Web Analytics Management

Metrics for Smart Content Management - Metric #6- Task Completion ROI

This is the last post in my series on metrics for evaluating the effectiveness of web site content. Over the last five weeks I’ve presented five metrics that you can use to benchmark site performance basaed on content and navigation effectiveness. Today's metric, Task Completion ROI, is built on the first metric I presented: Task Completion Effectiveness (TCE). It is actually pretty easy to calculate, and it is a great way to introduce concepts of value to content in organizations where this connection isn't seen clearly, such as government, healthcare, non-profits or corporate intranets.

Task Completion ROI: Measures ROI in Web site operations, content, applications, and labor relative to the budget spent on these initiatives.

Calculation of Task-Completion ROI: Number of task completions multiplied by a dollar value given to each task-completion type compared with the total spent on creating the task application or with the avoided-cost alternative to offline.

Calculation Note: Total expenditures will need to be determined based on budget allocation, set-aside funds, hourly labor rates, etc. Dollar values for completion of tasks will need to be assigned based on the calculated cost of completing the task offline.  Find out the costs for developing the application, the marketing program, the event, the video development, and then estimate a value for the completed task based on a time and materials calculation if the task had to be completed offline.  Then add a “goodwill” value.

Goodwill, according to the Merriam-Webster definition is “the favor or advantage that a business has acquired especially through its brands and its good reputation; the value of projected earnings increases of a business especially as part of its purchase price; the excess of the purchase price of a company over its book value which represents the value of goodwill as an intangible asset for accounting purposes.” So, you’re adding in “goodwill” over the costs. Sound like a stretch? Maybe a bit…think of this as a conversation starter or straw man for valuing content that may not have a clear dollar value. Even better, find out from the sponsors of the application what their business and revenue goals are before it ends up on the web site.

Example: There were 130 submissions of grant applications during the month.  Each submission is valued at $200. This amount is based on the time saved by having the applicant enter the data directly into the database, rather than having staff transcribe the data from printed forms. This results in avoided cost or savings of $26,000. 

Another Example: Calculate and compare the cost and success rates associated with promoting research studies and grants in the offline channels and the online channels to determine which one was most effective at attracting qualified applicants.

So, that about does it for this series. If you come up with other metrics for evaluating content sites, let me know. I think this is an area that could use some additional development.

________________________

And lastly, hope to see you at eMetrics Toronto at Analytics in Real Life on Wed, 3/20 with Anuschka Van Dijke from KPMG and Nancy Duley, from American Society of Clinical Oncologists.

 

 

 

Metrics for Smart Content Management - Metric #5 - Navigation to Content Score (NCS)

NCS is a quick way to see if your visitors are spending too much time trying to navigate to content. The fewer pages it takes a visitor to content, or the more direct path, the better. NCS complements the Content Engagement Score (CES) which lets you know whether visitors are using the most important content on your site.

Navigation-to-Content Score: Compares number of navigation page views to all site page views.

Calculation: Percentage of navigation-type pages viewed compared with all content page views. This is derived from dividing navigation page views by all site page views. A lower percentage indicates a more efficient visitor interaction with the site.

Calculation Note: As with the other metrics in this series, you need to identify navigation pages and measure them separately. Pages include: 
o    home pages
o    portal pages
o    search engines
o    site maps
o    pages provided for navigational purposes

Example: There are 40,000 page views of navigation pages out of 150,000 total page views for the month. This equals a navigation-to-content score of 27%. Here’s how a report might look:

You can make this a much more valuable report by running this for different segments of visitors based on geolocation, domain, first time vs. repeat, as well as adding visit segmentation attributes such as acquisition source. If you use this with the segmentation, you’ll be able to see which visitor groups are having an easier time finding site content.
________________________

If you're going to eMetrics Toronto, be sure to stop by Analytics in Real Life on Wed, 3/20. I'll be moderating a panel with Anuschka Van Dijke from KPMG and Nancy Duley, from American Society of Clinical Oncologists. They'll be sharing their stories on how to get resistant management to come around to the value of analytics. Hope to see you there...

