Breaking down your web traffic with Google Analytics

Transcript: 

 

Rebecca Matulka: Hi, all. This is Rebecca Matulka, Interior's Deputy Director of Digital. Thank you for joining us today to learn about Google Analytics.

We are lucky to have Tim Lowden with us, sharing all of his best practices and insights about Google Analytics. Tim manages GSA's Digital Analytics Program.

Tim has his presentation. Then we're going to save questions until the end. Feel free to start asking questions now, and then we have a portion at the end. There are handouts on the left. With that, I'm going to turn it over to Tim to kick off the program.

Tim Lowden: Great. Thank you, Rebecca. Thank you all for having me today to speak to you. This sounds like a pretty cool, little meeting that you all have. I was really surprised when I asked Rebecca, "How many people will be attending this?" She was like, "Oh, we'll definitely have over a hundred." I was like, "Oh, fun times." [laughs]

As Rebecca stated, my name's Tim Lowden. I'm the Program Manager for the Digital Analytics Program. Some of you may be familiar with what DAP is. I don't like to say Digital Analytics Program a million times. In true government fashion, you'll hear me affectionately refer to it by its acronym, DAP.

As I said, some of you may know what DAP is. If you do, I hope that this presentation will maybe make things a little bit more clear about what we do and how you can use our program.

Some of you may not know what we are. In that case, I really welcome the opportunity to try to explain a little bit about what the goal of our program is and how you can use it to really further your insights about how folks are interacting with your websites.

Firstly, the crux behind why DAP was created was questions like this. If I were to ask you "Which is the most-visited federal website across all of the federal government space?" some of you may know this off the top of your head. Some of you may be able to take some guesses at it.

In reality, this question and also just more broadly, what's the content that is being most viewed by the public across all of the federal digital sphere? Previously, we couldn't really answer that question. I'll explain a little bit more about what this means, but keep this in mind.

Here are some quick facts about a group of 5,000 federal websites. 47 percent of visits to federal websites come from mobile or tablet. Less than 45 percent of all visits are using Windows operating systems.

46 and a half percent of visits in the last 90 days are using Google Chrome as the browser. Last month, and here's the answer to the previous lad's question, tools.usps.com, which is where you can track your packages, had over 140 million page views on that site.

Those are things that prior to DAP, there was really no way of knowing an aggregate trend or number across many federal government websites. Let me actually just fast-forward to this slide so I can explain a little bit about why we can answer those questions.

Previously, before DAP was created, you can see my example here, site 1, site 2, site 3, site 400, all the way to 5,000. We have 5,000 participating sites at this point. Every site would have, if it had a Web analytics tool, it would have its own account.

As you all know from being communicators in the federal government, the historical way that federal websites are built and maintained are usually by contractors. There are these many different contractor teams throughout many agencies running sites.

Each one may or may not have had a Web analytics tool on it. If it did, it would be its own separate entity. You can see site one would have Google Analytics account one, or maybe not even necessarily Google Analytics.

There are other competitors as well, Adobe Analytics, and there used to be Webtrends. Webtrends and Adobe I think merged. GA was the largest even from about 10 years ago, but they would all just be different.

For example, if the CIO or CTO of the Department of Interior wanted to know like, "What is the most in-demand content across all of Interior's websites?"

The way that he or she would have to try to get that data would be to set some sort of monthly reporting structure, where some point of contact from every single one of these websites would have to send information.

If the sites were not all using the same tool, you couldn't even compare that information because it wouldn't be apples to apples, because they'd be recording different things.

This was an issue. Even down at an agency level, some agencies have dozens, if not hundreds of websites that it maintains. There was no real way to have reporting in aggregate on them

Then scoping it out all the way to the entire government, there was no way for administrations to be able to get a good handle on the technology that people were using to access federal government sites across all of them.

What was the most popular content and, for example, where the visits were coming from, whether that be via organic search on Google or via other methods, social media counts, things like that.

The goal for DAP was to one, make sure everyone was using the same tool and make sure that the tool was on as many websites as possible so that individual site owners would have metrics on how their own web pages were performing.

Also be able to scale that to the level across all of government so that we had a better idea of the content and the technologies that people were using to access federal information.

We went from previously this structure to now everyone pouring data into this single DAP Google Analytics account. What that allows us to do is look at things as granular as a single click on a single page but scope it out to the level that it encompasses all 5,000 participating websites.

That has been a really, really interesting adventure for us and provides a lot of insights that I think are valuable both to as far down as a content manager for a specific site and scoped out to big policy decisions regarding how we do digital government.

