly.  For example, to really understand whether it needs to add or cut coverage in East Cyberton, a news org needs to add up those East Cyberton stories in Local News, plus those East Cyberton Main St. retail development stories in Business, and those editorials and op-eds in Opinion about how ineffective the East Cyberton neighborhood council is, and….

 

Sometimes these hyper-codes are in content management systems but not in web analytics systems like Omniture or Google Analytics.  Knowing what you’ve got is great – but knowing how much traffic each hyper-coded chunk of content is equally if not more important.

 

Whether the hyper-codes and thus the data are there only makes a difference if a news org is willing to take a hard, nontraditional look at itself.   The data may suggest it needs to radically change what it covers and the way it allocates its news resources so it can produce “relevant, non-commodity local news that differentiates” it, as Neil Heyside’s PBS MediaShift story points out.

 

Picture 4 Heyside’s study of four U.K. newspaper chains has some interesting ideas about how a news org can cut costs but still maintain quality by changing the ways it uses staff-produced, wire, and free, citizen journalist content.  The news orgs in the study “had already undergone extensive staff reductions.  In the conventional sense, all the costs had been wrung out.  But newspapers have to change the way they think in order to survive.  If you’ve wrung out all the costs you can from the existing content creation model, then it’s time to change the model itself.”

 

Decision-making
If a news org doesn’t know, in incredibly painful detail, what type of content it has and how much traffic each type is getting, then it’s not arming itself with everything it can to mitigate the risks of making radical changes such as investing what it takes to succeed in hyperlocal news and in setting up pay walls.  Both are pretty scary, and it’s going to take a lot of bold experimentation – and data – to get it right.

 

 

 

Two million users? Not enough…

Xmarks Xmarks had lots of users but little engagement.  Apparently it forgot to figure out what the users needed or wanted.

From WebNewser:

“For the popular bookmark syncing service Xmarks, 2 million users was apparently not popular enough. Co-founder and CTO Todd Agulnick announced on the company blog Tuesday that, despite growing by 3,000 users each day, the startup was floundering and would shut down its service in 90 days….

Unfortunately, users who tried the system were looking for answers to questions rather than topical lists of sites.”

Believing in time-on-site

A Sept. 12 Pew Center study found that “Americans are spending more time following the news than over the past decade.”  Great news – or is this yet another misleading key performance indicator, as my previous blog post about time spent on site might suggest?

No – I like the Pew Center study.  It’s a study of attitudes and feelings.  Good old-fashioned survey research (with all of its mind-numbing statistical sampling), is an essential component of web analytics.  Web site traffic data is audience behavior – the “what.” News orgs have to havPew Research Centere attitudinal research to understand the “why” so they can attract audiences that aren’t coming to their sites.

The data you get from Google Analytics or Omniture is enticing, isn’t it?  (Work with me, here….)  Oh wow, we can track every click!  Ah, yes, we can track every click – but we can ONLY track clicks on OUR site, not on anyone else’s.

A time-on-site calculation can only be harvested for you if someone clicks on a page in your site and generates a page view that’s counted by your Google Analytics/Omniture account.  Time-on-site is the time in between the first page clicked on YOUR site and the last page clicked on YOUR site.

This means:

1.   If someone clicks on YOUR site and then immediately goes to another site (a bounce),
Picture 8 it’s not included by Google Analytics/Omniture in the time- on-site calculation.  It’s like it never existed, time-on-site-wise.

It IS counted as one visit and as one page view.  So that means that all of those people who come to your site regularly (you know, the ones we really like) just to get the latest on a story aren’t counted in time-on-site – and they should be.

2.    If someone is on his/her third page in your site and opens another tab anPicture 5d goes to another site for twenty minutes before returning to your site and clicking on another two pages, those twenty minutes are included in time-on-site – and they shouldn’t be.

3.    The time a person spends on the last page of your site isn’t counted.
Picture 7 If someone clicks through a few pages on your site and spends 15
minutes utterly absorbed in a story before leaving your site to go pay
bills online, those 15 minutes aren’t included in time-on-site – and they should be.