 

 

 

 

Metrics for Smart Content Management - Metric #4 - Content Distribution Index (CDI)

The Content Distribution Index (CDI) is another engagement measurement that you can use for benchmarking.  The CDI gives you a sense whether visitors believe content is valuable and interesting enough to share through Facebook, Twitter, email and other sharing links that you have on your site, such as AddThis.This is an easy to calculate metric and one that can help you rate overall relevancy of your site.

Content Distribution Index:  Measures whether content holds value and relevance based on visitor opinion.

Calculation: Content shares based on total number of sharing links usage divided by page views. A higher percentage of content shares indicates that content was distributed more successfully.

Calculation Note: Sharing events will need to be captured by tagging the article sharing links. As with the other metrics in this series, you’ll have more accurate results if you do not include navigation, search and index pages.

Example: During the month, there were 200,000 total page views and 12,000 article shares. This equals a content-distribution score of 6%.  Here’s how a report looked for one of our clients.
 

Taking a look at the results, you need to account for the number and type of articles that are incorporating the share function. A further analysis is needed to understand whether the article content is relevant and whether the position and presentation of the share link is clear to the visitors. Conduct this with a visitor/visit segment and you’ll be able to see which of your audiences is really engaging with content and which ones aren’t.

Metrics for Smart Content Management - Metric #3 - Content Engagement Score (CES)

In this series on metrics for assessing your site’s overall content health and content strategy, I’ve so far focused on focusing on valuable content as defined by Task Completion and Mission Critical Content Engagement. These metrics help you determine whether visitors are using the most important content in your site.

Today, let’s take a look at the other side of the coin…understanding content that appears to have little value; content that you would consider removing prior to a site redesign or a content migration. Let’s say you want to create a metric that will give you the opportunity to do a before and after comparison. To accomplish this, I want to introduce you to the Content Engagement Score (CES). Think of this as your site’s Body Mass Index (BMI). Once you have the CES, you can do more detailed analysis on why certain content may lead to more engagement.

We’re currently working with a client that is planning a redesign and consolidation of dozens of sites into one. They selected 3 current sites that were representative of all of the sites.  Using the cut-off of at least one page view over the course of 10 months, we focused on specific site sections.  After doing this analysis, we found that some sections had a Content Engagement Score as high as 100% and as low as 33%.

Content-Engagement Score: Measures the overall site-content engagement based on viewing of all site-content pages.

Calculation: Percentage of all content pages on the site or site section receiving at least a pre-determined page view compared to percentage of all pages receiving less than the pre-determined number of page views.

Calculation Notes: You have to be able to identify content pages and then decide on a number of page views that you’ll use as a cut off for what is considered as a minimally acceptable engagement. Using only 1 page view as a cut off may work for low traffic sites, but I’d recommend generally at least 10 page views as an initial cut off. Filtering out the following pages will create more accurate results:
o    home pages
o    portal pages
o    search engine results pages
o    site maps
o    pages provided for navigational purposes

Example: There are 25,000 content pages on a site; 20,000 have received at least 20 page views during the last quarter. This results in a CES of 80%.

Do an analysis of the 80% that are receiving higher than the minimum page views to understand what is driving visitors to that content and how this content influences Task Completion.

Metrics for Smart Content Management - Metric #2 - Mission Critical Content Effectiveness (MCCE)

Let’s return to the topic of content as being a measure of success. In my last post on Task Completion, I said that only sites that sell advertising can consider content viewing to be a goal. I admit that I overstated this a bit…if you’re a non-profit or government site, this could actually be pretty important. 
For non-profits, content that leads to donor contributions, program funding, volunteerism, and advocacy is highly important. Government agencies, branches and divisions have clearly stated missions against which they are funded.  I think of this as Mission Critical Content.

What is Mission Critical content? Content that directly or indirectly supports funding and revenue. For example, showing donors the readership of articles that they’ve funded.

Defining mission critical content can be a tricky exercise because you have to draw a line…or else all content can be viewed as critical. Here’s a few ways to go:

• Conduct an analysis to determine the pages viewed before visitors go to top tasks.  These pages are influential enough to bring a visitor to the task completion.
• Review the site to determine the pages that were designed to communicate mission. On a government site this could include: About Us, Our Mission, News, or posts from the head of the agency.
• Review the results of the top 100 pages and downloads on the site, cull out the non-content pages (home page, search, portals, index, and so forth). Review the resulting list to determine which pages are central to mission.