Moving back up a couple slides to what we offer as a program, what we do is firstly, we offer that tool to agencies at no cost. Here at GSA, we purchase a really large account for Google Analytics. We offer that tool to all the agencies at no cost.

The way that you implement the tool is to add a piece of JavaScript on your web pages. Then it will start collecting data and sending it to the DAP Google Analytics account.

I want to be clear that that doesn't preclude you from also using your own separate, individual Google Analytics account. In fact, I encourage that folks have a separate web analytics account as well, that is not just the DAP account, because there are pros and cons to both. Our tool, it's paid for by GSA and managed locally by our team here.

We also offer a lot of training. This is one of the main focuses, or foci, [laughs] that I have been part of since I took over as program manager. Really, I wanted to put an emphasis on teaching agencies, and folks at agencies, how to use this data to improve web experiences for the users, for the public.

We have an almost monthly training schedule where we either have a one-hour webinar or we have a three-hour introductory class and an advanced concepts class that we do. We do each one like once or twice a year.

We do one-hour webinars on very specific topics to try to help people understand how to make use of the data and ultimately make those user experience improvements.

We also manage analytics.usa.gov, which I'll show you in just a moment. That's the public window into the federal government analytics.

We built a public-facing dashboard that shows some of the more high-level data but doesn't have the ability to do analysis like Google Analytics does if you have access to our Google Analytics account. It's been a really, really great way to increase transparency and to show the public the power of this kind of program.

Finally, we offer one-on-one support to the over 3,000 agency users like yourselves to basically just help answer your questions about how to use the data. That might be as simple as "Can you help me find out how many times a certain PDF was downloaded on a page?" to how to use campaign URLs and various different things.

We are, unfortunately, only a team of two. It's myself and one other person, but we do have a lot of tickets that we help people with in terms of customer support.

I guess I had these in sort of a goofy order, huh?

Here are some numbers about the Digital Analytics Program in general. As I mentioned, it was created with the 2012 Digital Government Strategy.

Currently, we have over 60 agencies participating. When I say 60, I'm referring to cabinet agencies as one, so Department of Interior is one agency. You'll see when I show you the interface, everything for Interior is involved in a single view or profile. It's the same way for much of the other large cabinet agencies like HHS and Commerce, and things like that.

We have over 5,000 unique websites that are participating and sending data to this account for analysis. We record over 2.7 billion page views per month on those 5,000 sites. We have, as I mentioned, over 3,000 individual users across government who are accessing the data in our system.

We have approximately a thousand attendees to our various trainings annually. I guess you all could be considered part of that number now. [laughs] This is a special training for you folks in Interior, but we have trainings available to all our users as well.

If you become a user and gain access to the system, we put you on a LISTSERV where we keep you up to date with all the trainings that we do.

One thing that I wanted to mention before we go further was that not all Interior websites are currently participating, but we'd love it if they were. A lot of Interior sites are. The way that you can check is on pulse.cio.gov, and I'm going to go there just so you can see what it looks like here.

This is our website that was built by some friends of ours in another division of GSA. One of the things they wanted to do was track the websites that were participating in DAP. I think Interior is hovering around 60 percent right now.

If you work on a specific site with Interior and you would like to find out if DAP is already implemented on the site, you can go here and check. The link will be in the slides, I think, that are already available for you to download.

You can see there are quite a few that do participate here. These are all of the what we call second level domains, which is like something.gov, does include subdomains.

It's possible that you might also be working on a subdomain of blm.gov for example, that is participating in DAP but the main blm.gov site is not. There'll be contact information for anyone who wants to do a check or find out how to implement after this call.

I mentioned externally. Before I show you the Google Analytics account itself, I do want to talk a little bit about what we have externally-facing. That's at analytics.usa.gov, if you haven't seen it before.

What this site basically is is we took a lot of the data that we have and used an API to create this dashboard. As you can see, right now there are 296,000 people on government websites. This is realtime, so right now there are 4,000 some people applying for Federal Student Aid via FAFSA, we've got the Postal Service here, Medicare, Weather Service.

This morning, the number one most viewed site in the 5,000-site DAP data set was the OPM Status page, unsurprisingly.

[laughter]

Tim: I've seen the OPM Status page actually have 30,000 concurrent visitors at like 9 or 10 o'clock the night before a snowstorm, so that definitely happens.

The most visitors concurrently I've ever seen on a single site was the NASA page offering live video of the eclipse, and we had over a million concurrent visitors on that single page, which was incredible.

