Mission: Deconstruction

Now I've come to the hardest task in Web analytics: the actual analyzing part.

I'm most interested in finding out who is visiting my site in an attempt to discover how to bring them back again to consume my wonderful content. My main driver of traffic to this blog was a post on June 14 that I wrote on Facebook asking family and friends to check out my blog.

Since that time, I've had eight unique visitors who have made 13 visits with 43 pageviews. You can see from the chart below that they mostly occurred in one spike. The day I posted the blog link on Facebook is the day I got the most traffic of seven visits. I was curious as to who these folks were so I could reward these most loyal with high-priced Christmas presents.

Location

The most visits to the blog came from Morgantown, followed by Broward County, Fla., where one of my friends works. You might find the Moscow entry interesting. I know it to be from my one and only friend in Moscow, Idaho, who has clearly visited the site twice. In a conversation with him (market research, if you will), I discovered that he tried to comment on the site but was unsuccessful. I view this as customer feedback on the technical details of my site.

Then I got one visit from Lake Worth, Fla., where my brother attends college and one from West Palm Beach, Fla., where I grew up. This last entry should clearly be labeled “Mom.” Luckily I always get her something nice for Mother's Day.

I would add a caveat that I don't think location is always an important metric to check. But since the blog has had so few visitors and the locations vary so much, the metric works in this case. For my company's news site, the location data typically shows that most visits come from Morgantown. It's great that the city's population is so engaged with the site, but this isn't great for attracting new students from other locales. It's similar to the new versus returning visitors debacle. You want growth in each, but you can't merely attract returning visitors (Kaushik, 2010).

Use Dimensions

But I'd clearly want to know who the people in Morgantown are. I know a lot of folks in Morgantown, and I really would like to know who is deserving of better Christmas gifts.

So I went to Referral Traffic under the Traffic Sources tab. I did get most of my referrals from Facebook, but I also got three referrals from blogger.com and all of those visits were from Morgantown. I'm positing that these visits could be from friends who have blogs or a classmate who lives in Morgantown. I found out that these visits were from Morgantown by selecting City as a secondary dimension.

For non-ecommerce sites

Ever since I read Avinash Kaushik's suggestions on key metrics for sites that do not have an e-commerce component, I've been eager to try them (Kaushik, 2010).

Frequency and recency are both discussed in Kaushik's book, yet they must have displayed differently when he wrote it a couple of years ago (Kaushik, 2010). This spring, he wrote a blog post on the usefulness of analyzing the now combined frequency and recency reports.

Count of Visits

Kaushik likens visits to relationships, calling single visits one-night stands (Kaushik, 2012). From the Count of Visits report below it appears that I had 13 visits with 43 pageviews. I ran the report again with an advanced segment of returning visitors and had five visits from returning visitors. One of these included 16 pageviews. These are the visits I'd like to keep an eye on.

Days Since Last Visit

The Frequency and Recency tab also has a Days Since Last Visit segment to measure recency. There's nine visits today and two visits the day before. When you compare it with the following Returning Visitors segment, you'll see that a lot of these visitors have been recent but have been one timers. This isn't what I'm going for, but I know that I need to let my visitors know about my new posts to get them to keep coming back.

Visit Duration

In the Visit Duration tab, Google Analytics is showing me that most of my visits are lasting fewer than 11 seconds. I have had two visits that have lasted more than three minutes. To find out how to replicate the success I've had with these visitors, I should look deeper into where they came from and where they looked for so many minutes. I'll go into how to do this in a later post. I will add a caveat that Kaushik mentioned (2010). Because this is a blog, it will probably get a lot of one page visits. Analytics systems have no way to measure the length of visits that only last one page, so sadly no matter how long they are, they don't count for visit duration.

Page Depth

Most site visits included two pages. Only three visits were to more than two pages going up until 16 pages. I'm not sure that this statistic is going to change much because it is a blog, but for other sites, you'd want that number to be a lot higher (Kaushik, 2010).

