Segmentation, Part 3: Twelve Incredible Advanced Segments Tutorial

3.2 Introduction

Hi everyone, my name is Avinash. I'm the cofounder of Market Motive and the faculty for Web Analytics. I'm really excited to bring you the part three of our video series on analytics data segmentation. This part is perhaps most fun and exciting for me because we've covered the frameworks, we've covered the mental models, we've covered the technical details, we've covered the nuances. And now we get to leave away our basic approaches and do some really cool advanced things and look at a cluster of practical segments from which you will be able to glean inspiration and go apply it in your job almost right away. So I hope you're excited. I'm definitely very excited to share this dozen segments with you. To truly open your mind to the possibilities that exist. We're going to look at these in the three areas that we have established in our other videos. Acquisition, behavior, and outcome. So, we're going to keep following that model that we have already established.

3.3 Acquisition: Focus on a Specific Channel

So let's look at some really cool acquisition segments. So the first one is an example of focusing on a specific channel or comparing multiple channels where we spend money to acquire traffic or we spend effort and energy to acquire traffic. A good simple example of this is to create a segment for your email campaign traffic. So we have a lot of money being spent on email and so we just title our segment. We go on Traffic Sources and we click Filter Visits and we go and pick the name of our email campaign. And in this case that name we're using in our UTM parameters is UTM source equals pronto. That's the name of the email campaign that I am using, and so I go and choose that, and it will select it and put it over there, and boom, I've created a segment for one channel. Now, this is exact same by the way, in which you could create a segment for social media. From the dropdown, I am running a lot of campaigns. For social media, I use the term social media in the source. utmsource = social media. And I would click this. I would put it into this box and that would create a segment for one particular channel. But in this case we're going to focus on email, so we pick pronto and we click on save and we have our email campaign. You going to apply this campaign in many different places in very creative ways. For example, while we send emails for a very specific purpose, I'm always very curious what people do or what their intentions are different from what I had send the email campaign for. So when they land on the side, one of the reports I like to look at is my Internal site search report, and that's available easily in my analytics tool. So I go and apply my email camping traffic, and these are the top ten terms being used by people who come on email campaigns to my web site, and these terms give me very good clues about what my email traffic is interested in, because these are the words they are actually typing into the search engine on my web site. It's a great way to understand intent. And the difference between the campaigns I'm running and what people are actually looking for. So for example, I have not thought that action analytics is something that people would search for, or for that matter CXO because I very rarely write anything for CXO's. But that's what people are looking for. So this is a really good way to use Email Campaign Traffic and apply it to a non normal report because of course, normally you would apply it to your visits report and your conversion report. And you should do all that. But this is sort of an off the normal example of something you could do with your email campaign traffic. But of course you can do more. You can also create a segment for your social traffic. So it is important to understand that when it comes to social media, you should be creating two types of segments. One is social media traffic that you are driving to your website, and one is a social media segment for traffic that other people are driving to your website. So whenever you tweet URLs, you Google Plus URLs, you Facebook your own URLs, make sure that you use UTM parameters in those URLs. Or if you're using Site Catalyst, use the [INAUDIBLE] parameters from Site Catalyst, the CIDs, etc. Or when you use Web Trans, you use Its parameters. Whatever parameters are used for campaigns, use that when you share links on social media. And so in this case, my segment is self driven, which means I am driving the social media traffic. And I pick the UTM parameter I'm using, which is social media. And I'm going to choose that because that's my UTM parameter I use for my self driven campaigns. But if I was trying to create a segment for other people's ending traffic, I would probably pick this t.co. Among other things, for Twitter, I would use this as my source. Because that's traffic that I'm not driving, other people are driving. Especially on Twitter, but then I can things, the Facebook URLs, etc. And that would give me my non self driven traffic site. I saved that segment, in this case, I'm saying social media, source contained social media. I test the segment and make sure there are lots of people who are in that, and I hit save. I make sure that I've clicked on Filter Visits, because I want visits in this case, and not users. And I'm now able to apply that segment. So in this case now, you can see that I'm applying my email campaigns and I'm applying my self driven social campaigns at the top. I'm able to see how much traffic comes from it and I'm able to see the trends and patterns in the behavioral data and for example I can notice anomalies like my email campaigns always lag by a couple of days. The peak that I see from social traffic so it seems that people are discovering my content in social media while before I'm actually sending out the emails and that probably should be the reverse in the sense that my email campaign should be helping drive a lot of social amplification but that's actually reverse here so it's going to get me to rethink what time periods I'm sending my email campaigns and because I want that halo to b created in social media. So you can see how this one simple graph, and comparison helps me find some really interesting insights. Of course, I can then apply those segments to my conversion reports, because we want to be making money. And even though this is not an E commerce website, I'm able to focus on conversion rates. I'm able to focus on specific goals. I'm able to focus where email is doing better than social media, and where social media might be doing not as well, etc, so I can do all of those things, and I can find anomalies. So for example, while I have a lot more conversions from social media for this particular microconversion, you can notice that email drives more microconversion subscribers, then the social media site gives me a clue as to what people are looking for. And when it comes to speaking engagements type microconversions, email is a lot poorer at that and social traffic is substantially multiple times higher. The differences are very close for all of these other Microconversions I have for goal two, goal five. But for this one it's vast difference. And so again, it gives you a clue about the intent of the audience that is coming to your website. So this is an example of comparing two different segments. But of course, not only can you create advance segments that are focused on the overall source. All social media, all email campaigns. But you can actually create a segment that only focuses on one particular email campaign. So in this case I can say go to traffic sources, I can say visits, I can say campaigns, contains and I can pick this particular campaign I had done in the month of June, and I just choose that, and it allows me to focus my analysis on just that one email campaign. And then I can compare the performance of my May campaigns with my June campaigns as an example or my March, April, May, June campaign separately to see performance so you can actually focus on one particular campaign or in this case I can say I would like to only focus my analysis on the traffic I am driving from Google Plus or from Twitter or from Facebook, they're all here, right? And so I can just choose one of those and then just focus on that analysis and that allows me to compare Google Plus and Facebook, or it allows me to compare my twitter. With my RSS, with my mill campaigns, one particular campaign I ran during one time period, so there's a lot of power when it comes to you creating channel specific segments and unleashing the power of acquisition.