You can see here we also show some information about what kind of technology folks are using, the browsers that folks are using to access these pages, the top downloads across participating sites, some location information.

You can actually even go up here to the drop-down and filter this by agency. If I wanted to, I could filter for Interior, and any and all of you right now could go here and view what's most popular on the participating Interior sites right now.

It looks like Interior has got about a 40 percent mobile usage in the last 90 days and 9 percent tablet, which is a bit higher than the government average. Folks that visit Interior's sites like to use tablets more than other agencies. [laughs]

That's analytics.usa.gov, and it gets used by a variety of audiences, including the press or industry media on computer stuff. As far as I know, our data set on analytics.usa.gov is the largest data set of real analytics, Web analytics data that's made public anywhere.

We get a lot of things like this, like ZDNet talking about, "What's the most popular Web browser?" They gave us some nice little compliment here saying, "Most Web browser metrics aren't worth the pixels they're written in, but the US's DAP Program is based on hard data from real people, and the real winner is Google Chrome."

We see a lot of that, but also the broader, aggregate level analysis that we can do from this type of thing is be able to say, "OK, we know across 5,000 federal government websites, that less than," and I can give you the real number here, "Less than about one percent use anything before IE11."

If any of you are familiar with some Web development practices, you may know that supporting old versions of browsers can become very costly and cost a lot of resources in terms of person hours as well.

Instead of spending money to be able to retroactively display the page correctly in IE8, for example, you could make a decision across the Department of Interior maybe to stop supporting anything before IE11 because it's only going to affect one percent of your total user base.

That's sort of the decisions that you can make based on some of this data when it's shown in aggregate.

Hopefully you aren't like this, but I feel like when I was going through making this presentation, I was like, "People are just going to be like, 'Yeah, cool story, man,' like, 'OK, how does this affect us?'" Hopefully, I can show you how it can also help you as well now.

Beyond the public aggregate information, things that you can do as a user if you become a user of the Digital Analytics Program is you can segment visits to your specific website, whatever specific website you work on.

If it's participating in DAP, you can segment by technology, or location, or behavior. If you have outreach campaigns that are designated for specific locations, you can see how those campaigns are performing in a certain state, for example.

You can do analysis on mobile versus tablet versus desktop to see, if you did redesign and you redesigned to be mobile responsive and you can see if your targeted conversions, the things that you want users to do on your site, are being improved by the redesign, by segmenting for mobile device usage.

View where traffic is coming from, whether or not it's referral traffic, or coming from social media, or coming from Google. Get insight into user behavior, so what are people viewing on your site? Are they viewing five pages per visit or are they viewing two pages per visit, and is that a good or bad thing?

Creating reports for specific metrics and dimensions. Obviously, it's the government, so everybody wants reports. I feel like I'm always on my soapbox saying creating reports for the sake of creating reports isn't doing anyone any good, but hopefully, you can create reports that folks look at and analyze and you can take action based on it.

You can also track intended user journeys. If you have a specific journey that you hope a user will take to sign up for a newsletter, or complete an application, or view some sort of specific content, you can actually use the Digital Analytics Program, Google Analytics account, to do what we call a funnel report.

Where you can see, if you have five steps in that journey, it will tell you when people started step one, what percentage of them make it to step five, and which step they actually fall off at. You can pinpoint pain points -- pinpoint pain points, say that five times fast -- for the user, and then use other tools to be able to find out how to solve that.

Maybe if you find out that a lot of people are dropping off at step three of a five step process, you can then use usability testing, or heat mapping, or one of the various other tools that folks should be using, to improve user experiences in order to improve that.

One of the biggest things, I think, for comms folks is to be able to accurately find out whether or not your outreach is producing the results that you're hoping.

For you folks that are potentially sending out emails or operating social media accounts, the Google Analytics interface can help you understand if those are successful or not. Then make decisions about where you should dedicate resources or if you should put more, or less, effort into a specific channel of outreach.

Here's my go to example of one of the great case studies that folks have used DAP data for.

This is ftc.gov, and previously, and a while ago now, but a few years ago, ftc.gov, when you went to the home page, it actually had this photo that you can see, this little slider. This isn't the real, this is just a screenshot. The slider took over this entire space, the whole above the fold area of the home page was a slider with some photos that you could click and read articles.

What FTC did is they used DAP to do an analysis of the search terms that people were putting in their on-site search. When you go to the search box and you put in a search term, DAP can record the search terms that people are entering and find out what are the things that folks are looking to find on your site that they don't see it on the home page.

