Web Analytics and Optimization for Content Sites

August 21, 2007 – 7:13 am
Printer-friendly version Printer-friendly version

AddThis Social Bookmark Button

A common problem that faces many organizations hosting a strictly content-driven website is how to turn the data thrown out by any of the applications out there into something you can monetize. Sure, more page views mean more impressions which means more cash for your company, but what if you’re tagging a series of event-based activities (see Best Practices - Tagging Flash Files) which don’t generate additional impressions? Then the influx of page views you begin reporting isn’t actually contributing anything towards your bottom line, but it’ll certainly seem like you’re having a stellar quarter from your analytics reports.

(Side note here: this kind of thing can be solved with a separate profile that filters out all the “extra” activity you’ve tracked along the way)

GLE - DiagramThe metrics I follow for content-based sites that heavily rely on advertising dollars to fuel revenue can be grouped into three categories, which in tern actually lays out the analytics and optimization strategy for these sites:

  • Growth
  • Loyalty and
  • Engagement

All content-driven sites generally want to tell the same story of growing their visitor base and keeping them coming back constantly while they consume (engage with) more and more content on the site. Before I go any further, let’s take a second to get our heads around why this is a model for success.

Why should I care?

Money BagAt the heart of content-driven sites is ad revenue which is provided via impressions. Companies usually agree on a Cost per Mille / 1000 (CPM) impressions delivered and just like that, we have revenue. So at first the initial thought throughout the industry was “more page views = more impressions = more $$$”. This wasn’t a bad way to think about things, after all page views are a measure of engagement with your content, and page views is a metric that’s easy for anyone to monetize.

The problem lies with the fact that focusing on this one metric got people spending and focusing on the wrong things. In desperate efforts to drive up ad revenue that was likely tied to many people’s bonuses, site owners looked towards acquiring more visitors as their best bet to drive up page views. Efforts like SEO and SEM were seriously utilized in order to drive as many eyeballs to your site as possible - why? Because that’s all that advertisers cared about.

Leaky BucketBut there was an underlying problem with this method; driving a bunch of people to your site that hate what you have to offer is analogous to a leaky bucket. You keep pouring in money towards your acquisition strategies, but you never realize return on your investment when all your visitors just pop in and pop out of your site. Also, it becomes hard to charge higher CPMs to your advertisers if all you have to offer them is “500,000 people came to our site last month…but only 10% clicked through and only 5% ever returned.”

The other problem with the “grow first, ask questions later” model, is that without engaging content, you’re never getting the maximum amount of value out of the visitors you attract to the site.

Say I have a site with average monthly visits (not visitors) of 500,000, the average number of page views per visit of those visitors is 2, the average CPM for all my inventory is $10 and the average number of ads per page is 3. Using these numbers, my monthly ad revenue is (500,000 * 2 * 3) / 1000 * 10 = $30,000. Now assume that by focusing on optimizing my visitor’s site experience (see Bryan and Jeffery Eisenberg’s Book), I’m able to drive page views per visit up to 4. Note: I didn’t grow my visitor base at all, I simply got that visitor base to look at more of my content. The result? An incremental increase in monthly ad revenue of $30,000 a yearly increase of $360,000. And I didn’t even have to spend a dime on SEM and SEO - see what your CEO thinks about that!

What the above discussion hopefully demonstrated is a successful acquisition strategy is just one component to a content-driven site’s success. Once you’ve driven more people to your site, you want to keep them and you want them to love what you have to offer and there we have growth, loyalty and engagement.

Ok, I get why this is important…but what do I do?

Oddly enough, the easiest part of the growth, loyalty and engagement strategy is actually growth. There are plenty of methods out there to grow your visitor base (namely SEO and SEM), but it’s far more difficult to have a loyal visitor base and one that engages with your content.

What I recommend then is a backwards approach. First focus as much as possible on delivering engaging content to your visitors, then move onto looking at ways to improve loyalty and finally examine your growth strategy. The result of this strategy is a far higher return on all the dollars you’ll spend at your growth stage as well as, and this is far more important in my book, a far greater site experience for your visitors.

Great, but how do I know I’m delivering engaging content?

This is where web analytics fits in. The metrics (they’re actually key performance indicators (KPIs), but no worries) I prefer to use to gauge how engaging your content is are pretty simple (many of which borrowed from Eric T. Peterson):

  • Average Pages Viewed per Visit
  • Average Time Spent on Site
  • Percent of Visits Greater than 60 Seconds
  • Percent of Committed Visits (> 5 page views per visit)

The above metrics are available from most tools, but let’s take a look at how to grab them from Google Analytics.