3.4 Acquisition: Long-Tail/Multiple-Keyword Search

The second one I wanted to share with you is a bit more complex but really powerful. It's just focusing on your long tail or multi-keywords search course. Very often, we worry about our brand name or we worry about just one or two keywords. But the reality is very few people search on one or two keywords. Most people search on multiple key words because they realize search engines have become very smart and being able to answer questions. So literally people are asking questions of search engines and you can use the power of segmentation to understand that. As an example, a common query that people run to come to my blog is that they might go to Bing and they might type in advanced segmentation avinash download, because I have a lot of advanced segments I've created on my blog that you can just download from there. So you don't have to create it and you will open it in your Google Analytics account. And I noticed that a lot of people search for those. So if you search for this, and come to the blog, I would like to know what happens across four different words, and not just one word, avinash, or two words. Avinash. Those are the things that now normally people might end up obsessing about. So how do you focus on the phrases with such a number of words in them. It's actually really easy. You use the power of regular expressions and the way that it works is, you of course title in segment. In this case, I would like to focus on three or more words in a search query then I go to choose filter visits because I would like to focus on visits that were driven by multiple words in a search query and then I use matches regular expression. As my condition, I'm going to apply to my keyword, and then I type in this right regular expression that you see on the screen. This allows me to filter for only the search queries with more than three words in it. And I'm going to apply that to my keyword report and boom, there you go. As you can see, my report only shows me keywords now. With three or more words in the search query, so you can focus on that, right. So I can see some words here that I had not thought people would ever type. 1785 people typed in avinash kaushik hippo because I was one of the originators of the definition of hippo heist. [INAUDIBLE] person's opinion. But I never thought that that's how people type it in. Customer lifetime value makes sense but accuracy versus precision was something that I hadn't thought about. And so I exposes words for me and searches that I had not thought about. Of course I can scroll in the report and now I'm looking at five, six, seven, eight key words the people are typing. And as you can just take a cursory glance at this and it is so humbling, because there is no way that I would ever have guessed that these are the search queries that people are typing to find my content. And this happens for every single website. We get so obsessed about the top 10, 20, 30 keywords, but we don't focus our SEO efforts and PPC efforts on the opportunity that exists in the long tale, and I think that it's really criminal behavior, because there is so much gold to be unlocked here. And the amazing things that is less competition in the long tale which means you can actually rank better and get a lot more traffic to your website which is a very, very good strategy for you to execute. And the way that you would do is, you would use the power of regular expression. So, here are some of the regular expressions that you can use in order to filter down to particular segments with words in them. So as you can see if you use this particular segment and have two in it, then this particular segment because of the way it is constructed, will give you exactly matching three words in a search query. But by adding this little comma that you see here then you are changing the segment to say I would like to have all the keywords where there were three plus keywords, rather than just three which we got here. Or at the very bottom you can see how I can create a segment where I'm saying I would like you to create a segment for me with 2 or 3 words in the search query. So you have a lot of flexibility and power in using regular expressions to focus on your long. All of this does use regular expressions, and it's important for you to have the knowledge of regular expressions. And I had shared earlier that there is a handy dandy guide that you can use in order to focus on these regular expressions, and you can go look at that and download it from the web. And the download is available at this URL, zqi.me/regexga. If you type in that URL, the PDF will download onto your computer and and you can use that to learn how to create regular expressions. So that's our long tale segment.