Basically, what FTC did is they took all this information and they found out that, through some analysis of taking similar terms and putting them together, that there were really five main actions that folks were trying to accomplish in great numbers when they used the search box on the website. Those five things are what you now see in front of you in your Take Action section.

Through various different search terms, they bucketed everything into these five, and they did really, really wonderful analysis, beforehand and afterward, of how many folks landed on the home page and were able to complete one of these actions before they made this change and then how many people were able to complete them after they made the change.

They just cut this slider photo in half, and they added this Take Action section, and now anyone can easily click to find these.

I think they told me that prior to this change, three of the five actions were available from the home page, but only if you hovered up here and you had to drop down in one of these topics. Two of them were more than one click away. Three were like you could hover and click once, but it was somewhat hidden, and two of them were more than one click away.

Now that they've done this, they had wonderful numbers that showed that their users now have, much, much, like hundreds of percent greater completion rate of these tasks.

Also, their total amount of searches on their site search from the home page plummeted, completely plummeted, because now, the main reason that folks were coming to ftc.gov was to find these services that were more easily accessible to them from the home page.

That's just one example of how you can use some of the data that Google Analytics and DAP can help you get to make decisions that improve the user experience.

Another quick one is we had a nice little anecdote from Hurricane Irma, from someone who works at FEMA. One of the interesting things about the Digital Analytics Program is that, when you become a user, you are not limited to viewing only the metrics that are coming in for websites that are at your agency.

When we first created the program, that was the case. Previously, if you worked at Interior and you got access to DAP, we'd only grant you access to the Interior profile, but a long time ago, we decided that that was blocking some really interesting potential for collaboration across government.

Now, we changed the rules quite a while ago, where, if you gain access, you have access to the entire lot of DAP data, all 5,000 websites.

When something like a natural disaster like Hurricane Irma happens, this particular person at FEMA was saying that they were trying to get information on everyone accessing resources across many different agencies.

Usa.gov had information about what to do if you were affected by the hurricane. Obviously, FEMA had information up. State Department had information up for US citizens in the Caribbean that were affected by the hurricane. Various different agencies, there were like five or six agencies that had information about Hurricane Irma.

What this person realized was they could just go and get information about the traffic to those websites without having to do a data call to the agency, which probably would take a lot of time.

Instead, the person just went to the State Department profile in DAP and found the specific pages that were of interest and then was able to get statistics on them. This was something that this particular person was really excited about.

I think that this plays out a lot in inter-agency stuff too. Within Interior, maybe there are multiple sites within Interior that you would like to have stats on, but you don't have to go to a bunch of different accounts to be able to see them.

Rebecca: Also, Tim, sorry to jump in, is that you also don't have to share passwords, then, too, right?

Tim: Correct, you don't have to share passwords, yes. We try to grant access mostly at the individual level. There are certain agencies that request that they have of a shared or team address that's granted access. At that point, the password changing would be on the agency, the manager of that particular email address.

I think Interior uses Google already, so that eliminates one step, that you don't have to associate your email with Google. We can just grant access directly to an email address and then you would be able to login yourself and just be able to do it without having any extra account or anything that you need.

The sum of what we hope to do with the Digital Analytics Program is gain a better understanding of our users and therefore make better decision making through the analysis of the data and finally, better the experiences for the public.

Now I get to jump in, we can take a little look in the system itself. I can show you what's available to you. I want to make clear that I do a lot of trainings, and Google Analytics has a bit of a learning curve. I'm not going to try, in the next 15 minutes, to teach everyone how to use the system. What I want to do is just demonstrate some of the information you can get.

There are resources that I put at the end of this particular presentation. You can take some of our trainings, we've recorded them all, or feel free to get in touch with us and we can try to help you, as well. I just want to show you some of the things that we can do here.

Here is the DAP Google Analytics account. You can see here, I'm in the DOI agency profile. We also have agency profiles for all the other agencies, so here's HHS, and GSA, and FTC. All the participating agencies are all here, but the Interior profile is obviously of most interest to all of you.

What we can do is see all kinds of really interesting information here. Obviously, across the entire set of participating Interior sites, we've got counts on users and sessions and things like that. I think I've got the date range currently set to the last 30 days, just so that you know. You can change the date range over here.

One thing we can do is look at folks real-time. This doesn't have a ton of value for reporting, but it's kind of cool if you have a press release or there's big news and you want to see how it's performing. Right now, this is a really popular page all times of the year. The earthquake map from USGS, it's the top page right now among participating Interior sites.