Bounce rates

I was surprised to see that my site has a zero-percent bounce rate. Checking in with my Kaushik text, I remember that blogging sites typically have high bounce rates because visitors usually go to a blog site and then leave after they've read an update thereby showing many visits consisting of one page only in the analytics report (Kaushik, 2010). Maybe my visitors are just more thorough because they're friends and relatives.

What should I do differently?

The question you have to ask yourself is whether you should change anything based on your newfound analytics knowledge.

Since I have a blog, I want to increase readers, but I also want highly engaged readers; those who keep coming back, offer comments and get something out of my posts.

My blog is new, but I can still work on some preliminary ideas. I want my new visitors to return, so I have to make it easy for them. I'll encourage them to subscribe to my blog via an RSS feed. I'll also have to create content regularly enough to keep their interest. And I'll have to work on making my blog more interesting.

To that end, I've included screenshots of the Google Analytics reports I've referenced. Finding ways to use multimedia clips where necessary will also work. I'll work to keep a conversational style, and I'll work on directing my blog toward people who would find it interesting. My Facebook friends are mostly not marketing folks. But my classmates and workmates are mostly marketers. So I can try to link to as many of their blogs as possible.

An additional metric to think about is keywords.

I didn't run any reports on keywords because no one came to my blog by searching for it. But I could change that by making sure I'm including relevant words in my posts so that if anyone were searching for information on my blog, they'd be likely to find it.

On the Traffic Sources lesson in the Google Analytics IQ lessons, there's a helpful segment on correlating keyword searches with landing pages.

(Google Analytics, 2011)

References

Kaushik, A. (January 30, 2012). Google Analytics tutorial: 8 valuable tips to hustle with data! Retrieved from: http://www.kaushik.net/avinash/google-analytics-tutorial-8-valuable-tips-to-hustle-with-data/#rfm

Kaushik, A. (2010). Web Analytics 2.0: The art of online accountability & science of customer centricity. Indianapolis, Indiana: Wiley Publishing, Inc.

Google Analytics. (2011). Traffic sources. Retrieved from: http://www.google.com/analytics/iq.html

Facebook vs. Google

This seems like a really contentious week for my blog. There was content versus conversation and now it's Facebook versus Google advertising.

At the basic level, Facebook's and Google's advertising systems both offer the same thing. They allow you to send consumers to your website or campaign from a section of the online marketplace that a lot of people frequent. It's a lot like putting up a sign at the mall. If you check out this video on Facebook's ad formation, you can see that this tool aims to target users effectively (Facebook Tutorial, 2010).

Google's ad creation tool is also targeted (Google Business, 2009). You're not pulling from all the people who search Google who like Elton John or ride their bike to work like you can on Facebook (Facebook Tutorial, 2010). However, not everybody lists their interests on Facebook. I don't. Facebook has no indication that I am obsessed with Doctor Who, a BBC TV show, and that I would click on anything with David Tennant's face on it. Facebook only knows what you tell it, and if you're careful like me or you just don't spend enough time cataloguing all of your interests, you're not telling it very much.

Both are so easy to use that you could probably turn the sales process into a game for elementary students. I feel like I could purchase advertising on both sites in less than an hour, and I'm not really fast.

Google AdWords' targeting depends on choosing the right keywords. You have an idea of how well a keyword works through AdWords' tool. With Google you also have the option of driving traffic for free. This handy, dandy search engine optimization guide provided by Google should help you either way. It helps you diagram search pages and gives advice about page titles, summaries and URLs. It will also suggest the best ways to create your site navigation, the best way to use photos, and how to notify Google that your site is mobile.

The biggest thing you probably need to know though about these tools isn't regarding functionality. You might want to know about the return on investment you'll get. Facebook's advertising bluff is being called. And, as the advertiser, it's good for you to know where you stand. Google ads are clicked on 10 times more than Facebook ads (Haberman, 2012). Facebook ranks lower in targeting options and ad formats offered. Google's click-through rate is higher, and Facebook doesn't even publish its ad click-through rate. The data provided in this article comes from Webtrends' independent analysis. Facebook's advertising rates rose 40 percent in the first quarter of 2012 while its click-through rates fell by 8 percent.