3.5 Acquisition: Cohort (Retention) Analysis

The third one is to do cohort or retention analysis. This is really advanced and really cool acquisition segment that you can create. In this case, what we're trying to say is we would like to focus on people who came to our site between February 1st and February 28th. So we want to focus on that segment. And we would like to only focus on people who came from paid search. So we have created a cohort of people who came during this time period and my paid search campaigns. And the analysis that I want to do is I want to know if people who first came to my site during this time period just came during this period from the paid search campaigns that I ran, or they actually came back later and I was able to retain it. You are able to easily create this segment. And the second segment I've created, exact same time period, but in this case I've said I would like the cohort to include everybody who comes from organic search. So exact same time period, I want to know am I better at retaining traffic when I do paid campaigns or organic campaigns, which are not paid. And I can go back and apply it to my report, I'm looking at a 90 day period after my cohort was acquired and I can see that I do a slightly better job of retaining people in the following two weeks after my campaign, or even three weeks after my campaign has run. But I do a much worse job when it comes to paid search. In fact, once I stop running search ads their number quickly goes down to zero. I'm unable to retain the traffic on my website. I'm doing a little bit better with organic, but this is absolutely a shame, right. If you are spending so much money acquiring traffic you should create a micro conversion and engagement. A relationship with the client so you don't have to go back and pay for them again to come to your site. So cohort analysis is really really good at helping you understand in this scenario how good a job you are doing of retaining the valuable traffic that you're requiring. And now I would go back to my organic traffic and I will try and understand what am I doing in terms of retaining this organic segment. In fact now I can say, I would like here the people who came in the month of February. And then I'm going to apply that to the month of March, and see when they came back in March what were they doing on my website. And I can use that to go back and optimize my paid search campaigns. You can do this kind of analysis and do some very cool stuff in addition to what you have seen already. So in this case we are creating a segment for people we acquired during March. So in this case, the first visit is March. And the page search is easy to see if we did better in March than we did in February. And so now you can see the two graphs where I'm only looking at paid search. We actually spend a little bit less money in paid search, as you can see, during this line much lower than the one in February. But actually we did a much better job of retaining this traffic than we had done with the people we acquired in February. So the delta here that you see versus the delta you see here helps me understand that while I spent less money on paid search in the month of March, I actually ended up acquiring traffic that lasted with me longer. In fact some of these visits that came from March actually were from people who had visited my website in February. I just got to them with paid search again and managed to retain them over a time period. And then I can focus on other visits that they might be making to my website. So that's why when I apply to this report you see that my segment only focuses on CPC, but these same people actually came back to my site on organic, they came back to my site directly, they came back on my email newsletter. So I'm seeing that multiple channels are required to get to the same people. And this is where cohort analysis is so good. And helping you understand retention as well as multiple engagement points that are happening with your website.

