Tip:
Highlight text to annotate it
X
Avinash: Hi everyone
Nick: Hey
Avinash: Welcome to the next episode of "Web Analytics TV". My name is Avinash.
Nick: And I'm Nick
Avinash: And we're gonna go through your exciting questions from out moderator page. We've got
a really good clump of question this time around Nick.
Nick: We do, yeah.
Avinash: And the very first one is challenging and thrown at you.
Nick: Ok.
Avinash: And it's from Cesar Brea from Denver, from Dover, sorry. And he is asking, "I'm
trying to create a target towns reports using advance segments.
But getting the towns into the report requires a different logical structure than this capability
seems to support." So how can he create a bunch of towns rather than -- in advancing
addition in Google Analytics?
Nick: Right, that's a great question. There's actually a link in the question.
The link points to a support question. So apparently in Wikipedia, there's a list of
cities and towns that have a household income over certain range.
Avinash: Oooh
Nick: Right, so what we want to find out is give me this group of cities.
Maybe about 50 or 70 plus cities that have a high household income. So in advance segments,
there's a couple ways to define them. One way is, for example say San Francisco was
in that list, and you wanted to include it. You could say San Francisco for the expression
to match San Francisco exactly. And then click +or. And then say for the regular expression
-- for the advance segment to match San Jose.
Avinash: Right
Nick: Or another city, or another city. So you do each city individually.
Avinash: Right
Nick: So the challenge with that is that you have to create 70 of them, and it becomes
really challenging. So what you can do is instead of using actual ors, is you can use
a regular expression.
Avinash: Oh I see.
Nick: And regular expressions allow you to actually put or in the expression yourself
using a pipe symbol. That's the vertical line '|'. So you can say, "Match a regular expression
where you say San Francisco | San Jose." And you can put a bunch of those together. And
see, you maybe have 10 cities on one expression or another 10 cities, or another 10 cities.
So you can get a whole bunch of cities together in this one advance set-up.
Avinash: Right, I would still drag town over, then I would say matches regular expression.
Nick: Right.
Avinash: And then I can just put it all in one box essentially.
Nick: Right.
Avinash: And typically I think the way I end up doing it is I just create it in notepad,
you know.
Nick: Exactly.
Avinash: I just type it out in Notepad, and then copy that and past it Google Analytics
Advanced Segmentation. And it works fine.
Nick: Absolutely. And since this is actually a link from our group, whoever does create
it in notepad, paste the actual expression in there so that everybody else can get it.
Avinash: There you go, perfect, perfect.
Nick: So question for you Avinash. The index -- this one is actually from Tambu in Genoa
Italy.
Avinash: Oh, I know him. He's a good guy. Hi Tambu.
Nick: So index dollar sign index currently uses monetary incomes from goal value and
e-com revenue to determine a per page value. It's a per session calculation. If I link
my AdSense accounts, are these incomes added in calculation, even if they are per page.
Avinash: Actually, it's not. And so what you're noticing -- any about AdSense
revenue as well as you've set goals with goal values. Then what you see in dollar index
are just those e-commerce service revenue / goal values that you've put in there.
The AdSense revenue part is not there. It's happening sort of outside the ecosystem, not
within Google Analytics. So that is kept outside Tambu.
So feel free to use it aggressively to promote your website.
Nick: Great, alright.
Avinash: Here's a good one for you Nick. From Moshin in Pakistan. We're going really
Nick: International Today.
Avinash: Yeah.
Nick: Yeah, this is good.
Avinash: How do I attract affiliates of Google Analytics? I have an affiliate software installed
in a separate directory/affiliate. But I would like to know best practice of managing it
in Google Analytics. Now this is not that hard.
Nick: Right, so for affiliates, what they're saying is somebody will pay you for links
that point to your site. And they generate a lead or make a transaction. So you just
want to measure the different links that people send to you. So you can do that using campaign
tracking. There's a set of parameter for each link that points to your site. So you can
ask your affiliates that use these campaign parameter in their campaign's to send you
traffic that you're paying for. And when somebody clicks on those links, and Google Analytics
will interpret those parameters, and you'll see campaign reports, the source and the medium.