Once you figure out the content that is mission-critical, calculating the metric is pretty straightforward.

Mission-Critical Content Effectiveness (MCCE): Measures the number of times mission-critical content, including video, is viewed or files are downloaded.

Calculation: Expressed as a percentage based on total views or downloads of the top content pages of the site that most relate to organization mission, divided by the total number of all site content pages (do not include portal, search, or navigation-based pages). A higher percentage indicates a better MCCE rate.

Example: During the month, there were 25,000 total page views and downloads of mission-critical content and 150,000 total page views of site-content pages, including downloads. This equals a mission-critical-content effectiveness rate of 17%.
 

Based on the data, it would make sense that you could run this report for specific segments, such as visitors from targeted geolocation, organization type, content interest and branded keyword acquisition.  It would also be highly valuable to run an analysis showing the impact of mission critical content views on tasks such as donation, article sharing, membership and other top tasks.

Metrics for Smart Content Management - Metric #1 - Task Completion Effectiveness (TCE)

Management of content proliferation is important of course, but in evaluating content effectiveness it is also important to consider the goals of the site. All too often we hear that “reading content” is a site goal. While that may be true if you’re selling advertising, it is generally not true for content rich sites. (For more on this, you’ll want to read next week’s post on Mission Critical Content)The purpose of content is to drive an action or a task, such as a purchase, event registration, donation, newsletter sign up, lead generation form completion, using a selection wizard and so on.

Measuring Task Completion gives you an initial sense on whether the web site is effective in getting visitors to complete the tasks on your site that are associated with business goals. Once you establish Task Completion Effectiveness, you can use this as a benchmark to determine whether you content management actions are working. The goal: raise the site’s Task Completion Effectiveness percentage.

Task Completion Effectiveness (TCE): Measures completion of site tasks associated with your goals and mission

Calculation: Expressed as a percentage based on the total number of task-completion page views divided by total number of site page views. A higher percentage indicates more successful task completion.

Calculation Notes: The completion page for each task will need to be identified, such as a “thank you” page or a submission confirmation.

Example: During the month, there were 500,000 total page views. Of this total number, there were 10,000 task-completion page views. This equals a task-completion effectiveness of 2%.

In the chart below, you’ll notice that we’re using “visits” and not “page views” as a baseline metric. Why? This report was done in Google Analytics. GA bases its metrics on visits and can not correlate events to page views. While we could have created filters to obtain the page view data, we opted to go with what GA does best. Using visits as an alternative still provides a reliable measure of task completion effectiveness. If we had done this in SiteCatalyst, we would have been able to correlate the events to page views.

 

 

In this example, the TCE has gone down slightly over a 3 month trend. Next steps should include reviewing the page view totals of the task confirmation pages to see which ones are experiencing declines, and then determining what factors are influencing the drop. Is it because there has been more content added to the site, thereby “diluting” the completion page views? Less marketing and acquisition driving towards task completion? Or, perhaps tasks are seasonally driven? Like any good high level “health” report, the TCE will drive additional questions that require you to do some analysis and investigation.

As I mentioned in my first post, this metric becomes even more effective through segmentation of audiences by geolocation, subdomain, visit frequency, and acquisition source.

Metrics for Smarter Content Management

In your organization, what stakeholder group is most under-served by web analytics?

Certainly not marketers and product managers…they are an analysts’ best set of friends because the reporting can show real ROI.

The user experience designer or information architect? No, they get data that can really help with site optimization that brings visitors to the marketing and product focused site sections. 

I’m thinking it’s the content manager…the operations guy/gal whose job it is to manage the production, updating, archiving and retiring of content. Is there someone in this role where you work? Maybe they are a subject matter expert who has control over a particular site section. Maybe they manage a team of writers whose job it is to crank out the content. Maybe they are your intranet managers…a part of the site that generally flies beneath the radar when it comes to analytics. Maybe it’s the overall site manager. And maybe it’s you.

