AnalyticsGraphic

On Wednesday, July 27, Public Policy Communicators NYC members met to focus on the topic “Making the Most of Google Analytics.”  The session was billed as an opportunity to go beyond the basic mechanics of using this essential tool to how communicators can utilize its functions to better understand target audiences and the messages that appeal to them.  We definitely went beyond Analytics 101 to the next stage of learning, and many came away saying they had lots of new information with which to experiment.

Our discussion leader Greg Olson, an AdWords account manager at Google, working out of the Cambridge office and a member of the Google Analytics specialist team that trains and supports this office, was joined by Jai Sen, a digital media strategy consultant who has worked with several PPC-NYC members in developing their online strategies, to extend the scope of the conversation.

Olson started with a brief overview to make sure we were all on the same page.  In a show of hands, almost all in the room were using Google Analytics already, almost everyone in the room raised their hands to indicate that they were doing the analytics themselves (as opposed to being fed reports by I.T. staff), but only a small number had gone beyond the basic reports of number of page views for each URL on their site.

He said the Google Analytics platform can be used to:

-       Measure the engagement of users on the site

-       Make marketing campaigns more effective and accountable

-       Help determine if the content you’re creating is effective

-       Develop better awareness of when and why visitors are entering and leaving your site

-       Use Google Analytics to compare response to different versions of content, for example sending two different versions of an email and measuring which one got a better response

-       Improve audience interaction with your organization on the site

He started by encouraging nonprofits and foundations to take a step back and focus first on their organizations’ objectives and really question how the website is supposed to support the mission of the organization.  He used the example of PPC member Connecticut Health Foundation, which has the mission of improving the health status of everyone in the state by focusing on children’s mental health, health policy and advocacy, oral health and racial and ethnic health disparities.  CT Health Foundation’s Jenn Whinnem added that their primary audiences are grant seekers and health policy professionals in the state who work on those issues.  She said that her organization was happy with the numbers of grant seekers coming to the site and the information they were accessing, but wanted to bring more policy information seekers to the site and extend their time on site.  Olson said that there were many tools associated with Google Analytics that could help in better understanding how they are doing in reaching that objective and how effective the organization’s communications are in contributing to it.

Sen agreed with Olson’s point that the most important thing to do is make sure there is a clear discussion of overall communications strategy in order to get to a good plan for measurement. Strategy and measurement should function as a loop, with measurement leading to optimization and vice versa. An analytics snapshot can give an organization a sense of what users are actually consuming, providing insight into where the focus should be on content creation.

Olson said that an important step in that process is to really set key performance indicators for the organization’s objectives. Organizations should ask themselves what would be an actual measure of success.  For example with CT Health Foundation, what is the size and scope of the health policy professional community and what numbers or types of those coming to the site would constitute progress? And, since the organization is already happy with the number of grant applicants, the focus of the Google Analytics work would focus on URLs that were NOT associated with the  /grants/how-to-apply/ subdirectory where grant application information resides.

The next thing to do, Olson said, was to set defined targets, within a concrete timeframe, for example “to increase page views from ‘information seeker’ visits by X% in the third quarter of this year.” Once your organization has gained clarity about the types and number of visitors you aim to bring to your site and what you want them to do there, you can much better utilize Google Analytics to guide your work.

There are four basic categories of questions about visitors and related analytics measures to apply, he said. They are:

  1. Where do they come from? — Traffic sources
  2. Where do they spend their time? — Content reports
  3. Are they taking the right actions? — Goal tracking
  4. Is visitor behavior changing? — Alerts and intelligence

Olson showed two different reports that spoke to the “where do they come from?” question. The first showed an “All Traffic” report, the one which most of those just getting started with Google Analytics are familiar. This report shows the sources from which visitors came to a given page on your site. In the example shown, the most number of visitors came to a home page from Google, from a specific blog, from YouTube and direct (no referral – visitors may have typed URL or had it bookmarked, etc.). The “All Traffic” report gives you a sense of how visitors come to your site and, said Olson, may alert you to sources you may not have known about. For example, maybe a particular blogger has written about your work, which produced many visits. Perhaps that blogger should receive more attention from your organization and might be a good partner in communications. The second report he showed provided a different take on “where do they come from?” and might be less familiar to Analytics users. It is the “Location” report. The report Olson showed displayed a map of the United States with the states color-coded by number of visitors to the site. He said this report can be drilled down to the city level, or you can enlarge it to the international level. He said the location report can be especially useful in understanding where the people are who are really interested in what you have to say. You might find, for example, that an unusually large number of visitors are located in an unexpected small corner of the state, which might prompt some important questions about why that is.

Olson and the group next spent a good chunk of time exploring a tool called the Google Analytics URL Builder.  This is a tool that you can use to get a better handle on how effective certain planned communications mechanisms are in driving people to your site and – importantly and different from other tools – how they proceed through the site once there. The URL builder works by adding additional data (called a “query string”) to the end of a normal link that is tracked specially by Google Analytics. He showed how you can use the tool to create a specific URL, say for an eNewsletter, to see if that communication tool gets people to click on a particular story and then moves them toward donating, clicking a link to contact a congressmen, reading other reports or whatever actions you would ideally like your visitors to take. The URL created through the URL Builder allows the Google Analytics system to track how many visitors come to the site through specific links and associate those visitors with all of the other types of information reports.