So you can set the medium as affiliate, and the source as you know the name of the affiliate
and then the reports you can actually say of all my traffic, show me total traffic from
affiliates, and the broken out by the actual ones that you're working with.
Avinash: Exactly, so basically simply by leveraging existing Google Analytics campaign reporting
infrastructure, it's very easy to report affiliates.
Nick: Absolutely.
Avinash: OK, can we get a link to an article about...
Nick: Campaign tracking?
Avinash: how to use campaign tracking.
Nick: Definitely
Avinash: Good, good.
Nick: Great, so here's a question for you from Hazem in Raleigh North Carolina.
Avinash: Oh, this is my buddy Jim, Hi Jim!
Nick: So is there any way to utilize events as goals. For example let's say the objective
of a campaign or site is to generate a download of a PDF. So Legion, Wide Paper. Besides using
virtual page fuse, is there a way to use the event of downloading a PDF as the goal on
Google Analytics.
Avinash: Yes Jim, you can't actually use events as goals yet in Google Analytics. And what
I end up doing is simply leveraging the advance segmentation methodology. And apply, do that
to the events. And actually I have done this in my latest blog post on customer lifetime
value. I have an excel spreadsheet that's included to allow people to compute lifetime
value easily. I mean that spreadsheet that I have set up is something akin to categories,
downloads. So in my event tracking, categories, downloads.
And then activity is the customer lifecycle value .xls.
Nick: The name of the file?
Avinash: The file name. So now Jim, I can go back, and I've learned this from Nick of
course. I can actually go back and either create an advance segment that just tracks
the download events, so I can easily [Finger Snap] quickly get my number of conversions
if you will. Or I can then drill down further and see how many of those downloads were my
customer value lifecycle spreadsheets. So that's a very easy way. So while it's not
in the goal area, but you still have this available. Just leverage event tracking. In
fact event tracking, you might even have a more scalable way of doing this then simply
being able to track a one off. So give it a try Jim.
Nick: Definitely.
Avinash: Alright.
Nick: So here's another one.
Avinash: Another one.
Nick: Another very similar question,
Avinash: Sure
Nick: For you Avinash. This is from Ohio, from Illbit. How can you tie a video to a
conversion? For example, a user watched a video and then convert it.
Avinash: This is a great example, because, in my book "Web Analytics 2.0" you'll see
the event tracking described in video tracking. And that example comes from Nick. Actually
Nick helped me create the spreadsheet that you'll see in the book. And essentially we've
done, as well in this case is used advanced segmentation to do the reporting. And what
you'll see is we have set the category. In terms of event tracking, you get three different
buckets, if you will, of data that have a hierarchical relationship. So I've set up
categories, video. As in action is the name of the video itself, and the label is the
percent of the video watched. So I can go in and see how many people.
Let's say we did it over "Web Analytics TV"; we could see overall how many people have
watched the video. And by episode actually, we can see if people watched a complete video
of each episode of "Web Analytics TV", or they abandoned after 2 minutes.
Nick: Right.
Avinash: I hope not. But we can actually use that. So use advance segmentation, and use
event tracking, and you're all set. So the next one is from Beebe. I know all these people
this is great. Hi Beebe. Beebe's in
Nick: Vermont
Avinash: Vermont.
Nick: Yeah.
Avinash: And she asks: what is the best way to measure embedded flash for GA?
Nick: It's a tricky question.
Avinash: Yes.
Nick: So for people who don't really know flash that well, you can take flask, let's
say a widget, a new game, and you can have people embed it in Facebook, in MySpace and
different places around the web. So I think what you're asking is to track flash itself,
you can use events to track interaction. But I think the real question here is; how do
I track interaction by site placement? So, can I compare the number of interactions of
Facebook vs. MySpace? One of the tricky things with flash is again its very rich language.
But there's a lot of security concerns. And so the way Adobe built it is that the person
who is hosting -- the person, who allows you to embed the flash on their site, MySpace,
can disallow the flash from communicating with the page that's MySpace.