In much of my work with US Federal government agencies, financial services, pharmaceutical companies and non-profit organizations, the content managers and intranet managers tell me that they just don’t get metrics that help them determine when it’s time to clean up, spruce up or get rid of content that is old and in the way.  As one weary content manager told me, “The goal is to publish, publish, publish…not analyze whether anyone looks at the content or what they do with it.”

While I hear some content managers say with pride that they simply whack all the pages that don’t get page views, I don’t think this is the whole solution to the pruning problem.  I think it makes more sense to first get an understanding around relative site performance based on goals as a first step, and then take a more surgical approach that’s based on prioritization. This may indicate that certain areas of the site need more than just an elimination of pages and in fact may need a new content strategy. This seems like a more constructive approach.

Over the last few years, I’ve been working with clients who want to go this route.  I’ve developed six metrics that you can start using to evaluate site performance and direct your content management efforts.  I’ll be sharing them over the next six weeks in a series of posts that I’ll be publishing here. 

Let me say up front, that creating these metrics requires some prep work, like categorizing content pages vs. navigation pages. It's actually a good exercise...like doing an inventory. The goal: Do the analysis, create the metric and develop benchmarks that provide you with a good idea of overall site effectiveness that can be tracked and explained over time.

…But First a Word About Segmentation

While the metrics stand on their own, it makes sense to create a series of visitor/visit segments based on Semphonic’s two-tiered segmentation methodology. After an extensive review of site objectives, goals, top tasks and target audiences, you can create key segments. The metrics that you develop can be applied to these segments to understand whether each of these segments may use content that others don’t. This helps you understand immediately whether specific content types that do not get a high number of views, may in fact be content that is important to a highly valued segment. In the table below, we have some examples of segments developed for a site that reaches health care professionals, researchers and the general public.

As you can see from these categories, combining the known visitor data with site behavior gives you an opportunity to amplify meaning from the basic elements of data you’d be collecting.

OK…so categorizing pages and segmentation... two keys to making these metrics work. If you can do that over the next week, you’ll be on your way to applying the first metric I’ll be talking about next Tuesday…Task Completion Effectiveness. 

 

42 Rules for a Web Presence that Wins

42 Rules for a Web Presence that Wins by Philippa Gamse is a slender tome that gets right to the point on developing practical web site strategy.  Targeted primarily to the SMB market, I think there’s a lot in the book that larger organizations, as well as non-profit and government web site managers could very well be reminded of when it comes to web site management best practices. Specifically, as my paraphrasing of Philippa puts it:

• 95 percent of web presences leave money on the table, and usually the owner has no idea.
• Most websites fail to make the emotional connections with visitors who are crucial to creating and sustaining meaningful…and profitable…online relationships.

Having been working in the internet space since 1991, Philippa knows what she’s talking about and has a great historical perspective that transcends “flavor of the day” points of view that tend to proliferate in the online world.

What I like about Philippa’s point of view as a web strategist is that she is really focused on the bottom line and results, so she’s a natural fan of digital analytics, dedicating an entire section of her book to “Measuring Results.” Not only is she an advocate for using analytics, she continually and consistently ties it back to deriving intelligence that can be acted on. I can think of more than a few digital analytics managers who could use some of what’s here for ammunition to use for convincing stakeholders and senior leadership that there’s more to life than high level traffic reporting.

You can find out more about 42 Rules for a Web Presence that Wins or order the book from Philippa’s site Websites that Win International. I definitely think you should check it out.

So, will Web governance make your digital analytics program run faster, jump higher and make you smarter?

To conclude this year's Profiles in Analytics - Does Web Governance Drive Analytics, I wanted to leave you with some general takeaways on how to move towards a better Return On Investment on your program...regardless of whether your organization is driven by a solid governance model or not.

So, will Web governance make your digital analytics program run faster, jump higher and make you smarter?

Like any programmer will tell you “it depends.”

The Profiles in Analytics research indicates that enterprise-wide Web governance is generally a good thing. The operational processes, the roles and responsibilities, and the overall standards that come out of governance help get everyone in an organization on board with implementing strategy. This appears to help the institutionalization of digital analytics throughout the organization.