You can go here and you can take a look at various different things to find out what people are visiting at the moment, and you can also break it down by location. 89 percent of the visits right now are in the United States, and so on, and so on.

You can see traffic sources as well, which I think for the folks that are, since you mostly are communications folks, I think this is probably one of the big areas that you'd be interested in. Classroom.google.com, so there must be some curriculum that a teacher is using that leads folks to a particular page in the Interior. All kinds of interesting things here.

I'm interested to know what ducksters.com is. I'm assuming that's duck hunters. Anyways.

That's the real-time stuff. More in the breadbasket of what you can do with this content afterward is look at things like the All Pages report. Basically, the All Pages report is just going to give you information about each page on all the participating sites for the date range that you selected. Hopefully, it'll load for me.

While that's happening, I want to get an example to search for, so if I go to usgs.gov, and then I'm going to try to go to a relatively obscure page. Volcanoes would be cool. If I take this URL here, I'm just going to copy it, and then we'll go back and hopefully this is loaded. You can see here, the earthquakes page is number one.

This other is actually a bunch of very, very sparsely visited pages that are put together, like the very bottom part of this long list arranged by page views for the date range, will be squished into this other, unfortunately. You can kind of ignore that one.

These are the most popular -- Earth Explorer, earthquake.usgs.gov home page, various different things like this. Oh, search for soldiers in the Civil War from the National Park Service, that's a pretty interesting one as well.

What we can use this for is if you are interested in knowing how a specific blog post is doing or something like that -- I went to this Volcanoes page and I just copied it -- you could go and paste that particular page here.

Notice, in DAP, I don't use www, or HTTPS, or anything like that. It's just from usgs.gov. I search for that. We should be able to find information about this particular page. Here you go.

In the last 30 days, this page, it looks like there's some other stuff that comes from different ways that people get to the page, but 99 percent of them were to the main one here. 10,683 page views, a one minute 11 seconds average time per page.

Entrances means that this was the first page that people saw. Entrance means when you enter the site, this particular page was your landing page. They didn't see the home page first. They came directly here, probably from Google.

1,600 times, it was the landing page, and various different information. I'm not going to go over what bounce rate is at the moment. You can get a lot of information that's pretty quick if you are interested in specific pages.

Another thing which you can see is landing pages. I talked about entrances. This is correlated with that. You can see the top pages that folks are landing on when they come to the site. By far, the most popular is folks are going directly to the earthquake map. Number two is the earthexplorer.usgs.gov. It goes on here.

This is one is interesting. I don't know what that is. Yellowstone, photos, multimedia, webcam. I'm not sure what that one is. You can extend these lists up to 5,000 rows. You can export these lists into CSVs, or Excel, or PDFs, whatever you want to do with them. It's pretty helpful if there's a specific page that you're looking to find information about landing pages or all pages.

One thing that I wanted to show that I think is probably really interesting for you all is the channels report, which is related to acquisitions. How did people get to these websites or these pages on the websites?

You can see we've got organic search accounts for 60 percent. That's Google, Bing, Yahoo. Direct traffic is theoretically traffic that's either from a bookmark, or placed in the URL bar, like someone typed it or copied and pasted it. Referral is from another website. I'll click on that in just a moment.

Social is pretty obvious. That's from social media sites. Google has identified a whole bunch of them. Your obvious ones, Twitter, Facebook, Instagram, LinkedIn, things like that. It also goes into Reddit, Sina Weibo, which is the top social media in China, VKontakte, which is the social media that's like Facebook in Russia, and things like that.

Email, display ads. We don't usually see a lot of ads that are coming in in government, but this is the sort of breakdown. What you can do then is if I click on referral, you can actually get more granular with the data to find out the top referrers.

Again, we've got classroom.google.com. If I click through this, it might give us information about the specific pages that people...Ah, this is actually...It's classroom.google.com. This was the end of the URL. These are all the classroom.google.com URLs that are leading people specifically to an interior website and page.

What you're probably thinking here is, "OK, but this is for all the websites." You're like, "I work on one website. I don't really care about things in aggregate across the agency." I feel you there. One thing that we can do is create segments so that we can filter the data specifically for a specific website.

What I've done is I already created a segment for USGS. Here's the USGS family. I'm going to show you what this looks like in a second. Let me hit edit. What I did here is you're able to segment this data for anything you want.

You can segment it for mobile users. You can segment it for Chrome users. You can segment it for people visiting from Arkansas, whatever you'd like to do.

What I did here is I set a segment condition that says I want to filter for sessions where the host name, and I chose "contains" usgs.gov, rather than "exactly matches." We've got a whole one-hour webinar on segments that will help you understand all this stuff. Just keep in mind that you have the ability to filter for a specific site.