Avinash: Oh I see, I see. So it's a security sandbox that they've created.
Some sites allow you to do the communication, but other sites don't allow you to do the
communication. So the challenge is for the flash to know actually which site it was embedded
on. So using our flash tracking, by default if we have access to this property of where
the site is, we will report on it, and if not we'll just say it's not set. One way to
work around it is if you can embed your flash in an iframe. The iframe will respect the
same browser properties.
Avinash: Great. The referral in the iframe will be the actual pages loaded on.
So let's say you're loading iframe.
Nick: from site A, in MySpace, the referral on the iframe will be
Avinash: From the page.
Nick: for the page will be in that iframe will be in MySpace.com. So you can get actually
the site you've embedded your iframe on through referral. And then because you actually host
the iframe, you can actually reduce the restrictions on how flash communicates with the iframe
to actually get the tracking to work properly. So that's one option you can do.
Avinash: And actually two options I think.
Nick: Yeah
Avinash: One is that if you use embedded flash, and you use event tracking, and the site allows
you the communication to happen, then of course you'll get the data. And that comes through
fine. But in case it turns out that the site has set permissions in a way for flash, it
won't communicate to the site, then the second option,
Nick: Right.
Avinash: is to use the iframe link, and that works.
Nick: And just to be clear, when we talk about tracking flash, we actually have a library
that we worked with Adobe to build that allows you to natively track flash. So we handle
all the cookies in the library and everything. You build it into your actual widget, and
you make it run. That's what we're talking about here.
Avinash: Good, good. And we'll link to some more information.
Nick: Some information about it.
Avinash: OK.
Nick: Yeah definitely.
Avinash: That's great, thanks Nick.
Nick: Great, so the next question for you is from Steve-O in Brighton UK.
Again we're going international. I'm investigating how to combine my company's Google Analytics
and Urchin data.
Avinash: Oh goodness.
Nick: With the goal of linking this data to our non-web data via customer order ids. What
possible issues do I need to be aware of when trying to combine these three data sources?
Avinash: I feel sad for Steve-O. [Laughter] No, I mean, I have firmly believed and consistently
preached that when it comes to web analytics tools, you should practice monogamy, and not
bigamy.
Nick: Oh wow.
Avinash: Having one tool is hard enough. And to get one to work right is hard enough. Getting
two to four correctly is really impossible. So, first advice to Steve-O is if you can,
pick one. So pick urchin, because you want to have a hosted in-house environment. You
want to have everything in your end. You don't want to send the data into the cloud.
That's a good reason to use Urchin, use it. If you're ok with having data in the cloud,
just like every other web analytics tool, then just use Google Analytics, but pick one
please. It's really hard to make both of them work
correctly. Or it's very hard to make Omniture and Analyzer work correctly, or any of the
others. So pick one tool. Now let's discuss the second part which is
actually important, which is what is the issue if you take the web analytics tool data out
and then merge it with other sources of data inside your company. Specifically he mentions
customer order IDs, and the order ID is fine actually Steve-O because there not a PII connected
to it. Alright easy and 1234 garbage, whatever it is. The only thing you should be cautious
about when using Google Analytics is that the task prohibits you from collecting any
PII data and sending it over to Google. Or into Google Analytics, one of the reasons
people end up using urchin as an example. So you cannot collect any PII data. You can
collect any non-PII data; you can extract it through the API, which is an effort Nick
leads here. And then, once it's out of the system you can sort of merge in all that stuff
in your data warehousing environment over there. But be very careful and cautious when
it comes to PII data. Personally identifiable information and it being sent into Google
Analytics because that is prohibited. And things that fall into PII are e-mail addresses,
names, underwear size.
Nick: Credit Card numbers.
Avinash: Any of those things are PII, so be careful about collecting that with GNN, surplus
merging and things like that. Other than that, I think you're fine.
Nick: Great, so here's another question for you Avinash.
Avinash: Oh, yes.
Nick: So how do you feel about Google providing a feature to opt out from being tracked?
