The Definitive guide to Cohort analysis and Lifetime Value report in Google Analytics.
Let's admit it, whether you are a marketer, analyst or business owner, we all have huge obsessions with acquiring lots and lots of traffic to our website. And it's a totally a legitimate thing to do. But nothing in this world is free. So we spent enormous amounts of money on search and display ads, ads on TV, newspaper, radio etc.
While doing all this, most of us forget one very important thing. We forget to measure who our real customers are? Who is going to stick with us when time gets tough? We forget to understand what a real success looks like and how to measure it? We measure success based on conversion rate, cost of acquisition or even worse visits to our website. But all of these will only get us to the moon. But what we really want is to go beyond our own galaxy. Cohort analysis and Customer Lifetime Value analysis helps you to do that. It helps us to find the channels, medium or sources which bring customers that create value for your business in long-term. It helps us to separate the customers who only flirt with us to the customers who give their valuable time, attention and money to our business.
So if you are excited to learn and get most out of Cohort analysis and Lifetime value analysis, then keep reading this post until the end. It's for you.
Things covered in this post -
1. What is cohort analysis?
2. Understanding the settings and features of cohort analysis in GA.
3. How to do Cohort analysis in Google Analytics?
4. Customer Lifetime value report in Google Analytics.
5. Understanding the settings and features of Lifetime value report.
6. How to analyze the Lifetime value report in Google Analytics?
[ Important Announcement ] - Dear, in our upcoming posts, we are going to explain each and every report in Google Analytics. So if you interested to learn and enhance your knowledge of Google Analytics, then visit the website later after few days or subscribe to our blog to get it right in your inbox as soon as it gets published. Don't miss out on it.
1. What is Cohort Analysis?A Cohort is a group/segment of users who share common characteristics that occur within a defined period of time and it is done to understand and compare user behaviors over a certain time period. A cohort is simply just another way of segmenting users but it is based on date.
Examples of Cohorts - All users who are acquired during the Thanksgiving and Christmas holiday shopping season. Users who visited the website during 1 Dec - 31 Dec via organic search.
2. Understanding the settings and features of Cohort analysis report.
|Click on any pics to see in Higher resolution|
The new Cohort analysis report in Google Analytics has 3 main sections -
Section 1 - Menus(Settings) to create the Cohort Report.
Section 2 - The Line Graph.
Section 3 - The Tabular Data.
Section 1 - Menus(Settings) to create the Cohort Report -
A. Cohort Type -
Right now, the Cohort report in Google Analytics is in Beta version, so at this moment there is only one option here i.e. Acquisition Date. But don't worry, we can do a lot of great things with it. Just stick with us.
Acquisition Date is simply the date when your users first visited your website or App or when you acquired the users.
B. Cohort Size -
This determines the size of each cohort. It helps you to specify how do you want to group each Cohort? Do you want them to group by Day, Week or Month?
By Day - means that I want to include all the users who have the same acquisition day. Like, users whom you acquired on 10 Dec will be one cohort and 11 Dec will be another cohort and so on.
By Week - means all the users who were acquired during the same week. Like week 1,2,3 etc.
By Month - you get that.
C. Metric - The metric that you want to evaluate.
All the metrics in Cohort analysis report is categorized into three categories.
1. Per User - Like Goal completion per user, Revenue per user, Transactions per user etc.
2. Retention - User Retention
3. Total - Goal completion, Revenue, Transactions etc.
By Default, the user retention is selected in the cohort report.
What is User Retention? -
The number of users in the cohort who returned in the Nth time period ( Day, week, month) divided by the total number of users in the Cohort.
D. Date Range -
The time frame that determines what data will appear in the report.
The values of date range depend on the "Cohort Size" selected.
As you can see that I have selected the Cohort Size by week so my options are Last week, last 3 weeks, Last 6 weeks etc.
and If I change the cohort size by day then my options will be last 7 days, last 14, 21, 30 days.
Suppose if I choose a Date range of last 7 days as shown in the above pic and today is 5 Jan. Google analytics will start to count 7 days from yesterday and pulled out the data from the past 7 days.
|courtesy of Justin Cutroni|
As you can see that -
Day 0 - Dec 29
Day 1 - Dec 30
Day 2 - Dec 31
Day 3 - Jan 1
Day 4 - Jan 2
Day 5 - Jan 3
Day 6 - Jan 4
Day 7 - Jan 5
3. How to do Cohort analysis in Google Analytics? -
To make our analysis more effective and find better insights, we need to take certain steps.
What - First, we need to decide what are we actually trying to understand from this report. Are we interested in understanding retention, revenue, engagement etc.