3.6 Behavior: Content Consumption

Now let's look at behavior analysis. Let's look at behavior segments. Behavior is of course everything that happens after people land on our website. So the first segment we're going to look at is content consumption. So in this case, we're going to look at people who focus on one particular piece of content. So in this case, we're going to say we would like to filter for condition page. So we click on the drop down here. And we type in page, and we're going to pick the page. And we're going to focus on everybody who sees the toys new arrivals page on our website. Because we have a special page for toys new arrivals, and what we want to know is do people who see that page end up buying more, or buying the products that we are recommending, or coming back to the site again and again, or etc., etc., etc.? Because that might be a good place for them to visit us multiple times, so we go and pick the toy new arrivals page. And as you see, when I'm typing it in, Google Analytics automatically suggests a page. I don't need the whole thing, I can just pick the stem that I'm interested in. So page contains this and I hit save. And I have my segment ready to go. I can also test it to make sure that I've created a segment properly, and it says indeed I have, because 1.81% of the users to the site saw that segment.

3.7 Behavior: Specific Content Purchase Influence

The second one we could create for a content area is to understand the specific content purchase influence. There's one piece of content influence a particular purchase. In this case, what we're going to do is we're going to say that we are going to build on top of our new toys arrivals page content that people are seeing. So in this case, we already have that segment created that we have applied in this process. And then we're going to focus on the revenue where revenue is greater than one dollar. So essentially we are saying in any session where people just converted, show them to me. That's really what you're saying. And so, we do that and we can now go back and apply it to our report. And so you can see the conversion rate that happens for people who see the toy and new purchase page. At the moment you can see that the people who just made a purchase versus people who visited the page. What is the difference between the two. And you can see that people who see the new toys convert at a 70% conversion rate versus a 30% conversion rate for all people who make a purchase. Who may or may not have gone to this section and we can further specify and clarify this segment to ensure that to make the purchase segment only includes people who did not visit this page. In this case it was so low that it doesn't really matter. We can see how many transactions there were, we can look at the quantity, we can look at the unique purchases, we can look at the order value, etc, etc, etc. And see what that particular page, or a particular video on your site, or a particular car configurator tool that you have created, or any particular tool that you have on your website. And its impact on a business is easy to identify using a segment that we can create that looks akin to this one.

3.8 Behavior: Home Page Promo Effectiveness

The third one that we can create is for the Home Page Promo Effectiveness. This is one of my favorite segments, that really helps you understand something that we normally don't focus on. So, let's assume that we work at the NorthFace, and this is our website. And you can see that, on the homepage, we have a promotion for day packs. So, there are many ways to, of course, get to the day packs page. We could click on this link here. We could go to Equipment, and scroll down to day packs. We could come directly to the page. There are many different ways in which we can get people to the day packs page. Now the interesting question that somebody might ask is what is the effectiveness of showing this particular promotion on the home page? Does it really help us sell more day packs? Versus if we did not do this promotion. And that's a very good thing to think about. So most of the time real estate on our homepage is severely limited. We can't actually have as much real estate as everybody would like to see. So in that limited real estate scenario it is very important to understand the effectiveness of the promotions that we do and this segment allows you to understand that so what we're going to do in this case is use a sequence segment. So what we're going to do is we're going to create a segment. We're naming it homepage and day pack. And were going to say the first step is were going to pick a landing page, were going to say contains and this is the stem of the URL. I simply copied it from the browser by removing everything before this, where the .com stuff is. And everything, the stems that might be here. Parameters I just got rid of it and just a landing page contains this. And we are saying, we would like the first page in the experience to be the homepage of the website. And then we can say, and it is followed by, and you can say, page two contains this particular page. So what we're saying is people who go from the home page to go to the day packs page. This link you see is the day packs page. And what we can then do, is say we are going to compare this segment of people to people who enter on the day pack page. So here we're saying that step one Landing page contains this. So we have a segment called daypacks entrances. They've got a segment called Home Page to Daypacks Behavior. And we can compare the behavior of these two segments to try and understand If the engagement on the side, if the micro conversions, if the macro conversions are higher or lower in the scenarios where people are seeing the day pack promotion in the home page or not. So this is a really great way to create a sequence segment and use it to understand performance of your homepage promos.