In another report, four out of five Facebook users say they have not bought a product or service because of advertising or comments on Facebook (Jones, 2012). Before Facebook went public last month, General Motors pulled its Facebook advertising in favor of only using the site for free promotional efforts through Facebook's Pages. However, GM's competitor Ford has said it will increase its Facebook advertising budget.

ComScore argues that the poll data from the Facebook user survey is inaccurate because people have weak memories (Booton, 2012). This argument seems flimsy. If you read this article by Booton and contrast it with the data collected on Facebook's current strength versus that of Google's in advertising, I think it's clear where you ought to spend your money.

References

Booton, J. (June 8, 2012). ComScore: Stop bashing Facebook ads. Retrieved from: http://www.foxbusiness.com/technology/2012/06/08/comscore-warns-people-to-stop-judging-facebook-ads/

Facebook Tutorial. (April 5, 2010). Facebook Tutorials:How to advertise on Facebook. Retrieved from: http://www.youtube.com/watch?v=8jOBDIql4yc

Google. (2010). Search engine optimization starter guide. Retrieved from: http://www.google.com/webmasters/docs/search-engine-optimization-starter-guide.pdf

Google Business. (June 15, 2009). Getting started with Google AdWords. Retrieved from: http://www.youtube.com/watch?v=tx2L6EGa9DY

Haberman, S. (May 15, 2012). Can Facebook ads ever beat Google? Retrieved from: http://mashable.com/2012/05/15/facebook-google-advertising/

Jones, R. (June 5, 2012). Facebook's problem: Ads aren't grabbing users, analysts say. Retrieved from: http://marketday.msnbc.msn.com/_news/2012/06/05/12069407-facebooks-problem-ads-arent-grabbing-users-analysts-say?lite

Content, the king for now

Making an argument that content is in opposition to conversation in the marketing industry, as discussed in Novak (2010), is silly and narrow-minded. It's a tactic used by politicians when they want you to ignore their stupidity and focus on their opponents' dalliances. It takes the focus away from the problem. Why would you denigrate the content that you need to get any message across and set it in opposition against the marketing that could make it succeed? It's not constructive, and it creates an incomplete picture of the problem.

As soon as I read the Novak and Greenberg articles (2009), my mind went immediately to several projects I've worked on at WVU in which the partnership between content and marketing was shaky. Take this incredible truer-than-life story on a speech pathology professor who grew up with a stutter, overcame his disability, and went on to train future pathologists and counsel hundreds of patients throughout West Virginia for free (Mazzella, 2012). I'll admit that I wrote the story for our news outlet, WVUToday. But it's easy to make a nice story when you have such great material.

When I looked at the resulting conversation after the story's release on April 17, I felt it could have gone to a wider audience and generated more chatter. It was one of the five most popular stories on WVUToday before the onslaught of graduation coverage in the following weeks. The Associated Press wrote about the story after we sent it to media. That's not bad for the cost of paying employees to write, make photographs and post the story.

But consider that this isn't your average kind of story about a dean being named to head an association or a student who got really, really good grades and is saving a small town's water supply in South America. That's what every college writes about. They all think their perky run-of-the-mill stories with exclamation marks are going to sell. But this is a story that we go to great lengths to find and promote. It's one that people can relate to, and it could do a lot in terms of forming awareness and good will. I can't find an entry for the story on our main WVU Facebook page. That would have been a great way to share the story among alumni and others who are actively seeking out information on the university.

A challenge in our marketing efforts is that there are different people doing different tasks in marketing each story. Some of the efforts are handled in my office. Others, like Facebook and websites beyond WVUToday are updated by those in other offices. Those decisions are made by other bosses who may not have the same marketing priorities.

This is a key frustration that I'm sure is not limited to marketing. Yes, I could ask more people to market the story in an even nicer way next time. But there is a way to capitalize on the dispersion of marketing tasks (especially social media marketing) among various people in your organization. A Ragan.com article suggests that social media marketing is everybody's job. This goes beyond those whose task every day is to market with social media. This is actually a really good thing for my office because we have maybe one or two people whose sole task is posting to social media.