Measuring Engagement - Average Pages Viewed Per Visit and Average Time Spent on SiteAverage Pages Viewed per Visit and Average Time Spent on Site

Both of these guys are pretty easy as you can grab them right out of your dashboard report as soon as you open up your profile.

Just look for Pages/Visit and Avg. Time on Site.

Percent of Visits Greater than 60 Seconds

GLE - Percent of Visits Greater than 60 Seconds60 seconds is a threshold you can adjust depending on the content on your site. 60 seconds may be too long or too short for the type of content you have so you’ll have to use your discretion. Google Analytics also won’t often allow you to grab the specific segment you’re looking for so you may have to suffice for something different.

As for pulling this data, navigate to your Length of Visit report which can be found under Visitors -> Visitor Loyalty. Feel free to add this report to your dashboard for easy reference later on. You’ll likely have to do a little bit of addition to get the number you’re looking for, but that’s no biggie.

Percent of Committed Visits (> 5 page views per visit)

GLE - Percent of Committed VisitsAgain, this is a metric that can be adjusted if needed. 5 page views per visit may be a bit too much or too little to be considered a committed visit. To pull this information, navigate to your Depth of Visit report which can be found under Visitors -> Visitor Loyalty. Feel free to add this report to your dashboard for easy reference later on. You’ll likely have to do a little bit of addition to get the number you’re looking for, but that’s no biggie.

Cool! Now onto loyalty…

It’s logical to assume that really terrific, engagement content should always keep people coming back for more, but these days that may not be enough. What I try to remember is that the internet space is no different from any other and the dot com crash proved that. A truly unique value proposition that gives me (as a visitor) something I can’t get anywhere else and delivers incredible value at a low price (price doesn’t have to be in dollar amounts, what about my time and patience?) will always make me want to return. The name of the game is satisfaction.

Your best metrics to gauge how well you’re doing at keeping a loyal visitor base then are:

  • Percent Returning Visitors
  • Percent Loyal Visitors (> 5 visits per month)
  • Average Visits per Visitor

Percent Returning Visitors

GLE - Percent Returning VisitorsThe number of returning visitors (actually, this is visits in Google Analytics) as a percentage of your entire visitor population for a given time period. This data can be pulled from the New vs. Returning report which can be found under the Visitors menu. Feel free to add this report to your dashboard for easy reference later on.

Percent Loyal Visitors (> 5 visits per month)

GLE - Percent Loyal VisitorsAs usual, adjust the “> 5 visits per month” threshold to whatever makes the most sense to you and your site’s content. If the site is updated every day, loyal visitors may be those visiting more than 15 times a month.

To pull this data, navigate to the Loyalty report which can be found by navigating to Visitors -> Visitor Loyalty. Feel free to add this report to your dashboard for easy reference later on. You’ll likely have to do a little bit of addition to get the number you’re looking for, but that’s no biggie.

Average Visits per Visitor

Measuring Engagement - Average Pages Viewed Per Visit and Average Time Spent on SiteThis one is pretty well a no brainer, we’d all like to see the average number of visits per visitor improve but you’ll still have to work on baselines as to what’s appropriate for this KPI. You can calculate this metric directly from your pre-configured dashboard. Just divide visits / visitors.

Woo hoo! Now onto growth!

As previously stated, growth is probably the easiest place to improve upon in this three prawned strategy. Your most popular acquisition sources will likely be SEO and SEM, but there’s all sorts of partnering to be done with other content sites, affiliate networks and plenty more. Don’t let keyword buying occupy your entire acquisition universe and make sure to evaluate the success of your keyword buys by looking at the growth, loyalty and engagement metrics for just that section of visitors (you’ll have to create a new profile to do this).

I should probably note that I say “easy” in that growth has the most easily available options to you for improvement. However, actually picking the right targeted sources is by all means not easy and many marketers have to struggle with these problems everyday (hence the birth of behavioural targeting).

That being said, your two best metrics for evaluating how well your growth strategies are working overall are:

  • Percent New Visitors
  • Ratio of New to Returning Visitors

Percent New Visitors

GLE - Percent Returning VisitorsThe percent of new visitors (or visits in Google Analytics) can be an interesting metric to monitor when employing a high powered SEO/SEM strategy in that you want to ensure your efforts aren’t attracting an existing visitor base. This data can be pulled from the New vs. Returning report which can be found under the Visitors menu. Feel free to add this report to your dashboard for easy reference later on.