He then moved on to the topic of “Where do they spend their time?” and content reports.  He showed an Analytics report of a landing page that displayed for each source (what previous site the visitor came from) the number of visits, the number of bounces and the bounce rate (see here for a discussion of bounce rate versus exit rate). This report helps you better understand what sources tend to bring you visitors that immediately leave the site or proceed through to take the actions you desire. Because this report also shows the bounce rate over a period that you specify, you can better understand if certain communications you sent out produced visits that resulted in immediate bounces or longer stays with concrete actions on the site. Sen and Olson were careful to mention that, for specific types of pages, a high bounce rate is not necessarily a bad thing—some sites, for example, surface essential content on the home page, so if the user enters and leaves on that page, it does not necessarily mean that the page was not “successful.” The time the user spends on a single page, however, cannot be measured by Google Analytics (which calculates the time spent on a page by the next click on that page), so the “time on page” metric is complicated by a high bounce rate.

Next, Olson looked at a report that could be immensely helpful in answering, “Where are they spending their time?” For a given page on the site (say, a blog post), the report displayed the total number of page views, unique page views, average time on the page, bounce rate and exit rate. It also displayed the page views over a given period and the bounce rate for various sources from which the visitors came. He then showed a different way of answering the “Where are they spending their time?” question with an Analytics report on search terms.  This report showed what visitors are looking for on the site. It showed the total number of searches, the percentage of times visitors had to refine their search, the time they spent on the site after the search, and the depth of their action after the search.  The report also shows the top search terms, the number of times those terms were searched, the percentage of those who left the site immediately after the search, etc. Olson noted that the site’s search mechanism does not have to be Google-based to be able to do run this analysis.

To get at the question “Are they taking the right actions?” Olson strongly encouraged users to take advantage of the “Goals” function in Google Analytics. A “goal” is a website page that serves as conversions for your site. Some examples of good conversion goals are:

  • A ‘thank you’ page after a user has submitted information through a form. This can track newsletter signups, donation receipts, email list subscriptions, job application forms, or contact forms.
  • An ‘About us’ page
  • A particular report, press release, news article or blog post.
  • A page that leads visitor to a “contact your congressman” function

(With some extra code, goals can even be file downloads or on-page actions).

Using the goals function allows you to track how far visitors move toward the actions you want them to take on your site – and let’s you know a lot about the characteristics of those visitors in terms of where they came from, how they proceeded through the site, etc.  Goals do not need to be static, in fact, they should change over time.

He then showed an example of a Goal Overview Report that displayed the goal conversions over a given period, the total number of goal conversions, the value of those conversions (for example the total dollar value of tickets sold or donations made), the goal conversion rate and the abandonment rate.  This report also showed the top sources from which visitors came and the total conversions and conversion rates for each.

Olson then moved to the question of “How do I identify traffic trends on my site?” He showed a “Daily Events” report. Event tracking is a relatively new (2009) function in Google Analytics.  Any significant marketing event should be tracked so you can identify correlating traffic trends. You might do this with, say, a new press release, a newly launched campaign, the beginning of utilizing online advertising, an important eNewsletter or a major capital campaign. You may have these dates noted elsewhere, but Google Analytics annotations makes it quick and easy to see how communications events overlap trends in your site metrics. You can even create annotations of key external condition. For example, a major debate in the media about your issue or a disaster in your service area. These external events could be just as important to explaining trends over time as your own marketing and web efforts.

In summarizing, Olson said that making all of these tools useful hinges upon proper tracking, advanced segmentation and micro-conversions.  He showed the “advanced segments” tab in Google Analytics and talked briefly about the choices you can make to designate key groups and what you want them to do, which can be a big help in better understanding how those groups are using your site.

A discussion of filters followed. It was noted that you might want, for example, to look at site traffic without employees’ usage of the site.  This is best accomplished using the filter feature. Click this link to learn more about using filters in Google Analytics. Olson also briefly touched on the difference between filters and advanced segments in Google Analytics. Advanced Segments provides a way for historical data to be grouped by your criteria to investigate differences between groups. Filters in Google Analytics, on the other hand, provide a way to process the raw data coming into your account before you even see it. This can be used to look at site traffic without employees’ usage. Filters are applied at the profile level. The data cannot be re-processed, and any information filtered out in this way cannot be recovered, and so it is important to maintain one unfiltered ‘master’ profile for your site.

You can also use Google analytics to better understand what types of platform users are accessing your site and what they’re doing on it. For example, are people who come to the site through iPhones and Blackberry phones exiting the site immediately because they have a hard time viewing it? Click on this link to learn more about what can be learned about various mobile platforms through Google Analytics.

Sen pointed out that Google Analytics is the best analytics tool out there and recommends it to all clients. However, because of the way it works, it does have one important limitation that he wanted the group to be aware of. Because Google Analytics tracking code is contained in web pages, it does not track downloads of files, for example PDFs. Properly configured, it can track clicks to download links or buttons, but often PDFs are linked from outside sites, and Google Analytics would not capture this data. Sen recommends using Google Analytics in tandem with a logs analysis tool (such as Webalizer or AWStats) that analyze the server’s record of downloads. Sen and Olson cautioned that logs analysis tools yield much less accurate metrics on all other data (like vista, visitors, and page views), so Google Analytics is the better tool for all metrics other than downloads.

Lastly, Olson made several recommendations for links to more information on various aspects of Google Analytics usage.  These included:

For me, one key take-away from the session was that getting your organization deeper into analytics can be a very productive way of clarifying what exactly you want your target audiences to do on your site and how your site should really function to move those visitors to specific actions that support your organizational mission.  In the past, many of us have just settled for basic numbers about how many site visitors we had and couldn’t really, or didn’t bother to, ask fundamental questions about how our constituencies were actually using the site.  Optimal utilization of Google Analytics will force your organization to confront challenging questions about what its priorities are for how visitors should move through the site and what should be the end result.