Which Metric - Depending on our end Goal, we need to select the most relevant metric, as we see there are lots of metrics available to us.
Which Segment - The Last step is applying the segment that is meaningful to your business. Maybe you want to understand the difference in Behavior between email and RSS readers or Paid vs Organic traffic etc.
To get started -
First Go to the Audiences > Cohort Analysis - to access the report in the Left Navigation.
As in my case, Recently I published an Article on Remarketing - The Ultimate guide to remarketing with Google analytics. And I have seen the highest number of traffic to this blog after publishing that article.
But the question is " So What? "
I get lots of visits but Is that traffic is relevant? Did I attract readers who stayed with me and become a loyal reader of this blog? what percentage of the readers are engaged with my website over a period of time?
Cohort analysis can help us to answer these kinds of questions.
Depending upon the nature of the business, you set the Cohort size, metric and Date range.
In the above pic, you can see that I have selected cohort size by week and user retention metric and last 6 week in the Date range. Here, I am interested to know what percentage of users come back to my website after they first interacted with it and how long they keep coming back again and again to read my articles. If you are a publisher and publish contents daily then you might want to see the data by day to understand user behaviors.
If you look at the table data, you can see that Dec 10 - Dec 16 cohort has the maximum number of users (859 users) as I published the article on Dec 10.
And in the heatmap of Dec 10 - Dec 16 and Dec 17 - Dec 23 cohort, you can see that only 3.73% and 4.72% of the readers respectively come back after 1 week of acquisition. Then the numbers again reduced after the 2 weeks. And the Nov 26- Dec 2 cohort has the highest retention rate relative to all the data in the table hence very darker in color.
And in the line graph above, it's clear that two of my high user cohort's retention faded away after week 3 but the Nov 26- Dec 2 stayed longer than those other two cohorts.
Overall if you look at the table data above, you can see that my website ain't able to retain users after the 4th week.
Now, I know from the above data that I have to keep writing at least 2-3 articles every month ( In Dec I wrote only one) to keep these users coming back again and again to read my articles. Now, I know beyond what point, I am going to lose all these readers that I acquired with so many efforts. Now, I know that I have to move my lazy ass more and write more articles every month to retain users for a longer period of time.
One great feature of new cohort analysis report in Google Analytics is that I can click on any particular cohort and create a cohort segment out of it and apply to my various reports in Google Analytics to understand more about these cohorts behaviors. Like What contents are they most interest in? Where these people are coming from? Which contents leads to more engagement and help me to generate more newsletter signups? How much economic value is delivered to my business from these cohorts? etc.
Until Now, we only look at the users at an aggregate level but to go one step further, we need to apply segments to this cohort report. Some of the traffic sources that send the highest amount of users to my website are GrowthHackers community, direct traffic, and social media. Segmenting the cohort report with the traffic sources helps me to understand, which traffic sources sending me more relevant users and which are not. Without segmentation, I can't able to understand the difference in their behaviors from the aggregate data.
How to create the Traffic sources segment.
1. First Click on +Add segment at the top > New Segment.
2. Then select the source for which you want to create the segment. In my case it's GrowthHackers. And choose Filter users at the top.
Why User-based filter?
Because the cohort analysis report is user-based, so if you apply a segment based on sessions, you can get unexpected results that do not include 100% of users on day 0 as you would expect. As you can see in the above pic.
Now, let's apply the two segments and try to understand the difference in user behaviors between these segments.
If you look at the above pics, you can see that Nov 26- Dec 2 cohort has a very high retention rate than any other cohorts in both the segments hence more important for me. Right?
Not so Fast!
Now, look at the raw numbers at the left for this cohort. Only 3 users. Though this cohort has a high retention rate but I can't make a decision based on 3 users only. So be careful when doing your analysis. Always look at the raw numbers because percentages and averages have a great power to mislead you.
Now let's take a look at the Dec 10 onward cohorts. if you look at the data table, it is obvious that readers from the GrowthHackers community are performing better than the readers who came from social media. So focusing more of my efforts on GrowthHackers community is more profitable than the social media.
Now, we can take our analysis to the next level by asking again " So what?"
ok, some % of users come back to my website again and again for some period of time but are they delivered any macro or micro value to my business? what was the outcome of all these efforts?
As it's a content website, so I can use the Goal completion or session duration metrics to understand the outcomes and engagement and if you are in an E-commerce business then more relevant metrics for you will be Revenue per user or revenue. But you can also use Goal completion if you want to. As I said before depending on your end goal ( What are you trying to understand) pick the metric that gives you the direct insights.
From the above pic, you can see that most of the Goals are completed in the week 0. 105 goals by the GrowthHackers community and 63 goals by the Social Media. Then the number of goals completion start to fall after the week 1 onwards.