Avinash: Oh yes, so this is a blog post that the team has written a month ago I believe,
and just saying they were working on a feature that would allow this to have happened. At
the moment, we don't have any more information then what is available in the blog post. I
mean Google, as a company believes in having users having choices. As we learn more, we're
gonna make sure that we share it with you. So keep an eye on our blog. And perhaps we
will address it in a future episode of "Web Analytics TV". But at the moment, we simply
don't have enough information to share with you for now.
The next question Nick is for you. Another nice tricky one. And it's from, let's see,
Avingunpreeden in India. And it asks how to make a custom report with top landing pages,
the page titles for each of them and the unique visitors to them.
I cannot figure out how to display page titles on the report. I want to do this because my
URL's are not descriptive. That's a good one, yeah.
Nick: Yeah it is. There's actually some subtleties in here. So you can create a customer report
with the page, and then in the actual custom report builder, you can add multiple dimensions
to your report. And if you look at the UI, it's kind of slightly down the right a little
bit, you add another dimension. And you actually generate the report; you'll see all the page
titles with all your metrics. You click on the dimension, and it drills into all the
page titles for that.
Avinash: Ah, yes yes yes.
Nick: So you can use the page, and then you can get the page title, and you can get that
report. The one thing to keep in mind is that the relationship between page URL's and titles
is not necessarily 1to1, its 1toN. Avinash: Yes yes yes.
Nick: So what you'll find is when you click on a page, you'll see many titles usually
for that page.
Avinash: Yeah, so you have to be cautious, I mean one of the most important things is
that your website has to have clean page titles. And sometimes, we often see that they'll represent
it using the same page title on multiple pages. And then that's why have to be very careful
about it. But you can follow the custom reporting methodology that Nick's highlighted and it
works just fine.
Nick: It works really good.
Avinash: OK, that's great.
Nick: Great, so a question for you Avinash, this is from visitor in Charleston. How to
break down non-paid search engine visits that come to our site via specific buckets of landing
pages. For example, we just added zip code landing pages for a site. Each zip code page
ends zip.aspx. And so the goal of this is how do we measure impact of new pages.
Avinash: Ah, actually, this is a good one. It's slightly non-obvious, but he uses one
of the features I am very very fond of in Google Analytics. And, essentially the first
thing you want to be doing is actually make sure that you've created a segment for your
non-organic and unpaid traffic. It's very easy to do. Just go into advanced segmentation,
create a segment for your organic search traffic for all search engines, or only for Google.
Save it, now go back to your -- the landing pages report. There's a dedicated report for
the top landing page you'll find it in left Nav in Content. Content top landing pages.
And then what you want to be doing is two steps. The first step is on that particular
standard reporting in Google Analytics you want to go up top right and choose your organic
search segment. Step number 1.
And you apply it. And now what you're looking at is the top landing pages, all the top landing
pages for search only organic. Second thing you want to be doing is at the
bottom of the table, you will notice a very thin strip that's -- that has a number of
links for allowing you to filter, or go to the next stage and all that stuff at the bottom.
And you'll see a link called "advanced filtering". And I love this feature, because what it allows
you to do is the pool of data you've collected in the report, it allows you to do what I
call, " In-line filtering". So click on that link and it will open a little box at the
bottom allowing you to filter for anything you see in the report. In this case you're
gonna choose the landing page name. Landing page, and add a condition matches
exactly, contains
Nick: The zip.aspx.
Avinash: starts with, ends our whatever. And say zip.aspx, or just say zip and contains
if you want to use. And just hit apply and BAM! Now what you have done in two seconds
is really you've created a customized landing page report, only for the pages that contain
that particular URL, and you have applied your organic search segment. And you have
what you want. So it's a very simple process due to two things. It's really cool. I think
that not a lot of people use this feature in the advance table filtering at the bottom
in order to do their analysis. I was just doing an analysis of all my branded keywords
that get me conversions greater than 15%.
Nick: And you were able to use the filters.
Avinash: In two seconds I had the data
Nick: Real fast, really cool.
Avinash: So that's really a great way to go.
Nick: Great.