Ratio of New to Returning Visitors

GLE - Percent Returning VisitorsFinally, the ratio of new to returning visitors provides an excellent way to understand what strategy your site is pursuing (whether intended or not). For example, a ratio of new to returning visitors of 3 would say, “for every visitor that returns, I bring 3 new ones in”. Ideally speaking, you’d like to see a ratio of new to returning visitors of around 2 in order to keep a healthy growth rate, but ensure that enough of your growing audience is still returning.

A ratio that’s higher than 2 indicates your site is in “growth mode” and your acquiring more visitors than you’re keeping. A ratio that’s lower than 2 indicates that although you’ve got a very loyal audience, you may want to look into some ways of growing it out.

This metric can be pulled from the New vs. Returning report which can be found under the Visitors menu. Feel free to add this report to your dashboard for easy reference later on.

Now tie it all together…

Just in case anyone got a little side tracked with that run thru let’s put those metrics in perspective again:

Growth

  • Percent of New Visitors
  • Ratio of New to Returning Visitors

GLE - DiagramLoyalty

  • Percent Returning Visitors
  • Percent Loyal Visitors (> x visits)
  • Average Visits per Visitor

Engagement

  • Average Pages Viewed per Visit
  • Average Time Spent on Site
  • Percent of Visits Greater than 60 Seconds
  • Percent of Committed Visits (> x page views per visit)

Finally! Now that we’re all on the same page about why growth, loyalty and engagement are the aspects to a good strategy for content-based sites and we have the accompanying metrics to evaluate the success of that strategy we’re set for success! Of course I’m leaving out a big part of the strategy for success which are the actual changes to make in order to improve these metrics, but that’s supposed to be the fun part that’s left up to you!

As always, questions and comments are more than welcome.

AddThis Social Bookmark Button

Printer-friendly version Printer-friendly version

  1. 6 Responses to “Web Analytics and Optimization for Content Sites”

  2. When you say “set a threshold” is there a way to automatically set a target for that specific value?
    Great article!

    By Matt on Aug 21, 2007

  3. hey there
    very interesting article!

    You make some good points. However, from the point of view of someone concerned only with revenue (and an emphasis on short term revenue growth) your logic is a little flawed.

    Loyal visitors do not click ads.

    The ideal visitor, from a revenue point of view, arrives at your site and clicks the first ad they see. This type of user almost certainly comes from a search engine - and likely via a keyword you poached with clever SEO ;)

    Think about it from a CPM perspective. Users that stick around for > 5 page views are killing your ability to charge a high CPM.

    This points out the flaws in your example of 4 pages views per user (vs 2). Sure, you’d have more impressions but your CPM would be cut in half too.

    ;)

    Take digg for example, they are full of literate and loyal users. The users understand what is (and isn’t) an ad. I would be willing to wager they have one of the lowest click-through rates (or CPM) of any site in the top 2000 or so on Alexa.

    You say an ideal rate might be 2:1 for new vs returning users. I’d prefer to see 100:1 on my sites.

    By neil on Aug 21, 2007

  4. In response to Matt’s post, the thresholds I’m referring to in the above post actually have more to do with the type of content you have and not so much forecasting results (which is an entirely different post altogether!). Sorry if this doesn’t directly answer your question - but it does get my mind racing about thoughts for new posts so check back soon!

    The idea with the “threshold” is to set it once based on a discussion you and your team should have about the type of content you have on your site. Then, after maybe one or two quarters, re-examine that threshold and see if it still makes sense (but don’t just adjust it because you’re constantly under performing on that KPI!). The metrics above are definitely not to be considered a “final state” for your organization, just a good starting point!

    By mike.sukmanowsky on Aug 21, 2007

  5. In response to Neil’s post - you definitely raise some good points and any advertiser reading your post would probably nod their head in total agreement.

    Where my thought process comes from originally however is that I’m not designing and building a site with the objective of persuading my users to click ads. The way I see it, my objective as a site owner is to deliver the best content andd online experience I can possibly give my audience, and then allow a space for advertisers to promote their relevant / targeted products and services (notice the word “targeted” for all you behavioural marketers out there).

    I guess what I’m saying is that I see the “click” part of advertisements more a job of the actual advertiser, not of the site producer who I’m (at least attempting) to keep in mind with this blog.

    Thank you for the post though! I really REALLY appreciate the feedback and the open discussion!

    By mike.sukmanowsky on Aug 21, 2007

  6. Hi Mike,

    Congratulations on starting your blog. Wish you all the best.

    Nima

    By Nima on Aug 21, 2007

  7. In response to Nima’s post - thanks Nima! All the best to you too. FYI I’ve added your site to my blog roll :).

    Mike

    By mike.sukmanowsky on Aug 21, 2007

Post a Comment