Tip:
Highlight text to annotate it
X
Avinash: Hi everyone. Welcome to Episode 4 of Web Analytics TV -- a rebranding, I think.
Nick and I are going to try and answer your questions about Google Analytics, web analytics,
anything else. So please continue to go submit your questions
on our moderator page, because that's exactly where all these are coming from.
And if you don't have a question, go there and vote for the questions that you would
like us to answer, because we get more questions than we can actually answer.
So please continue to do that. With that, I'll turn it over to Nick to ask
the first question.
Nick: So here's the first question. And that's from Shailendra Dubey from Indore,
India, where I think we have a lot of fans.
Avinash: Yes.
Nick: So. "In all traffic sources sections, currently there are three sections -- direct
traffic, referring sites, and search engines. Is there any way to check traffic from social
media sites rather than creating some filters and advanced segments?"
Avinash: At the moment, actually, there isn't any way.
I think that we have a new social media website popping up every three hours in the world.
So, it's kind of hard to keep track of this imoffers and evolving space.
So the best way is that, unlike many other web analytics tools, Google Analytics allows
you to do advance segmentation on the fly. There's really no limitation. You can create
it. You can go back in history and look at your
data. So the best thing -- even more than using
filters, actually -- is just to create a segment. And as a new website pops us that you would
classify as 'social', simply create an advanced segment and go apply it in the report that
you care about and bam, there you have the data.
Nick: You know, one thing to add to that is, you can actually define campaign tracking
parameters. So you could say, in the link that points
to your site, UTM medium equals social. And so, for some of these URL shortening services,
you can actually add the campaign-tracking parameters in them.
So when somebody clicks the URL, the parameters appear on the final page.
And that way, it's another way to track them as well.
Avinash: Exactly. Actually, atKL.AM, I think is another URL shortener that's doing that.
So many of them. And in that case, they'll show up under the
campaign reports.
Nick: Right, right, right.
Avinash: Okay, cool. So Nick, the next question is for you.
"One question that always distracts my mind. Many times in Google Analytics reports, we
see keywords recorded by visits, bounce rate, average time on site.
But visits are recorded, but those other metrics are not recorded.
In some other cases, for keywords, there is all of the other metrics, but no visits.
So in some cases where keyword. Let's say, I might see zero visits, but I
see the other pieces of data. In other cases, I'm seeing the visit, but
perhaps not the other pieces of data. So when are the rare exceptions when that
does happen, Nick?
Nick: Yeah, so I think the question really -- and we actually see this a lot is, when
you look at ad words in Analytics, the way that we calculate ad words clicks in business
in Google Analytics is different. And so, when you're looking at keywords from
ad words, you might see clicks in cost data, but zero visits, which is weird.
And then, if you see visits, you might see zero clicks in different metrics.
So they might not align. And that's actually a pretty common problem.
And, you know, the challenge that we have is, the way clicks are measured and visits
are measured are very different. And, in the reports, we have to make them
align. So in clicks, whenever you click an ad, that
increments the number of clicks from that source.
For visits, we define a visit as, when the browser closes or 30 minutes of inactivity.
So what that means is, you can come to a site, leave for 20 minutes and return, and you're
still in the same visit.
Avinash: Same session.
Nick: So what happens then when you click an ad, you come to a site, leave, click another
ad, and come back to the site? Well, we have to make the visits -- it's the
same visit, but two clicks. So the visits align, and the first click gets
the keyword in the visit, but the second click gets zero visit.
It's an orphaned keyword source.
Avinash: Exactly. So I guess this happens in those rare cases
where somebody might search. Let's say, they'll search for 'web analytics,
Avinash.' And it'll come to my blog, and they'll see
-- maybe they land in the wrong page. They go back.
And now, maybe they type in 'web analytics engagement.'
And now, they're landing on the right page, and they stay.
So in the rare cases where people flip-flop back and forth between search engines and
the site in a very rapid fashion, then this might occur.
Because I guess a challenge for ad words in Google Analytics is that, it has to give that
credit to one of the clicks. And if it happens in the same session, this
accruance can happen. So that's the behavior that's happening in
your case. So good question.
Nick: Yep. So another question for you Avinash. And this one is from Micah F.K. from Fremont.
"Is there a reason why Google Analytics only displays search keywords under the ad word
section -- under the traffic sources menu -- and not the actual search query as well
as without having to create a profile folder?"
Avinash: I wish I could ask somebody this question, because this is a wonderful question.
And I wish that it would show the actual search way.
This is a request from our users that we've had for a very long time.
And we would love to send this back to the team kind of with your endorsement, and hopefully
in the future, you'll see it. So we're going to send this on to our team.
Thank you so much. I have a question for you, Nick.
"I'm using parentheses, Question mark, parentheses Q equal gobbledygook, something, something
funny. Direct or actual search keywords from PPC."