3.9 Behavior: Frequent Visitors

The fourth segment in the behavior area is frequent visitors. So this is of course, we always have the segments for new and returning visitors and I do dislike them to some extent, because while new visitor itself is value. Returning visitors, I think is a cop-out in a way. Because what it shows us is that there is a general amorphous glob of people that we're calling returning and we don't know anything specific about their behavior. So we are going to go and try to understand what they're doing and the insights are very poor. So in this case, what we're going to do is we're going to say, we're going to focus on a specific behavior when it comes to returning visitors, a specific behavior. And the behavior we're saying is in this case, we would like visitors who have made ten visits to our website during this time period. Now, ten is a lot. Ten is a lot, but I'm just trying to make a point that rather than treating everybody who came twice and keep treating everybody who came five times and treating everybody who came 100 times the same, makes no sense at all. In this case we're saying, we want to focus on visits that are greater than ten. And now this is really cool, because these are people that really frequently come to our website and what we want to know is hey, as a result of their visit to our website, is there anything of value that gets created for them or for us? Additionally, what we can say is I would also like to know about the visits, people who have visits greater than ten, but have not visited in the last four days. Because let's say, I want to send them a campaign or a promotion or something in the last four days and so we can add that into the mix. And we can add other parameters, if they would really like to. But in this case, we're focusing on people who usually visit a lot, but recently have not visited a lot so we're going to focus on that and you can go and analyze their behavior. This is so much better than actually focusing on returning visitor, because returning doesn't make any sense. It is quite easy then for you to say, I would like to focus on people who come between three and five times during a month or I want to focus on people who visit more than six times and have made a purchase. Or I would like to know people who visit more than ten times and I'm going to apply it to my all traffic sources report to understand what drove them to my website originally, etc. So this is a really great behavior segment to analyze and helps you understand how engaging et cetera, the content on your website is.

3.10 Outcome: Specific Product Buyers

The third category that we want to look at are outcomes segments. I think that a lot of people focus on conversions, et cetera but I often don't see as much creativity in their analysis as might be optimal. So we're going to focus on some really cool examples of outcome segments that can help you analyze the data. So the first one is to focus on specific product buyers. This is very interesting, right? So we have a lot of products that we are selling, and some products are worth more than other products, and some products have higher margin than other products, or more popular than other products, or not, etc, etc. So in this case, what we're trying to figure out is can we focus on analysis of just particular products. And so in this case what we go in and do is go to the e-commerce area and we are going to focus on the sale of pants. Everybody who comes to our site and buys pants and so you can see we're saying product contains pants. We could be selling many different kinds of pants. So in this case we might say, we can also sub segment by corduroy pants or we can sub segment by some other things, but in this case we're saying we're just interested in the pants category, shirts or underwear or whatever we're interested in. And so we select that and we save our segment and we can go and analyze all kinds of things about people who are buying pants. Where do they come from? What is their average order size? What is the traffic source? What are the campaigns and key words? What pages do they see? What is the average duration, visits to conversion? How many pages per visit? So many different things we can analyze by focusing on a particular product rather than all revenue.

3.11 Outcome: Multiple Items (+ Combo)

A second category we can look at are people who buy multiple items. I'm going to show you the segment. I'm going to also create a combo segment. I'm going to combine different things. It's going to be really exciting, okay? So in this case we're saying we are really interested in people who buy more than five quantity of items in a session during one visit. So we have a filter on Quantity, then we're saying greater than 5. So, very simple, visits during which quantity of purchase in that visit was greater than five. So that's a really, really cool way to identify people who buy a lot of products. But we won't stop there. We're going to go to traffic sources after having applied our condition. We're going to go to traffic sources and we're going to add one more segment to the mix, and that segment will be we would like to focus on people who only come from Google. Because at the moment Google's very important to us, and so we're focusing on Google. But of course, you could just as easily focus on Facebook or you could focus on email campaigns or affiliates. It doesn't really matter. But I wanted to use this as an example to show you how you can make your segmentation even more focused by having this combination strategy where the first segment we've created is the people who buy more products in one visit, five in this case. And then the second one is we've created a segment and added Google to it. And now we can apply this segment to our analysis and have such a powerful understanding of an audience that we should love and adore from a source where we're spending a lot of attention. And we can go back and see if our search engine optimization paid search and other kinds of analysis is good in this case. So that's a really, really great segment. And it's a combination segment because it focuses on two things that are important to our business.