If I said "exactly matches," it would only show me information for folks on usgs.gov exactly, but because I said "contains," what's interesting here is that you can see, because it autofills for me, now it will contain data from anything that contains usgs.gov. That's all of the subdomains as well. Earthexplorer, volcanoes, all that stuff. That will all be included in this segment.

I could narrow it down just to one site for the purposes of this demonstration. I'm going to make it more broad and say the usgs.gov family. The second-level domains and all the subdomains.

When I apply the segment, what we'll see then is it then takes the data and filters it all into a view that would be more similar to what you would have if you had a site-specific Google Analytics implementation. In this case, it's not a specific site, because we've got a bunch, but you could do it for a specific site.

Now, if I go to channels, the main channels report here, now I have information on the specific usgs.gov family where that information or where those visitors are coming from.

One thing to note about this is that because I made the segment multiple sites, theoretically, folks could be going from usgs.gov to waterdata.usgs.gov, and that would count as a referral, because any separate site is measured as its own entity. For usgs.gov, we can see here we've got organic search, 48 percent, direct, 32, and so on.

Then we've got social media. This brings me to a point that I wanted to talk about, especially when I'm dealing with folks in government communications, that I feel like I have to make. That is, if you don't already use campaign URLs, please, please do so. I'm going to talk in a minute about that a little bit.

You can see all these different social media platforms that are bringing traffic to the specific sites. If I wanted to, I could actually add a secondary dimension, and it could show me exactly where those places are delivering traffic.

Once this loads -- I know that I'm breezing through this, but trust me, you all can learn how to use this platform -- now I can see Facebook directs people to the earthquake page, 8,000 users and 23,000 sessions over the last 30 days. Reddit brought a lot of people to earth, how much, and water.usgs.gov, and so on.

These are cool because you can see whether or not social media is bringing a lot of traffic to your site.

More specifically, for those of you in comms, we have campaigns. This is where I want to talk a little bit about campaign URLs. Let me go back to my slides for a second. Here is where I get on my soapbox again and make a few statements.

If you include links to your website in emails without campaign URLs, which are, if you've ever seen it, a campaign URL would look like this. ?utm_source=twitter, these kinds of things. Some people call them UTM codes.

What this does is when the page loads, Google Analytics sees a specific code, utm_source, utm_medium, utm_name, and it says we have to record this data.

That is super, super powerful, because what happens is, if you put a link in an email and you don't have a campaign URL, if the link is clicked on in a browser-based email client like Gmail on a browser, it won't come in as email. It will actually come in as a referral from Gmail.com.

If they're not using a browser-based, like if they're using Outlook, a desktop client, it will come in as direct traffic, because when the visit loads the page, Google Analytics tries to look where it came from, and there's nothing there, because it didn't come from the Web.

You get a skewed view of whether or not your email campaigns are succeeding if you don't use campaign URLs. I've got a whole one-hour webinar on campaign URLs I included in the resources here.

Another big one is link shorteners. If you use any link shortener other than t.co, which is Twitter's native link shortener, all the traffic will appear as direct, because for lack of a better technical way to explain it, the link shortening service just bookmarks whatever you put into it.

In the same way that if I click a bookmark and go directly to a site, that would come in as "direct traffic." If you use link shorteners, that also happens. Even if you're posting the links on Twitter and Facebook, it will look like it's coming in as direct traffic if you don't use campaign URLs before you shorten your link.

Thirdly, adding campaign names can help you better assess how your outreach efforts are paying off. Campaign name is something that when you create a campaign URL, you can actually list a campaign name.

You can see there are quite a few campaigns being used for the usgs.gov family. These are the names that are associated with those campaigns. They have a campaign where they literally put in the URL, "Latest earthquakes." It would like this. &utm_name=latest+earthquakes, like this. That's how you would create this.

There's a whole tool, Google Campaign URL Builder. There's a tool that will help you create these for when you are sending this information out. You can put in the website URL, the source medium, and name, and then it will spit you out your specific URL that you can then shorten.

This allows you to understand where your traffic's coming from. If you have long-running campaigns -- like obviously latest earthquakes is not a single day campaign, it's something that goes over time -- it allows you to be able to do reporting on the long-term success of the campaign, but then also drill down to what specific channel led to that.

I could have a campaign URL that specified latest earthquakes, and then I have three separate URLs. I put one on Twitter, one on Facebook, and one in an email, and then you would pass that information into here so that you could read it and understand it.