There does appear to be a gap between where enterprise-wide governance ends, and where digital analytics governance begins. Whereas Web governance creates a space for digital analytics to grow, the depth of influence and making real change is only beginning.  The dichotomy between the considered priorities of digital analytics teams and the time spent on tasks shows that there are constraints that detract from analytics reaching its full potential as a way for organizations to learn from their experience in the digital channel.

How to clear the constraints?

• Invest in a Web governance process that provides insight on where digital analytics should sit in the organization. The level of influence that digital analytics has within the organization is tied to the office or department in which it resides and the influence of an executive champion in fostering cooperation.
• Conduct an audit of digital analytics operations to understand priorities, roles, responsibilities, tasks and processes. Senior management and digital analytics management need to be on the same page. There needs to be frank and open discussions about constraints and roadblocks that prevent digital analytics in being an effective voice in providing insights throughout the organization.
• Facilitate cooperation between digital analytics team and potential partners within the organization where there are opportunities to complement and leverage resources. Digital analytics programs that succeed depend on a network of resource support beyond the direct team. IT, Procurement, Marketing, Content Management, Business Intelligence, Finance and Operations are among the groups that a Digital Analytics team must interface with and work with to operate effectively. It helps to have these relationships facilitated by senior leadership so it is clear to all that analytics are recognized as a management priority.

This year’s Profiles in Analytics research suggests that it isn’t enough for organizations to pronounce that they are “data driven.” There is a lot more work to do around overall Web governance, development of analytics management expertise, and figuring out the best place in the organization for digital analytics so that it can have real influence.

Interested in continuing the conversation? Email me at [email protected].

 

Has Digital Analytics Made it into the Digital Production Lifecycle?

Continuing on with Part 5 of this year's Profiles in Analytics - Does Web Governance Drive Analytics...or get the whole analysis, survey data and digital analytics profiles at Semphonic.

Has Digital Analytics Made it into the Digital Production Lifecycle?

For all of the aspirations on how to create digital cultures and data driven organizations, when it gets down to basics, simply making sure that data collection occurs is a back burner issue. While Web governance oriented organizations may have processes in place and top down commitment to ensure that analytics is part of the development cycle, that it is only at around 60 percent says that there is room for improvement.

While non-governance organizations have roughly the same compliance, that there is such a high level of dissatisfaction with the process of implementing code changes, tells you something about an underlying issue that can sabotage whether accurate data is available when senior leadership wants to use it for appraising the role of the digital channel.

The Details

It stands to reason that organizations with established Web governance have more formalized processes around production of new content and applications for the digital channel, than organizations that don’t have strong governance.

Web development process. Nearly twice the number of organizations with governance includes digital analytics in the process “all the time” versus those that don’t, as seen in the chart below. Still, at 25 percent, there is clearly more work that needs to be done in this area. Interestingly, once you combine “all” and “most” of the time responses, you get roughly 60 percent in all organizations.

The research shows that there is still a great deal of work that needs to be done in the area of integrating analytics to the digital development process. While there is an increasing amount of project management best practices aligned with enabling analytics, analytics itself is often implemented after a project has launched. With the advent of mobile applications and social media initiatives, there are new opportunities for using analytics, and at the same time a repeat of the poor practices that have left analytics out of the fixed Web development process.

Implementing data collection code changes. The research showed the biggest difference between governance and non-governance when dealing with the process for implementing changes to data collection code, as seen in the chart below.

Anecdotal research tell us that the application of data collection code, considered to be the least glamorous element of analytics, to be a universal and significant hold up in the analytics process.  Because there are often other projects in the queue ahead of analytics, or analytics is considered expendable if content or applications need to meet a deadline for launch, data collection coding may be held up for weeks and even months in certain organizations.  

The “not acceptable” level of satisfaction among organizations without governance is a situation that if addressed, could show immediate improvements to the availability of analytics and potential value of metrics created with the data.

Next week: So, will Web governance make your digital analytics program run faster, jump higher and make you smarter?

Get the entire research at Semphonic.