Avinash: Here's a good one, my favorite [laughter], well it used to be my favorite report site
overlay.
Nick: Oh, it's your favorite.
[Laughter]
Avinash: It comes from IB
Nick: Your favorite to make fun of, or to actually use for analysis
Avinash: [Laughter] for analysis.
Nick: OK. And IB from Ft. Lauderdale FL. says, "The site overlay report is not showing transaction
goal or revenue data." Oh my. "Even the data is in the regular Analytics reports, how can
-- how exactly does Analytics extract the transaction data for the sire overlay report
to use?" And how can he or she get it to show up Nick? So what's going on here?
Nick: Yeah, it's tricky. So all site overlay does is it creates a report, and it looks
at all the URL's. You know, so, we have a navigation report that says, for a particular
page, what were all the next pages people went to?
Avinash: Right, right.
Nick: And you can look at the navigation summary. So it uses that exact report, and says for
this page, what were all the next pages people went to, and give me all the URL's for that.
And then give me for those different URL's give me the transactions, conversions, clicks
and so forth. So once we have that list and all the metrics, Google Analytics will go
and get your page, render your page and it will go through all the links on your page
and find the next pages that match our report. And then if we match, we show a little pretty
box, and it overlays, if we don't match, we don't see it.
Avinash: Right.
Nick: So there's probably an issue or the report that we're generating is not actually
matching the link that's on your page.
Avinash: Ah, oh I see.
Nick: And that's why it's breaking this.
Avinash: Right, right, so it's the sorta the VI tracks page views or something. So ideally
in this case, they would probably get a GAAC. One of our Google Analytics Authorized Consultants
to come help.
Nick: They can analyze.
Avinash: Consult with them a little bit.
Nick: Right, they can help analyze where points can be breaking in that whole flow, so that
way you're not getting the data. And the other one is if you want cursory information is
looking at the navigation summary cuz it's very similar in terms of the type of data.
Avinash: So look at the navigation summary report, or go get a GAAK, and they can easily
diagnose this issue.
Nick: Help you, right, definitely.
Avinash: Great, great.
Nick: Great, so here's a question for you Avinash. And this is from Hidekei from Japan.
Avinash: Hello Japan.
Nick: It's been really international.
Avinash: Yes.
Nick: What's the difference between Absolute Unique Visitors and Unique Visitors?
Avinash: [Laughter] oh god. Yes.
Nick: I've been puzzled with number of these metrics, which are not always equal to each
other.
Avinash: [Laughter]
Nick: I've been puzzled by the number of type of metric's edited in the same things.
Avinash: Goodness. Goodness.
Nick: The help doesn't make me understand the difference, thanks in advance.
Avinash: Well Hidekei you're not alone. Let me really just first say you're not alone,
and I don't know how to say it in Japanese, but I would have said it in Japanese, you're
not alone. So the metric that was traditionally available in Google Analytics was Absolute
Unique Visitors. And it was only available in one place. Which is if you look at the
dashboard that is available after you click on visitors, and you get a dashboard, those
only -- dash -- place where you could see Absolute Unique Visitors. And it was essentially
the number of distinct, persistent cookies in your, off the visitors who came to your
site. And it was roughly, but not really, but roughly
close to the number of people who might have come to your site. That was the Absolute Unique
Visitors. But in last October, I think at E-Metrics
I had announced that we have a new, we have a much more robust functionality in Google
Analytics to do Unique Visitor reporting in every report, in everywhere which we did not
have before. This new metric was called Unique Visitor,
instead of Absolute Unique Visitor. And this is the metric that is available to
you when you go into custom reporting. It's actually not in any standard report, you have
to go to custom reporting, and you'll see it under the metrics available. You drag it,
apply it in any report. It's one of my favorite metrics to use. So the answer to your question
is there is very slight difference between Absolute Unique Visitors and Unique Visitors.
And what we're gonna do is take the Absolute and deprecate it.
We're gonna send it away to a nice farm where it can live the rest of its life out. We're
gonna -- the Unique Visitor metric will be the one that will be available in Google Analytics
everywhere. So please use the Unique Visitor metric. And the definition is simple, it's
trying to count the distinct persistent cookies, sorry for the technical things. But approximately
the number of people who come to your website. So use Unique Visitor in custom reporting.