Another thing to notice is that Dec 10-Dec 16 cohort for the Social media segment has the highest number of users as well as numbers of goals completed in the week 0 compared to GrowthHackers but the numbers started to fall very quickly in the number of users and goal completion by Social media ( Look at the Cohorts and Week 0 column). Compared to that the numbers of users and goals completions increased from the GrowthHackers community.
Overall, not only the user retention from the GrowthHackers community is better for this website but it is also more engaged with the website relative to the social media.
So, it's obvious that allocating more resources to GrowthHackers and then the Social Media is a right thing to do. Without Segmenting the report, we can't get these kinds of insights that help us to make the right decision for our business.
And if you have an E-commerce website, select the Revenue or Revenue per user metrics as we discussed earlier.
Here, you will also do the analysis the way we did earlier.
From the above pic you can see that after 2 days of acquisition, the Revenue per user falls down to nearly $0.0 for all the cohorts. Which is not a good sign at all.
Again, you can apply segmentation to compare the behaviors of users for the different segments.
Here, I just applied two simple segments to understand the organic and paid traffic, which is very important for any businesses.
In this pic, we are looking at Revenue metric for the organic and paid. If you look closely, you will see that most of the revenue is generated on the day the customers were acquired and after that, it's nearly come down to $0 for both the traffic sources. Although, the organic traffic is performing better than the paid traffic.
4. What is Customer Lifetime Value?
As customers are becoming less loyal to brands and more loyal to their overall experience. Focusing on Customer lifetime value enables you to find the customers that matter most and have a better relationship with them.
It helps you to compare the users acquired through different marketing channels ( organic, paid, email, display etc.) and find out which sources bring the high-value users for your company in the long term.
5. Understanding the settings and features of Lifetime Value report -
1. LTV Metric - The primary lifetime value metric you want to analyze.
Available metrics in the report -
1. Goal completion per user (LTV).
2. Pageview per user ( LTV).
3. Revenue per user (LTV).
4. Session duration per user (LTV)
5. Sessions per user (LTV)
6. Transactions per user (LTV)
7. Appviews per user (LTV) - if it's an app
you can also compare these metrics with one another with the help of compare metric option. Look at the pic above.
2. Acquisition Date Range -
The Date range during which users were acquired. Use the acquisition date range to set the cohort that you want to examine. As in my case, I can select the cohorts date during which I published the article. You will select date range depending on your business requirements like Christmas holiday or black Friday campaigns etc.
The graph shows the lifetime value per user for the metric over a period of 90 days( Maximum lifetime) in increments of day, week and month.
In the above pic -
Day 0 - is showing the cumulative revenue per user on the day of acquisition.
Day 70 - is showing the cumulative revenue per user on the 70th day after the acquisition.
The table shows the number of users you acquired during the Acquisition date range, along with the two additional metrics related to the LTV metric you have selected at the top.
In this case, it's - Revenue per user(LTV) and Revenue(LTV).
6. How to analyze the Lifetime value report in Google analytics -
For the sake of simplicity, let's set the acquisition date range to be Dec 1 - Dec 31, 2017.
in this example, we want to understand which channel is responsible for acquiring users with the highest Revenue(LTV) for the Google Merchandise store.
From the above pic, it is clear that the number of users acquired from organic search is highest among all but the referral traffic is responsible for the highest Revenue(LTV) $264,788, followed by $57,213.51 for the direct traffic compared to only $31,499 for the organic search traffic.
And if you look at the last row for the Affiliate traffic, you can see that the lifetime value delivered by it is totally $0.0
From this report, you know which channel is helping you to succeed in long-term and which isn't?
Now you know which channel you should shower your love and affection and which channels need to be divorced :)
want more fun?
Change the dimension from channel to campaigns.
just look at the red rectangle in the above pic. scratching head :)
Fire each and every one of them.
Before you do, take a look at the Assisted conversion report in the Multi-channel funnel section in the Google analytics for the real truth.
By default in Google Analytics, Conversions are attributed to the last campaign or source or medium etc. who converted the customers in the last stage. But we all know that it's not the real truth. Most customers come from various channels, do lots of comparison shopping, product research etc. before making a final purchase. So always look at the assisted conversion report in Google analytics to make better decisions.
Good husbands/wife don't make impulse decisions :)
Read this post to learn more about Assisted conversions in Google Analytics - How to quickly optimize your Google AdWords campaigns, keywords and ad groups?
Read it from top to bottom to get most out of it.
I Hope you find this article helpful.
Please share your perspective with all of us below in the comment. For more posts like this subscribe to our blog by clicking here or on the form below.