I know that's a lot, but I have to make my pitch about campaign URLs if you folks don't already use them. Please feel free to reach out to me for more information on that, if you need it.

I don't want to go too much more time, but there's one more thing I want to show you that's a big value-add for DAP, and that is that we automatically track a bunch of events that out of the box, Google Analytics implementations do not track.

We've hard-coded into our JavaScript library that we track things like outbound link clicks. That would be a link click that is leading outside of your domain, like maybe from usgs.gov to opm.gov. We track downloads of PDFs, and docs, and all kinds of things, jpegs, zips, xls.

PDF is number one. If you have a specific PDF that you want to know how many times it was clicked on your site, you can go here, you can search for it. You can find out how many clicks that PDF has had in the last 30 days or whatever you set your date range to.

We also track things like mail-to clicks, as well as telephone clicks. For those of you that have websites with a telephone number where mobile devices can click that number and it will automatically call, we track those and what pages those happened on as well, and how many times.

This is all done for you if your site participates in DAP. If you are implementing a Google Analytics out-of-the-box implementation, all of this would have to be coded, and you would need developers to be able to put the information on the web pages that need to make these things function.

We've already gone ahead and done all that for you so you can get a good idea of what's going on on your pages after people load them.

Hopefully, now you went from, "Cool story, bro," to, "Not bad." I know that I had to go through a lot of this pretty quickly. My main points for everyone to understand are that if you use Google Analytics, there's a lot of data. If you use DAP, there's an even larger amount of data that you all can have access to, and it's free of charge.

If you would like access, or to try to speak with us to help get your site participating in the program if it doesn't already, you can email us at dap@support.digitalgov.gov. We're happy to try to either get you access and/or help get your site participating as well.

In the slides, you'll see that I have included some of the resources I've mentioned. We've got our GitHub repo implementation instructions, our giant YouTube playlist of all our recorded trainings.

I think we have some 20-some, 30 videos if you're looking to learn more, as well as specifically the campaign URL webinar -- that's really important -- and the DAP 101 if you're just getting started.

48 minutes. Super close. All right, I'm ready for questions.

Rebecca: All right, Tim. Thank you so much for that. As a reminder, the side deck is in the handouts tab. We are taking questions.

One of the first things that everybody wants to know is how do they get on the mailing list for these additional trainings and get access? Just to email you, or through the email address that you just had up there?

Tim: Yeah. If you email us and request access, there are a couple of steps that you need to follow.

Folks at Interior -- I don't know if the entire agency is using Google -- if you don't use Google as your mail client, you'll have to do one step where you take your government email address and you associate it so that Google Analytics can make an account for you.

Then you have to fill out a registration form so that we know who has access, and where they work, and things like that. Once that's done, we will grant you access, and you'll get a confirmation email that will also let you know that we add you to our Listserve. The Listserve is how I do all the marketing for our upcoming trainings.

I will say that we usually take December off from trainings. We just wrapped up the whole year, but come 2019, we'll be planning a whole new set of trainings and, like I said, we usually try to do one monthly.

Rebecca: One of the other questions is wanting to know what's the difference, when you were talking about segments, of creating a segment versus just using the search window to filter that.

Tim: The difference is that when you put a segment on, it will hold as you go through any of the reports. For some of the reports, like the all pages report that I showed you, where we were looking at them all, they have a search box right here.

Some of the other reports...There are tons of them. One of the other ones I showed you was channels. If I'm showing you the channels report, there isn't a way that you could segment out the channel stuff with a search box. You would have to either create a custom report or create a segment that would filter all the data for you in the interface.

Right here, this search box would only relate to the results here. Assuming this segment was not put on, this data would be representing all the participating interior sites, and all this search box would do would be to show you where direct, or referral, or social was. You can already see them, because there's only eight rows.

With the search box, you couldn't say, "Show me referrals from usgs.gov." That's why we use the segment -- because it's the easiest way to just filter all the data, and then you could go anywhere in the reporting interface and it would still apply that filter.

Rebecca: The access question is can contractors get access to DAP?

Tim: Great question. Contractors can get access to DAP, as long as we have written approval from their federal supervisor. You can request access and c.c. your supervisor, and if your federal employee supervisor approves that you get access, we can grant you access.

Rebecca: The other question is wanting to know is it correct that all tools -- Analytics 360 -- demonstrated are available at analytics.usa.gov?

Tim: No, if I understand that question correctly.

Rebecca: One's public-facing and one is you get a little bit more.