3.12 Outcome: Short or Long Purchase Cycles

The third example is long or short purchase cycles. Clearly this is really really important because some people buy right away and other people buy a lot later. So it's really important for us to understand what is happening in terms of driving people to buy right away versus people who are going to take longer amounts of time to buy from us, so let's create this segment. So in this case what we're doing is we're saying that we are interested in people who take two days to convert on our website. And it seems about normal so we're going to say we're interested in people who take two days to convert. So we have the days to transaction greater than two. Right? So we're saying people who take more than two days to convert. And then we click a test button to make sure that our segment is created properly and we need to have some people in that category. And then we create another segment which would say, we get people who take more than 30 days to convert. And then notice that there are indeed people, who we can see by clicking on the test button, who take more than thirty days to convert. That's really crazy right? Now what we've done is we are focusing on two segments where we can learn what kind of products are different between people who take greater than two or by the way you can quickly flip this thing. You can flick this thing less than two, so we can say less than two and greater than 30. Of course you have full flexibility here, in terms of the numbers and the conditions you use, but this would allow us to understand all the particular products that are longer consideration cycles. Other people in particular geographies that take more or less time to covert. Other campaigns we're targeting that have longer or shorter conversion cycles and you can use them to do everything. You can use that to optimize your landing pages so that it's not buy now, buy now, buy now! For people who are taking greater than thirty days, and you know, have calls to action, clearer calls to action for people to take less than two days, so you can use greater than or less than. You can do greater than or equal to or in between. So many different options you have. But this is a great way to focus on the purchase cycle. And using outcomes to determine how best to optimize your content and how best to optimize the acquisition that you have inside your company. So, really, really cool way to draw this contrast between two different kinds of outcomes that get delivered on your website.

3.13 Outcome: Micro-Conversions (+ Combo)

The fourth segment is also a combination segment where we're going to have two different things. And it focuses on micro conversions, because we've obsessed enough about e-commerce and macro conversions. So in this case where the segment that I've created uses goal value. And so these are microconversions for which I have assigned goals. And so we're going to say goal value in each session, is greater than $2. So we're saying we have achieved an economic value of at least $2 during the session. We would like to focus on those sessions, visits to our site, and we would like to focus on the country of Germany. So you can filter people who come from the country of Germany. And you can test and you can see that indeed there are a bunch of people who fall in that criteria. And this allows us to then go back and optimize our strategy for acquisition and content for the country of Germany. So it's a really, really great way to analyze the behavior and focusing on outcomes that get delivered on our site. So there's the country, and we can say Germany, and we can really, really focus on what we're trying to do.