A good way to look at this strategy is to consider this blog post. I brought up a great story you'll remember that you found unintentionally just by reading my post.

I've put the content versus conversation in a more accurate light, defining the problem as not one of content versus conversation but one of using your content effectively. You need both. People will find out quickly if your content is superficial, but that's another story.

References

Hanelly, A. (May 15, 2012). Why it's time to get rid of your social media guy. Retrieved from: http://www.ragan.com/Main/Articles/Why_its_time_to_get_rid_of_your_social_media_guy_44882.aspx

Mazzella, D. (April 17, 2012). At WVU, one man fights for words, hope. Retrieved from: http://wvutoday.wvu.edu/n/2012/04/17/at-wvu-one-man-fights-for-words-hope

Tidying the Web toolbox

I will mention Google Analytics a lot on this blog because it's mentioned a lot in the industry, and it's free, so it's a good place to start in evaluating how well you are speaking to your online audience. But there is more than traditional analytics software out there to measure the impact your content and your efforts are having on the great big world of people out there.

Analytics can show you a lot about your site, but there are other indicators, particularly when it comes to optimizing search keywords that you can find when using a different Google product: Webmaster Tools (also free). (Kaushik, 2010) Here is a great landing page for Webmaster Tools and other webmaster assistance on Google. Using this tool, you can compare the number of impressions (when your listing is viewed) versus the number of clicks you receive on your listing in search tools. This is where you cross over into search engine optimization. You are analyzing the performance of your site, but you're also making sure that your site is easily found by the people looking for what you offer or those who will hopefully be pleasantly surprised. Once you see which keywords are working and which are not working, you can refine pertinent portions of the content on your pages and keyword text around your videos.

You may find it helpful that some Webmaster Tools data is incorporated in the most recent version of Google Analytics. (Gerson, 2011) Through this tool, you'll be able to tell which kinds of Google searches are bringing you the most traffic, whether that's Web, Image or Mobile search. It shows you impressions, clicks, average position and click through rates for each of the above categories. This service is limited to one portion of Webmaster Tools, so you'll get the deepest information from looking at Webmaster Tools itself.

Google Analytics will also incorporate social engagement with its new social reports. (Gerson, 2011). Using this tool, you can see how many site visitors recommended or otherwise engaged socially with your content on Facebook, Twitter and Google+. This of course will not take care of all of your social insights but it does add another element of your metrics into Google Analytics.

Of course, there is still the paid element of your online presence to consider. Kaushik, and it would seem all of those in the know, recommend that you engage in paid search campaigns (paying for search placement at the top or side of a search page) as well as organic search campaigns (optimizing keywords for free). (2010) Google Analytics and Google's AdWords service (purchasing ads will cost you) do separate your organic and paid clicks in search engine results. However this is not the case with other analytics software and it will not affect paid search campaigns outside of Google. So if you bought paid search ads in Microsoft and Yahoo!, you'll have to tag them and can use Google's free URL builder to do that.

Yes, it is true that all of these services I mentioned come from Google. Though the company has its hands in what seems like every cookie jar, that doesn't mean you have to use its tools. But if you did, these are some ways they converge. And even just using Google as the example, this analysis has shown that often analytics services won't see everything. In this case, you still have to account for social media, paid search campaigns and some aspects of organic search campaigns.

The social media information provided through Google Analytics is just at the surface of what you should be discovering about your social campaigns, but this is a start. I'm sure we'll be digging into more social analytics down the road.

As a taste of social media analytics, check out these links: A new analytics tool just for Pinterest, called Pinerly, is being released to latch on to how and why this platform has become popular and how it can benefit your business. Older platforms like Facebook are also continuing to add to their social media analytics arsenal of tools, such as in this new Facebook ads tool.

References

Gerson, R. (November 30, 2011) 10 new Google Analytics features you need to start using. Retrieved from: http://mashable.com/2011/11/30/google-analytics-new-features/

Kaushik, A. (2010) Web analytics 2.0: the art of online accountability and the science of customer centricity. Indianapolis: Wiley Publishing, Inc.