Tim: You get a lot more. With analytics.usa.gov, what you see is what you get. There's no slicing and dicing of this. It's just sets of data. You can download the data sets, but it's just flat tables of data. There's nothing dynamic about it other than it changes every few minutes when we run a new scan.

With the Google Analytics interface, once you get proficient in using it, you can break down anything you want. You could say, "Show me how many people visited from California, using Google Chrome, on Tuesday, and visited this particular page, the volcanoes page."

You can get that data succinctly, whereas here, it's just sort of there are this many people on the pages right now. Here's from 7 days and 30 days. You can't get much more granular than this. The Google Analytics interface is really where you want to do your analysis.

Rebecca: How does DAP work for internal sites? We're on Google. For example, there's something called Inside NPS, which is a Google site, through the platform.

Tim: That's another great question. I'm glad it was raised. Unfortunately, DAP is specifically designed to be able to help us understand how the public interacts with federal government websites. Therefore, we don't allow people to put it on intranet or sites after an authentication.

It's very specifically supposed to be only for public-facing pages. Intranet sites should not be implemented with DAP. Also, just as another caveat, there's an ongoing discussion right now about whether or not third-party services like Google Analytics should be on any web page after an authentication, even from the public.

If you have a site where you have users that do authentication, even if they're public -- they're not federal workers, but they have an authentication -- we also have a policy that DAP should not be on any page after an authentication. That's something that we can work out with your developers if it affects you.

The main bread and butter here is that the DAP script tag and the information that's recorded in the DAP account, because it's managed by GSA and has not been...Basically, we don't want to get in the realm of accidentally collecting any PII or anything like that.

If you want to track things on your intranet sites and/or on sites where you have authentication, I would recommend working with your team or your supervisors to look at getting a separate implementation, your own implementation of Google Analytics that is not...

It's independent from DAP. Then work with your agency teams, like privacy and security teams, to make sure that everything about that is following your particular agency's policies.

Rebecca: One of the things that came up a couple times is -- I know you said you do a more detailed campaigns training -- how does the campaign URLs work with link shorteners? Specifically, one person asked about the Park Service's official link shortener, which is go.nps.gov, which I'm not sure if you're familiar with or not.

Tim: I have not heard specifically of go.nps.gov, but I know our colleagues at usa.gov used to also have go.usa.gov. Basically, the way it works is this. If you create a website URL, like if I just type in usgs.gov. Then I say the campaign source is newsletter. The medium is email. The name is going to be Weekly Newsletter, something like this. You'll see here it gives me this particular URL.

Then I just take this whole URL, with all of this UTM, source, medium, and campaign there, and I shorten that. You put this big thing into your link shortener. Then, when someone lands on the page, the link gets expanded again. It shows up as what you had here.

For those of you who don't understand this little question mark, you can put anything after a question mark on a URL, and you will still go to the same page. Anything after the question mark is what we call query parameters. The parameters here are designed to send information to various different systems.

It's usually how search works. It's how we can send information to Google Analytics. UTM, source, medium, and campaign are the specific codes. When the page loads and Google Analytics runs to collect the data, it looks for those.

You could put a bunch of random stuff in here and get the same page, but that wouldn't send any data anywhere. That's how it works. You just need to create your campaign URLs, shorten them afterward, and then put your shortened links in your emails or on social media, wherever you may put it.

Rebecca: Our last question is you mentioned you do consultations. If someone is interested in having a consultation to see if their website is a good candidate for DAP, if they're not already on it, do they just email you at the support email address as well?

Tim: Yes. Please do. This gives me a good opportunity as well to cite OMB Memo M-17-06 that actually mandates that all public-facing federal .gov websites have DAP scripts on them. If you need a good argument to make to your supervisors about why DAP should be participating, it's actually mandated by OMB.

It's just that OMB doesn't have the manpower to run around and enforce it. I can send you the information if you need it. It's M-17-06 that requires all federal public-facing .gov websites to be participating.

Rebecca: Great. Wonderful. Thank you so much for your time today, Tim. We are right at our three o'clock mark. Unless there's anything else, we'll end this webinar.

[pause]

Tim: Sounds good. Thanks so much for having me.

Rebecca: Thank you so much. Everybody stay safe out there.

Tim: Bye now.

Rebecca: Bye.

[silence]

11/16/2018
Last edited 9/29/2021

Ever wonder who's reading your web content and what they want to see next? Learn how to use Google Analytics to figure out where to focus your efforts in content creation based on what people are reading and searching for. This class covers the different data points to look at, how to create reports and what's important to share with your leadership.