3.14 Acquisition + Behavior + Outcome!

The last segment brings everything together. I just wanted to get to blow your mind and tell you that while I'm about to attempt this crazy. So boys and girls do not do this at home. [LAUGH] But I wanted to show you that when you have particularly tough questions in your business, you can use segmentation to really do even crazy things. Things that you would otherwise think are not possible. So let's bring all of this together in one place. So here is the complex segment that we have created. So the first thing that we're going to do is we're going to say, we would like to focus on people who are from the Americas. So, from North America and South America so that's the first segment we've created. That's a sourced segment so that's a good way to think about an acquisition segment that we have. It's all people from the Americas and then we are saying that we would like to focus on people who use mobile traffic, right, so who are on mobile traffic. So then, it allows us to understand, as we discussed, this could either be an acquisition segment or behavior segment, depending on your target. In our case, we are running specific campaigns targeted to Americas, for mobile, and so this is an acquisition segment that we're focusing on. And then we are saying, in addition to that, for our mobile, because we've done something very special for them, we are going to focus on people who are visiting our website for more than 10 seconds. Usually bounce rate is measure by people who see only one page view and no more, but in this case I automatically said more than 10 seconds, which means it is my way of defining bounce in this scenario. Where if you stay on the site for ten seconds, it would mean that you at least saw two pages. Because without seeing two pages, time during your session does not get recorded in any web analytics tool. So I'm saying the behavior I would like to analyze is people who at least stay on the site for ten seconds, which means at least they have seen minimum of two pages maybe even more. And then I'm saying I would like for the traffic to be one where I'm spending my money on the mobile search advertising campaigns. So in this scenario, we are going to focus on people who come on our search advertising campaigns where the key phrases they are typing have more than five words in them, which means there is a degree of specificity to what they're looking for and what we are going to provide them with. And lastly, we're saying we're going to focus on this traffic. This segment we want our people who have at least completed one goal on our website so you could see the last one here. Goal completion per sessions greater than one. So at least one goal during the session, right. So you could see how in this case we have acquisition, behavior and outcomes all combined into one segment. And in that segment it allows us to have very sharp focus on our Americas mobile search ad campaigns that are targeting key phrases typed by people which use five words or more, who come to our site on a mobile platform and also deliver at least one micro-conversion, so it's really great. And you can see while I have created a really insane, if you will, segment there are indeed people in my business who meet this criteria and are indeed converting on our website. So you can see now I am able to see what keywords these people are typing, and you can see every one of them. Grid and and you can see their keywords. I can see the number of visits, duration, and etc. And as you can see, the bounce rate is 0% because I wanted people who were not bouncing. And I should not have been worried about using the time and wondering if it was only people who saw two pages because in this case, you can see that the average for this audience, on a mobile platform no less, is 11.75 pages during the visit on a mobile platform where lots of content is not very easy to read. And so these are my loyalists of the loyalist of the loyalist people even if there are few of them but it shows me the kind of content, even in a mobile platform, that causes people to see 19 pages, et cetera. In fact, this person, this is a nuance, this person has been on the site, this one person, how to do web analytics, this person has been on the site for three hours and eight minutes. Three hours and eight minutes. That's amazing, right, that's amazing. To think about and remember in Google Analytics as is the case in omniture, Site Catalyst, Adobe, IBM, etc if you open a website, if you leave it and don't touch the browser or the page for 29 minutes, it terminates your session. It will close your session, because it knows that you're not engaging with the site anymore. So, in this case, what this means is that this person is actually active for three hours and eight minutes, because their session was not terminated. Google analytics in this case detected that they're on the site, they're engaging, they're reading content. There had 19 pages during the time period. And essentially using my blog as a book I suppose. But again it shows you how even from these quote unquote crazy segments you can find some really cool insights and find areas and learnings that you can use. To make the experience better for other people and also focus on creating new programs and marketing activities that will drive higher outcomes for your business through this deep engagement.

3.15 One More Thing

Amazing isn't it? Really really amazing. I want close by one other quick thing. Throughout this CD on segmentation, we are focused a lot on tools like Yahoo Analytics and Google Analytics and we've talked about omniture and etc. But the reality of the matter is segmentation is everywhere, in everything you do. It's not just when you're in Google Analytics. For example, this is a very common report that people have. In this case, we have an Excel spreadsheet that we have created. In this case it is for the month of August. And for the month of August what we are showing is that there were certain number of people who visited different types of websites. So we're showing the number of visitors to the site and the number of visits to the site in the first row and so that's nice we know the distribution of where people go. But even this data in aggregate is quite useless, even if it isn't segmented by sub domain, support site, and Microsoft, right. because what you could do to improve this particular table is to segment the data, and by that, what I mean is that before the total visitors. You could simply segment the ones that purchased and now you suddenly you have two rows of data that give you amazing insights. So we got 5 million visits to the site, but of that only 12,900 converted and that's like such an amazingly humbling number if you will. While on our sub-domain which looked pretty paltry compared to our main site for 5 million visits. In this case we are able to show that they received 1.9 million visit but drove 84,843 conversions. So you can see how a simple segmentation allows you to give a really great insight to what was once a lonely row in Excel. Underneath it, when we see unique visitors, you can see that we have segmented it by direct traffic and ratio of new and returning. Not just returning but this ratio of new and returning. It gives a sense for what is going on in the website so do simple segments applied, that give you some really, really great insights that you can use in your analysis and make things smarter for your business owners. Here's another example, one of my favorites, where you can see that this is a nice little graph, let's say it's revenue, it could be profit, it could be economic value, anything really. But let's say it is revenue and you can see that this graph is very lonely because all it says is revenue is up and down. We can make this graph better by applying segmentation. So in this case what we are doing is changing the graph a little bit and now we are able to see segmented data. Where we can see the direct is the main driver of the traffic. We have a slightly better sense for where a big component of our traffic comes from, then we can see that affiliates are very important to us. And in this case, YouTube was so important, even in some months more important than affiliates and then is paid search. Obviously, then you begin to ask the question, well, what about organic search? And organic search doesn't even show up on the horizon. It's like, wait, why don't we have any conversions for organic search? Or, that they are so small that they are in the other packet? Let's go fix it. So you can see how this lonely graph doesn't raise a lot of good questions, but this graft that is segmented raises a lot of really good questions and help us come up with some really interesting and valuable insights from our performance.