Seeing the light

Researching Web analytics for a news site and a definitely non-ecommerce one at that can be a bit like always being the “different” kid at school. When I first started learning about the topic I felt like I had to cover my ears whenever the discussion turned to conversion (when a visitor to your site makes a purchase or completes a task). (Web Analytics Association, 2008) I knew in general that conversion applies to any website, yet my previous classes typically put forward an ecommerce site such as Best Buy or Amazon where the customer researches something they want to buy and—hopefully for the company—buys the product and continues to return and buy even more products.

These basics of Web analytics made me wish I had a commercial product to sell to make conversion calculation easier. But if you look at the situation from a different angle, it's a similar process for non-ecommerce. And you really have to see it this way because without looking at your revenue or the value created from your online presence, there's A. no demonstrated point in having a website and B. no way of correctly estimating what the site is achieving.

The first real, cogent process I've seen for measuring value generated from a non-ecommerce site was in Google guru Avinash Kaushik's book “Web Analytics 2.0.” (2010) Typical benchmarks such as page visits won't really tell you much that's valuable, he says. You should be looking at visitor loyalty, visitor recency, length of visit and depth of visit, all metrics that can be found in the free Google Analytics reporting tool.

1. Visitor loyalty analysis measures how often your unique visitors come to your site. (Kaushik, 2010) But more than that, it will allow you to segment your visiting audience and concentrate on those who keep returning to your site. You can find out how many people really are your friends, where they're coming from and what they're seeing on your site.

2. Visitor recency shows you how many visitors have come to your site in a period of time by measuring the time between visits. (Kaushik, 2010) It would be ideal if they visit you multiple times in a week, for example, and even if they aren't, you have a way to track how your efforts are increasing or decreasing the frequency of visits.

3. The length of visit doesn't tell you the average time on site, a traditional method of measuring Web traffic through analytics software such as Google Analytics, but it will tell you how many visitors are on your site for a given period of time. (Kaushik, 2010) Through this tool you can discover that key group of people who stay on your site for more than a minute.

4. The depth of visit will tell you the number of pages in each visit to your website, not an average number of pages. (Kaushik, 2010)

Once tied to your business objectives, these metrics can take you far. (Kaushik, 2010) You can better measure conversions if you know important visitor activity such as the how often they visit, how long they are on your site, how many pages they visit, how frequently and recently they visit and how loyal they are to you. These tools allow you to track real progress for a site that lives off brand awareness and the spread of knowledge, your kind of sales.

You don't have to read the book to dig into these metrics, but can find them on Kaushik's blog, Occam's Razor in this article.

The last step in this equation is to assign values to actions taken on your site. (Kaushik suggests making use of offline values assigned to actions such as giving a brochure download on the site the same economic value as what it costs you to send a brochure in the real world or acquire an e-mail list. (2010) Google suggests two ways of doing this: assigning values to your goals as Kaushik demonstrates, or set up ecommerce tracking in Google Analytics. (Clifton) Setting up the tracking may help you in making more accurate counts of items that convert customers to your goals.

References

Clifton, B. (n.d.) Measuring non-ecommerce site value. Retrieved from: http://support.google.com/analytics/bin/answer.py?hl=en&answer=1011367

Kaushik, A. (2010) Web analytics 2.0: the art of online accountability and the science of customer centricity. Indianapolis: Wiley Publishing, Inc.

Web Analytics Association. (September 22, 2008). Web analytics definitions. Retrieved from: http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0CK8BEBYwAQ&url=http%3A%2F%2Fwww.webanalyticsassociation.org%2Fresource%2Fresmgr%2FPDF_standards%2FWebAnalyticsDefinitions.pdf&ei=J6nLT97JE66L0QGI6rmkAQ&usg=AFQjCNFP1XPpBbnyL03FKfksg7A87siKIg&sig2=ap3gX0XDbyx-Cwj3G442Yg