Great connection between those two.
Nick: Yeah, right, right, right.
Avinash: "However, this contains lots of special character gunk which have to remove in Excel.
Please, could a new URL decode parentheses filter be added to Google Analytics?" It looks
like another feature request.
Nick: Another feature request. It's something that I think it sounds really
good. We'll pass it on to the rest of the product
team. And we'll evaluate it.
Avinash: There you go. So you're helping us create a better version
of Google Analytics right here.
Nick: Right here.
Avinash: [chuckling]
Nick: Okay, so the next question here is for you from Phil from London.
"Does Google organic index use the URL sorting Google Analytics assuming data sharing is
tipped. If yes, does this mean that using G.A. will
help get pages indexed faster for SCO?"
Avinash: Ah, so let me look right into the camera and have a dramatic pause and say,
"No." Pause again. "No. It does not." Google Analytics data -- even if you are doing
anonymous or data sharing is not -- NOT -- used by the SCO team in any way.
Complete differences -- you can see how this would be quite sub-optimal.
Because most websites in the world don't use Google Analytics and there's no -- we would
put them at a disadvantage in any way. So let me have a dramatic pause again.
Say, "No." There's no connection between Google Analytics and SCO and indexing and faster
or slower. No, I'm really sorry, but it's very, very
key. It's a great question. But there should be
no confusion. This does not happen, and it will not happen.
Nick: Great answer.
Avinash: Okay, so Nick -- here's one of my favorites.
I love site search. So here's a question. "Suppose that I've added custom search engine
from Google to my website the co-op search engine.
How can I track search results from the custom search engine through site search in G.A.?
So there are two cases here, Nick, that I would love for you to address.
In one case, what happens in the co-op search engine -- I had it on my blog -- the free
version, where somebody searches -- it actually takes them to Google.com, where the results
are displayed. And it kind of comes back to my site.
So that's case one. Case two is where you can implement a co-op
search engine -- the purchased version -- it's a hundred dollars a year, I think.
Where all the results are displayed inline on your URL on your website.
Nick: I see.
Avinash: "So in those two cases, in each case, can you track and how would you do it?"
Nick: That's a great question to ask. So let's just talk about how the custom search
engine -- well, the internal site search works. So, what it uses is what we call get parameters,
or query parameters. And when there's a certain parameter like
'Q equals a search term', Google Analytics will pull that out and say, "This is actually
coming from the search engine page." And it'll do all the calculations for it.
And that's how you can figure it in the administration screen.
So as long as you could find a way to put the query parameter on the page that the actual
-- the search results are on -- Google Analytics will pick that up and then create all the
reports. So for example, if it's on your page, there's
a way to create virtual page views, where you can add the query parameter and the term
and so forth and fake that and make it look like it works.
And so, if you can get the data into a format that Analytics can read, then we could do
the reports.
Avinash: Yeah, same thing -- in case number one, actually, you cannot track using Google
Analytics or any other web analytics tool, because the person is leaving your website,
going to another site, coming back to your website.
And what happens on the third party site, well, that's not tagged by you.
So Google Analytics or Omniture web ____ anybody wouldn't know.
In the second case, Nick is absolutely right -- easy to track specified parameter in the
admin settings and sing 'happy birthday' and there's your data.
One more for you, Nick. And this is, "How do I replace all instances
of a plus sign with white space? I'm only able to replace first and not all
the instances." Hmm. that's a good one.
Nick: That's a great question. So what happens is, in some of the keywords
report, you might have like tennis plus shoes for tennis shoes.
Avinash: Ah, I see, I see.
Nick: And so, what happens is, in Google Analytics we have this thing called 'profile filter'.
And what you can do is, for any of the fields of data that we collect, you can change the
values in there. You can search and replace as a profile filter
and you could replace 'search for the plus' and replace it with a space.
So, that way, when we start processing the data, the end result doesn't have the pluses.
So it's a simple matter of setting up a profile filter.
Avinash: Profile filter, okay. One more for you, Nick, actually.
"Can we have some information on RegEx implementation within Analytics please?
What can and can't be used in both input and output sections of advanced filters?
This one comes from Andy in the U.K.
Nick: Oh, great. Great question, Andy. Profile filters are pretty complicated, but
very powerful if you understand what they do.
They pretty much allow you to manipulate any of the data we collect before it gets processed.
So, like in the previous question, we could actually replace the plus with a space.
The advanced filter allows you a lot more flexibility.
You can use -- there's two fields -- a 'field A' and a 'field B' that you can use to create
expressions. And then, there's an output that you can use
to manipulate data, and then rewrite any of the fields that we process.
So for example, in field A, you could use any regular expressions.
And you can use parentheses to create a back reference.