3.16 Example: Getty Museum

And my last example comes from the Getty Museum in Los Angeles. If you've never been, you should go, it's really a wonderful museum. We've been a couple of times. It's really a great place, but one of the things Getty Museum wanted to know from its digital business is they wanted to understand, are we focused on producing the right kind of content, content that people are interested in? And so they created this, rather than looking at a graph that looks like this, where we're just showing how many page views we had and how many visits. I mean this is kind of not very useful. We created a segmented graph. Where we are able to track two different things. For each of the categories Wetgetty focuses on, education research, collections, conservation, and grants, we are able to focus on how much content is in each area, which is the green bars you see. And how many people consume that content, which is the blue bar we saw. So we can see that we have an enormous amount of educational content, and indeed a bunch of people consume it, 23%, but it's nowhere near to the amount of effort we're putting in to producing all this content. And while look at the research area, it is the highest place where people are going and checking information, but we have such little content in that area just 4% of content in that area. So it helps us understand that maybe we want to focus some resources away from education, and focus them on research, so we can create more content. Clearly our audiences very, very interested in that content. So that's a really, really good way to focus on understanding how to optimize the content that we have and our content strategy for our website. So these are three simple examples of knowing that segmentation should be everywhere in everything you do. Rather than just inside Google Analytics or just inside catalyst.

3.17 Thank You

There is nothing, nothing that is more powerful than segmentation when it comes to going from data to information, and for information to insights. Segmentation forces you to understand business goals. Segmentation forces you to focus on particular areas that are of value to the business and drive specific insights that can take specific action, which is a thing that web analysts so struggle with. I hope that you found the three videos to be of value. I hope you found our last dozen segments to be inspiring and that you're going to go and create your own segments that will help you truly rock at your job, and accomplish things otherwise would have been very difficult for you to accomplish.

3.17 Thank You

There is nothing, nothing that is more powerful than segmentation when it comes to going from data to information, and for information to insights. Segmentation forces you to understand business goals. Segmentation forces you to focus on particular areas that are of value to the business and drive specific insights that can take specific action, which is a thing that web analysts so struggle with. I hope that you found the three videos to be of value. I hope you found our last dozen segments to be inspiring and that you're going to go and create your own segments that will help you truly rock at your job, and accomplish things otherwise would have been very difficult for you to accomplish.

  • Disclaimer
  • PMP, PMI, PMBOK, CAPM, PgMP, PfMP, ACP, PBA, RMP, SP, and OPM3 are registered marks of the Project Management Institute, Inc.

We use cookies on this site for functional and analytical purposes. By using the site, you agree to be cookied and to our Terms of Use. Find out more

Request more information

For individuals
For business
Name*
Email*
Phone Number*
Your Message (Optional)

By proceeding, you agree to our Terms of Use and Privacy Policy

We are looking into your query.
Our consultants will get in touch with you soon.

A Simplilearn representative will get back to you in one business day.

First Name*
Last Name*
Email*
Phone Number*
Company*
Job Title*

By proceeding, you agree to our Terms of Use and Privacy Policy