In the 2002 film Comedian, Jerry Seinfeld was followed by a documentary crew during a year-long comedy tour. The first shows were not pretty. It reminded me that the polished acts of even the best stand-ups don’t arrive in a flash of inspiration. They’re forged over months of toil in the hot furnace of hundreds of two-drink-minimum focus groups.
The film also shows Seinfeld talking about his trade with other comics – usually over more drinks, in the back of these same clubs. While they commiserate, others are heard onstage. This background noise, like the distant din of battle, got me thinking about how impossible their jobs are: They’re trying to say something that causes a bunch of strangers to laugh.
The shop talk often alludes to “selling a joke,” and it’s true … change some of the language and these conversations could easily be between salespeople.
This documentary reminded me of two immutable facts:
- The origin of something that meets general public approval is not really an act of Creation – in the Biblical sense. It’s more Darwinian — a gradual evolution.
- This requires feedback, often harsh, and plenty of it!
In a comedy act, feedback is instantaneous. We web-based merchants of mirth (well, we’re merchants of something) aren’t as lucky. We rely on analytics.
Is it enough to simply count sales or other types of conversions? Probably not. So we study behavioral evidence and ask questions. Questions like: Are we converting more of one group than another? Do social media visitors convert more than other visitor types? And if so, why? These are the questions that segmentation addresses. It’s just one area we can look into with standard analytics.
Here’s where I have a theory: Even after exhausting all of standard analytics, you still haven’t reached an optimum conversion level for your site. I say this because there is one step in the four-step process of selling/converting that web analytics does not easily track. As I blogged about here a few weeks ago, in Counting nose prints on your shop window, and last week on my own site, visitor interest isn’t being adequately measured.
How to measure Content Interest Index
Those posts make a case for Content Interest Index (CII). Here you’re going to learn how to measure it.
In that prior post I suggested you measure interest by looking for signs on a page that the content is doing a good job of causing people to take small actions — actions that only interested people would take.
In the real world, if you’re showing your products behind a plate glass window, the action you’d look for is getting a closer look at your products; If you’re showing those same products on a web page, the interest actions are saving information for future reference and sharing that information with others.
Here are a few of the interest actions you can tally up to calculate the CII of a page:
- Bookmarking a page locally
- Saving the page to a social bookmarking site
- Sharing the page on Facebook
- Sending a page to a printer
- Emailing the page to a friend or colleague
- Referencing the page in a Twitter post
- Your site needs to receive enough traffic to calculate CII to a statistically reliable confidence level. That means pages get hundreds of views every week.
- You’ll need an analytics system that uses a cookie-based approach, such as Google Analytics and Omniture.
- You’ll benefit the most if your site has a content management system that archives old versions of your pages.
Selling your content the way Seinfeld sells a joke
How does Seinfeld and his peers develop consistently funny material? Constant feedback. Audiences push the “Like” button every time they burst into laughter. Jokes with the most “Likes” are retained. Others are tossed or retooled until they hit their mark.
Content managers would ideally learn from this process. They should use the same rigor with their web-based stories, descriptions and graphics. So why don’t they? Until now, they’ve been trying to “sell their jokes” to an empty club. Before the CII they couldn’t hear the laughter … or the boos. As Peter Drucker famously said, “You can’t manage what you don’t measure.”
But measuring interest actions is only half the story.
To keep with the stand-up comedy analogy, when Seinfeld is listening for audience response, he automatically takes into account audience size. If there are a dozen people in the audience, getting 10 people to laugh is a big win. If it’s a theater of 2000, those same 10 voices laughing in a sea of silence would trigger major flop sweat.
Similarly, when generating a CII for a page, you need to factor in a second number. In addition to the number of “Likes” (and shares, email pass-alongs, etc.), you need to look at the number of total page views from that same time period. That’s why I prefer to take these steps in generating a site’s CII:
- 1. Score the interest actions for the page
- “Bookmark” and “print” actions get a lower score, for instance, than “Retweet” actions. The score is based on how interested you think a person is who takes this action versus another. It doesn’t have to be perfect. Just consistent over time.
- 2. Add these scores for that time period
- This gives you the final interest score. It’s the amount of interest exhibited during that time period.
- 3. Divide by the total number of page views
- This factors in the audience size, which varies from one page to another, and one period to another.
- 4. Turn this number into an index
- Indexes in statistics are usually whole numbers that hover around the norm of 100. Because the number you’ve calculated above will be a tiny fraction, make it more manageable by multiplying it by a constant, such as 100 or 1000, to raise it so something greater than 1 and less than 200.
- 5. Repeat for every other page that matters
- These are all pages “surrounding conversion funnels.” Another way to say this is that you should calculate CII for all pages that immediately lead to a call-to-action.
When you’re done, you get a list of lifeless numbers. They take on significance only when they are compared to the CIIs from past periods, for the same pages.
Similar to Seinfeld, you don’t freak out if the laughter for a joke under development is faint. You just ask yourself if the laughter is greater tonight versus last night, and what you might have changed tonight to cause this difference.
Remember that with web content, changes are more gradual. I recommend that my clients review CII every quarter, or every six months. If a content manager sees a leap in CII for a page, he or she should look at what they’ve changed and keep doing it. Conversely, a sharp drop in CII for a page calls for going to the archive and restoring a prior version (if possible!).
Making the data visible
As a digital marketing analyst I know that much depends on how this data is presented to content managers. That leads to my next post on this topic, coming later this month. There I’ll be showing some breakthrough ways to visualize this (and other) data to make content changes as intuitive and instantaneous as it is for professional comics.
Maybe they’ll never get a feature length documentary about you, but you’ll know you’ve honed your content with the discipline of a show business legend.
Related articles by Zemanta
- Counting Nose Prints On Your Shop Window (socialmediaexplorer.com)
- Evaluating Content Performance For Reaching Site Goals (verticalmeasures.com)
- Excellent Analytics Tip #18: Make Love To Your Direct Traffic (kaushik.net)
- Make The Most Of The Traffic You Already Have! (The Only Resource You’ll Ever Need To Read On Lowering Your Bounce Rates) (codemyownroad.com)
VIP Explorers Club
- A Guide to Building the Best Possible Facebook Custom Audience with Jon Loomer
- How Marketing Will Change at the Incite Marketing Summit
- Book Review: ‘The CMO’s Periodic Table’ Reveals the True Elements of Marketing
- How to Use Reddit to Drive Traffic to Your Site
- Trump Launches National Snapchat Filter and Millennials Are Furious