So, time-spent-on-site is always either over- or under-counted.  And you’ll never know which – this makes this metric utterly unreliable as an indicator of success or failure.  So you can’t make a decision with this data because you can’t know whether your action – a section added, the number of long-form videos reduced – caused time spent to go up or down.

More importantly, these days it really doesn’t matter how much time people  actually spend on a news site.  What matters much, much more is whether people are engaged with the news, whether they believe news sites are an essential component to their lives, so much so that they come back repeatedly, rate a story with five stars, leave comments, click on an ad, and otherwise use the site.  It doesn’t matter whether they spend three seconds or three hours.
More say they graze for news

That’s what makes the Pew Center finding so exciting (surely you’re still with me on how great web analytics is?).  People actually said they’re spending more time with the news now than they did a decade ago.  It doesn’t matter whether they actually are (!) – they believe they are.
I wish every news org could afford its own Pew Center-like attitudinal research study so it could track how engaged its own targeted audiences are (or aren’t), and to understand how to get and retain new audiences.  The information wouldn’t always be pretty, but at least it would be data that could make a difference.

Decision-making with data

Wow – a web analytics story in the New York Times!  And one that probably didn't spark mass hysteria by saying that looking at web traffic data is the end of journalism as we know it.  

Jeremy Peters' September 5 story, "Some Newspapers, Tracking Readers Online, Shift Coverage," nicely framed the issues about using web analytics to make decisions about coverage.  At one end you have journalists who completely ignore audiences and report whatever they want.  On the other there are those who "pander to the most base reader interests." 

This was one of the few mainstream media stories I've seen that recognized the new role that audiences have in informing – not dictating – news decisions. 

Actually what was the most interesting to me was that Raju Narisetti, the Washington Post's managing editor of online operations, said that he used "reader metrics as a tool to help him better determine how to use online resources….the data has proved highly useful in today's world of shrinking newsroom budgets." 

Picture 1 In other words, Narisetti didn't just look at overall site traffic numbers like total visits or unique visitors and say "Oh, that's nice," or "Whatever! I've got a paper to put out."  He had to cut his budget, he saw that long-form videos had little traffic compared to other types of content, and then he decided to cut "a couple of people" from that department.

Narisetti made the decision – the data didn't.  He used the data.  In my rather narrowly focused web analytics world that is a really, really big deal.  Data just wants to be useful.  Seriously.  If it could talk, it would say that it doesn't want to be any part of a daily e-mail "about 120 people in The Post's newsroom" get that lays "out how the Web site performed in the closely watched metrics – 46 in all."

Forty-six! 

For data to be useful you have to ask it specific questions that you know it can answer.  A metrics report won't tell you what department to cut, and by how much.  However, it could tell you what types of stories get the most and least audiences, and which story types appear to have growing or declining audiences.  You can also get indicators – such as the percent of long-form videos people watched to the end, rated and/or commented on – of whether you're engaging audiences.

Then the hard part – the decision – is up to you.

 

Page views: bad metric #3

Why do news organizations persist in using total page views as a measure of success?  Perhaps because if you're afraid of numbers then you're even more afraid of bad numbers, or numbers that tell you that your site isn't as successful as you want.

As with unique visitors and time on site, page views is a deeply flawed metric for understanding how a news organization is growing and retaining audiences.  

If the number of page views goes up, it could be a good thing.  Or, it could be bad.

If page views go down, it could be a bad thing.  Or – you guessed it – it could be good.

We would all like to think that a soaring number of page views means lots of people are eagerly pawing through our sites reading everything that's written.  However, how many times have you gone to a site and clicked on, say, 12 pages, fruitlessly looking for something? 

This is counted as:

1.    One unique visitor
2.    One visit
3.    12 page views

And one dissatisfied person who may not come back.

The page views metric rewards the bad design and navigation that many news sites have (sorry).  Most news sites persist in using section titles that are the same as their legacy media product (e.g., "Local News," "Life"), leaving audiences – if they're so inclined – to have to click numerous times before landing on a story about a particular city or activity like gardening.   

Or, a site breaks up a story into multiple pages, which can be annoying to a reader and reduces the possibility the reader will read the entire story and rate it, e-mail it or leave a comment.  What could be counted as one page view with a comment is counted as, say, five page views. 