And we apologize that at the moment, you've got this funny other thing in the visitor
dashboard. And we'll work on taking care of that.
Nick: And just to say one thing, this metric is very interesting. Most packages will do
daily Unique Visitors.
Avinash: Yes, yes.
Nick: With custom reporting, you can have unlimited, almost unlimited number of reports,
right? You can have certain page, with a certain source of traffic and so for each of these
customizations report will recalculate the unique visitors. So, you should definitely
leverage that metric.
Avinash: I like that, yes. Actually some of the paid web analytics vendors will only provide
you with daily, weekly and monthly unique visitors. They're actually unable to provide
you unique visitors across arbitrary dates. Or for let's say the last four months. You
actually cannot get, even with the paid search web analytics, paid analytics tools in the
marketplace today. You cannot get that unique visitor number. You can in Google Analytics.
And Nick will link to a post in my blog, it's called a standard metric's revisited daily,
weekly and monthly unique visitors. And that will teach you more about the unique visitors
and how they work. But it's a good point in Nick, thanks for adding it.
Nick: Yeah, no problem.
Avinash: Here's one from New York and Aaron, Nick for you. A simple one.
It says how do we get the homepage entrance values to be aggregated. Now my visits are
being reported separately on a /, and /index.php. So this in admin config.
Nick: Right, and in the admin config for Google Analytics, you can specify your default page.
Avinash: Right
Nick: So if your server, when you go to a forward slash, if the server will just by
default go to the index.php page. You can define that in Google Analytics and we'll
dedupe those different reports.
Avinash: Right so just go to the admin settings and you can take care of this.
Nick: Yeah.
Avinash: Oh here's one of my.
Nick: This is one of your favorite ones here.
Avinash: Yeah.
Nick: So, this is from Zack-O in Boulder. I'm making an advance segment for visitors
who spend more than 20 minutes on my site. I use -- so that's a long time.
Avinash: That's a loyal, loyal people. I don't know what you have on your site.
Nick: Probably has some really good videos, we gotta check out this site.
I use this time on site metric with a condition greater than 20.
Avinash: Yes.
Nick: When I run this segment, I see average time on my site is 8 minutes.
Avinash: [Laughter]
Nick: And so how is this possible? Let us know.
Avinash: It's magic.
[Laughter]
Nick: Right.
Avinash: Well Zack, you're making a mistake that I personally made as well -- and probably
corrected -- I'd still keep doing without knowing it. But when you go into your advanced
segmentation, which by the way is the right thing to do there. And you were writing 20
in the greater than, you're using minutes. Google Analytics advanced segmentation actually
uses number of seconds, not minutes. So you would do 20 times 60, and the resulting number
is the one you would put into the advance segment. And then Google Analytics will
Nick: Give you the right number.
Avinash: pick only the sessions where the average, who spent more than 20 minutes on
your website. And fingers crossed, the average you'll see will be greater than 20 minutes.
Nick: And it depends on the content.
Avinash: [Laughter]
Nick: We don't have the link, so we can't tell.
Avinash: But it's a very simple mistake, I think you're just putting a time in minutes
when you should be putting time in seconds. It's a simple one Zack, but a good one. We
should have upgrade GA to provide some helpful hints in advanced segmentation.
Nick: Some help to that, yeah we'll look into that.
Avinash: So Zack won't have problems like this. So that is a good question.
Nick: Great. We also got a feature request from Mark in Bethlehem Pennsylvania. So Mark,
thank you so much for your feature request. We'll pass it on to the team.
But with that we've reached the end of a whirlwind episode of "Web Analytics TV". Hopefully you
all had fun. Please go back to the blog post where this video is posted. Click on the link
for the moderator so you can submit a question for the next episode.
And the rest of you who don't have questions to submit, please look at the questions already
submitted and vote on them because we only -- we answer questions based on the priority
and your vote. So please go ahead and do that.
But with that, happy analytics.