And so, when you're in the output field, you can reference the back reference by saying,
"money", the field name -- for example A -- and then, the number of reference.
So Money A1 will actually go back and say, For anything dev match in the first parentheses
in field A and put that data there. And so, what this allows you to do is, rewrite
different fields. It's pretty expressive, yeah.
Avinash: Okay. So, perhaps, what we'll do is, we'll have Nick write a little explanation
in the blog post that contains this video. And we'll also turn a link out to a help article.
Because RegEx is actually very powerful. They can do a lot of wonderful things.
So maybe you could link out to a couple tutorials.
Nick: Sure. I think our friend, Robin Strife, actually put a whole series together on how
to use that.
Avinash: There you go. We'll link out to Robin. Hi, Robin. And that will be great. Okay, good.
Nick: Great. So, question for you, Avinash. And this is Rudolph from the North Pole.
Avinash: [laughing]
Nick: He's getting to his Analytics to see where to put the right packages.
"I need to write a report for my boss which shows user engagement.
Are there any guides or templates on the best way to present this?
Also, is it possible to segment users by behavior?" For example, split the good and bad visitors,
right?
Avinash: [laughter]
Nick: He needs to know where to send lumps of coal vs. packages.
It's a great question.
Avinash: Well, thank you, Rudolph. I have to tell you, this is a bit of a high
horse for me. Because I have --
Nick: So how do you feel about engagement?
Avinash: I know, I know. [Laughter] Well, to be serious for a second, actually,
I really do think that we use the term 'engagement' quite loosely, I think.
Or we play fast and loose with it. And I am sort of on the record saying that
engagement is not a metric -- it's an excuse. And I was like old chapter in my new book
on that particular thing. So, it's important to realize that -- quick
tutorial, okay? -- it's important to realize there's two things
you can measure with web analytics cliksing tools.
Even if you think about engagement, which is the degree of engagement, if you will.
So you can measure how many pages somebody saw, how much time they spent, how often they
came, things like that. But remember what you cannot measure using
cliksing tools is the kind of engagement, which was, that at the end of five pages that
Nick saw was that a good engagement -- he was happy.
That I only saw two pages, but I was happy, and Nick was sad.
And there's no way for you to measure the kind of engagement.
So just be aware that -- don't use the term 'engagement' say, "engage visits are people
who visit our site three times or more." Call it visitor loyalty.
People who sit through the entire movie, say, "time on site more than x minutes or y minutes."
So try not to use the term 'engagement' with users.
Now, that said, it's kind of funny that you're asking about good or bad visitors.
Nick actually has written a wonderful blog post that takes, actually, traffic from my
account for Facebook and Twitter and helps me understand which traffic is of more value.
Let's not call it 'good' or 'bad'. So Nick, you want to talk a little bit --
are you using the API for that, I think?
Nick: Right. So we put example up. You can read it on the blog. We'll link to
it. And it goes out into the API, and it looks
at different segments of traffic. It looks at the tall visits -- people from
Facebook and Twitter. And then, it grabs a bunch of metrics for
each of those segments. And then plots little charts that allow you
to compare on pages per visit, bounce rate -- typical metrics that we call engagement
metrics.
Avinash: Sure.
Nick: And allows you to compare between the sources across a variety of them.
So this actual -- the code for that is open source, and you can use it in your own organization.
Avinash: And you've -- and I've seen the visualization. It's very slick. It's very simple.
And, you know, you could easily slap it on a Powerpoint slide and present it to hippos
and I'm sure they would be impressed. So that was a great job, Nick.
Nick: Great. So here's a question for you. So, "Google Analytics has many reports.
Not all of them provide instant insight. One report that never brought me any interesting
info is the Content Drill Down report. What insights can I get based on this report?"
Avinash: Oh, this is from our buddy, Daniel, in Israel.
Okay, so Daniel, I think this is actually pretty -- to be perfectly honest, I think
there are reports in Google Analytics that I wished would just go away.
I can't think of a time -- like every web analytics tool, I think Google Analytics accumulated
some. I wish there were no report on screen resolutions.
Like, I don't want it. But I actually think that the Content Drill
Down report is probably not one of the useless ones.
And I say that, Daniel, because -- actually, the Content Drill Down report is misnamed,
actually. What it should be called is a Directory Report.
And so, what it does -- it takes the directory structure.
So let's say, Kaushik.net/Avinash/2009/month or Kaushik.net/Avinash/analytics/CompetitiveIntelligence
. I have all the data in directories, so.
And I have a dynamic website. And what this particular report does is, for
people who have clean websites with well-defined directory structures, it actually helps them
understand which content in which directory is being consumed more or less.
So, it's another way to sort of drill down into your data.
So I was analyzing a hospital's website the other day.
And using this report, I was very quickly able to figure out that nobody cared about
their newsletters and health information. What people cared about was their online webinars,
because the directory structure was so clean.