If a site is redesigned and readers can find what they want with fewer clicks, total page views will – should – go down. 

And, dynamic content, or content that changes on the page without the reader clicking anything – isn't counted at all.  Streaming stories, videos, widgets, Flash applications, podcasts?  One page view.

To truly grow, a news org must understand every action its audiences are taking on its sites.  These are challenging times that require news sites to experiment and try many different things.  Not all things will work, which means sometimes the numbers will be bad.  But – you guessed it – that's a good thing.  We have to know if something's not working so we can fix it. 

 

“A trend is a friend”

Someday I’ll give up web analytics and move on to something real like pottery or something, but until then I’ll keep fighting the good fight to get news orgs to stop using monthly unique visitors as an indicator of success.

It’s tempting, I know, to count UVs because the number of monthly subscribers is the standard for print, and the number of people Nielsen says watched a program is the standard for TV.

But technology has made everything different.  Strategy is more important than ever, and understanding audiences, not just counting them, is essential.

UV overcount

UVs are counted by counting the number of cookies, or computers,
that go to a site.  This means UVs are always significantly overcounted
or undercounted.

If one person uses three computers, it’s counted as
three unique visitors.

UV undercount Conversely, a number of people going to one computer – for example, at a school or library – means that UVs will be undercounted.

Neil Mason, a web analytics guru who does some pretty thorough research for his clients, noted in a recent ClickZ column that UVs are usually overcounted.

(So that’s the reason why news orgs use total UVs – would they use this number if it were consistently undercounted?  Don’t think so…)

Mason notes that while the UV metric is “particularly important for those sites that are dependent on advertising revenues as a major source of
income,”  it “must always be treated with caution and never taken at face
value.”

Believe me, it’s rare to hear a web analytics expert use the word “never.”

Gingerbread man The advice from the ever-pragmatic Mason:  “A trend is a friend.”   Analyzing significant increases or decreases over time will give news orgs the information needed to build audiences.

(Trends are indeed friends, but don’t even think about using counts of Facebook friends and Twitter followers!)

 

An ounce of Prevention

I exercise regularly and don’t smoke, but I avoid reading Prevention.com. It’s just too annoying to be reminded of all of the other “smart ways to live well.”  When I see “by the editors of Prevention.com,” I have this picture in my mind of a bunch of really healthy people popping out cheerful stories like “5 Vitamins Your Bones Love” and “10 Reasons You’re Always Exhausted.”

Prevention Now I have another (annoying) vision, thanks to MinOnline‘s story about how well Prevention is using web analytics.  Of course a staff that is so pragmatic and probably always mentally alert would resist “going for the cheap link grab and traffic spike” – the junk food of web analytics.

I haven’t been sleeping well because I think too much (the top reason people don’t get enough sleep and are therefore exhausted), so I’ll just plop in these two paragraphs verbatim from Steve Smith’s MinOnline story, “At the Building Prevention.Com, Only The Abs Are Flat.”

Page views rpt Prevention stays “on its own brand message and [courts] the kinds of audiences that it and
its advertisers really want. ‘We got back to engaging with our customer
in the ways we knew they wanted us to engage with them,’ says [vp/digital Bill] Stump.
Fishing for any and all eyeballs and courting simple traffic spikes in
the search-driven universe doesn’t pay off in the end. ‘You get waves of
traffic, but the tide goes back out and what are you left with?’
Instead, by keeping to the needs of the ‘core customer’ in everything
that goes out to syndication or into the e-mail newsletters, prevention.com is courting the
people who tend to stay.

Page views per visit “Now, each big wave raises the sea level for all of prevention.com’s metrics, says
Stump. In the last two years, overall page views climbed 60%. In the
last year, the number of visits per user went up 12%. But it is the
engagement metrics of which Stump is proudest. ‘The number that warms my
heart,’ he says, ‘is page views per visitor that are up 49%.’ That
means the new visitors are sticking with the site and drilling much
deeper than they ever have before. ‘In general, advertisers want an
engaged audience. They want the metrics that show that people value your
brand and come to you for something that is unique. We own natural
health and fitness and beauty. We are the authentic voice.'”