Nick: Right, right, right.
Avinash: Bam -- in two seconds, I could see where the visits are happening the most.
So this is a useful report for those kinds of websites.
If you're truly a dynamic website, this report is not useful for you.
Nick: You need to have a clear directory structure to make this report valuable.
And that could be the reason why you haven't found [inaudible]
Avinash: Exactly. And you are super-smart. And you have no directory structures and so
this probably is not going to be the right one for you.
Nick: Right.
Avinash: Here's a question for you, Nick, from Nozero in New Jersey.
"I'm an e-commerce site. I want to use Google Analytics to have an
affiliate account. Can we delete the G.A. cookie or modify it
based on the date it was dropped? Delete add utm_xyz, if xyz was a source more
than 30 days ago? So he's starting something very complicated.
Nick: Right. So it sounds like what he's trying to do is, change the definition of visitor
or change the duration of what we typically call for a visit --
Avinash: Ah, I see.
Nick: -- and so forth. Generally, it's bad to change one metric on
a condition. Specifically 'visit', because it's tied to
so many other metrics. So one segment of users is actually -- has
a different definition of visits, then all the different reports that rely on the averages
are all going to be weird and off. So just keep that in mind.
Avinash: Even actually the loyalty report, which is my beloved report, shows the distribution
of the number of visits. If you just do this selectively, you would
screw up that report as well -- there's no averages, distributions.
It's one of my favorites. It'll just get corrupt.
Nick: That being said, we have just released three new methods that change the cookie behavior
of the visitor, the session, and the campaign. And that's across all visits, visitors, and
campaigns. And you can use that method.
We'll put a link to the definitions of those. And that will change the length that the cookie
persists for each of those. And that's something that you might want to
take a look at.
Avinash: So there is a way, but use with caution.
Nick: Right.
Avinash: Okay, good, good.
Nick: Definitely. Great. So here's a question for you, Avinash.
Twitter visits are shown as referral. Please can GA.GS be updated to automatically
change those to 'social'? Also, do you think there's conversion higher
fee should also be updated. For example, PPC manual to organic to social
to referral to direct?
Avinash: These are great questions. These are great suggestions, Phil.
I actually personally am not sure if we're quite ready to have a social segment.
The evolution is happening too fast.
Nick: There's too many of the sites. It's hard to keep up.
Avinash: Exactly. So these are great suggestions. We'll keep them in mind.
As we touched earlier in the video -- you can easily create a custom segment that will
allow you to group the data and look at it the way that you would like to.
For now, I think -- I humbly believe -- that that's a more optimal method of doing this.
But of course, Google Analytics keeps releasing new features and new front-end things.
So, we'll keep this in mind and we'll see what would be the best time to do this.
Nick: Definitely. Great answer. So we have one last question here.
And I'm just going to read it verbatim, so we can all hear it and enjoy it.
So five evangelists, the J query team have. "If you had a group of mini me's, would you
be able to win '.net' awards and more effectively take over the world?
And what requirements would need to be DEMONstrated to be eligible to become mini-evangelists?
And this is from Dr. Evil from Volcano Headquarters in the Caribbean Islands.
Avinash: [laughing]
Nick: What do you think?
Avinash: [clap, clap, clap] We've had Rudolph, and we've had Dr. Evil.
This is a great episode I think. Well, it actually takes -- the journey to
web analytics to me -- in all seriousness -- it's actually easy to get started with
web analytics. And the journey becomes tougher and harder
as you go along. And you have to put more and more effort.
But the fruits of your labor become better and better as you go in this journey.
So you can indeed take over the world. Hopefully not for evil purposes, but for good
purposes.
Nick: Don't be evil.
Avinash: Don't be evil. There are many good books out there.
Read those books. And I love tons of blogs. If you go to my blog, and on the right-hand
side -- right hand navigation, at the bottom -- I have my top ten, favorite web analytics
blogs by practitioners who check those out. Specifically for Google Analytics.
Actually, we have a G.A. IQ program, where you can watch videos and review a lot of content.
And then, take the G.A. IQ test and become a certified sort of a Google Analytics person.
But even if you don't actually get the certificate, go through the content.
So between books, the G.A. IQ program, and the blogs, I really think that the knowledge
is out there for each and every one of us to become a mini or a major evangelist of
data. And indeed, I think, ______ rather than sort
of taking over the world, we can become superstars in our company and get promoted faster and
help our companies make more ROI and take our power and use it for good, so.
Nick: There you go.
Avinash: So that's what I encourage you to do.
So, with that, we -- both Nick and I -- wish you a very happy, holiday season.
We'll see you in 2010. Please go to our moderator page and submit
your questions about anything Google Analytics and anything web analytics.
Have a happy new year.
Nick: Take care.