Be Real-World Smart: A Beginner's Advanced Google Analytics Guide

NectarBeing book smart is good. The outcome of book smart is rarely better for analytics practitioners then folks trying to learn how to fly an airplane from how-to books.

Hence, I have been obsessed with encouraging you to get actual data to learn from. This is all the way from Aug 2009: Web Analytics Career Advice: Play In The Real World! Or a subsequent post about how to build a successful career: Web Analytics Career Guide: From Zero To Hero In Five Steps. Or compressing my experience into custom reports and advanced segments I've shared.

The problem for many new or experienced analysts has been that they either don't have access to any dataset (newbies) or the data they have access to is finite or from an incomplete or incorrect implementation (experienced). For our Market Motive Analytics training course, we provide students with access to one ecommerce and one non-ecommerce site because they simply can't learn well enough from my magnificent videos. The problem of course is that not everyone is enrolling our course! :)

All this context is the reason that I am really, really excited the team at Google has decided to make a real-world dataset available to everyone on planet Earth (and to all intelligent life forms in the universe that would like to learn digital analytics).

The data belongs to the Google Merchandise Store, where incredibly people buy Google branded stuff for large sums of money (average order value: $115.67, eat your heart out Amazon!). And, happily, it has almost all of the Google Analytics features implemented correctly. This gives Earth's residents almost all the reports we would like to look at, and hence do almost all the analysis you might want to do in your quest to become an Analysis Ninja. (Deepak, would you kindly add Goal Values for the Goals. Merci!) You'll also be able to create your own custom reports, advanced segments, filters, share with the world everything you create, and all kinds of fun stuff.

For consultants and opinion makers you no longer have to accept any baloney peddled to you about what analytics tool is the best or better fit for your company/client. Just get access to this data and play with the actual GA account along with Adobe and IBM and WebTrends et. al. and suddenly your voices/words will have 10x more confidence informed by real-world usage. No NDA's to sign, no software to install, no IT resources required. Awesome, right?

In this post I'll highlight some of my favourite things you can do, and learn from, in the Store dataset. Along the way I'll share some of my favourite metrics and analytics best practices that should accelerate your path to becoming a true Analysis Ninja. I've broken the post into these sections:

I'm sure you are as excited as I am to just get going. Let's go!

How to get Store Dataset Access?

It is brilliantly easy.

Go to the Analytics Help Demo Account page. Read the bit in the gray box titled Important. Digest it.

Then click on this text: ––>ACCESS DEMO ACCOUNT<––

Looks scary in the all caps, right? That is just how the Google Analytics team rolls. :)

You'll see a tab open, urls will flip around, in two seconds you'll see something like this on your Accounts page…

google analytics accounts view

Click on 1 Master View and you are in business.

If you ever want to remove access to this real-world data, just go back to the page above and follow the five simple steps to self-remove access.

Jump-Start Your Learning.

You can start with all the standard reports, but perhaps the fastest way for you to start exploring the best features is to download some of the wonderful solutions in the Google Analytics Solutions Gallery.

You'll find my Occam's Razor Awesomeness bundle there as well.

It is a collection of advanced segments, custom reports, and dashboards. You'll have lots of features incorporated in them. You can customize them to suit your needs, or as you learn more, but you won't have to start with a blank slate.

You can also search for other stuff, like custom reports or attribution models.

Another tip. If you are a complete newbie (welcome to our world!), you probably want to start your journey by reading about each type of report, and then looking at the Overview report in each section in Google Analytics. At this point you'll be a little confused about some metric or the other. That's ok. Go, read one of the best pages in the Analytics help center: Understanding Dimensions and Metrics. Go back into GA, you'll understand a whole lot more.

This is a beginner's advanced guide, so I'm going to do something different. Through my favourite reports, often hard to find in your company's GA dataset, I'm going to push you beyond other beginner's guides. I'll also highlight frameworks, metrics, custom reports, and other elements I feel most Analyst's don't poke around enough.

1. Play with Enhanced Ecommerce Reports.

It is a source of great sadness for me that every single site is not taking advantage of Enhanced Ecommerce tracking and analysis . It is a complete rethink of ecommerce analysis. The kind of reports and metrics you'll get straight out of the box are really amazing.

Go to the Reporting section of our Store Demo account, click on Conversions in the left nav, then Ecommerce, and now Overview. You'll see in an instant the very cool things you can track and analyze…

ecommerce overview

With a little bit of smart tagging you can track your internal promotions (buy one Make America Great Again hat and get one Stronger Together hat free!), transactions with coupon codes, affiliate sales and more. Very nicely summarized above.

Next go to the report with new things that will help you drive smarter merchandizing on your mobile and desktop websites. Go to Shopping Analysis and click on Shopping Behavior…

shopping behavior analysis google analytics

I adore this report.

Most of the time when we do funnel analysis we start at the Cart stage (third bar above). We rarely hold people responsible for Traffic Acquisition accountable, we rarely hold people responsible for Site Design and Merchandizing accountable. The former are promoted on silly metrics like Visits or Visitors or (worse) Clicks. The latter are promoted based on silly metrics like PageViews.

The first bar to the second shows the number of visits during which people went from general pages on your website to product pages (places were there is stuff to be sold, add to cart buttons). A lame 26%. See what I mean. Insightful. How are you going to make money if 74% of the visits don't even see a product page!

The second bar the third is even more heart-breaking, as if that were possible. Of the sessions with pages with product views, how many added something to cart. A lousy 17%. One. Seven. Percent! On a site were you can do nothing except buy things.

See what I mean? Question time for your Acquisition, Design and Merchandizing team.

Do you know answers like these for your website? That is why you need Enhanced Ecommerce.

I won't cover the last two bars, most of you are likely over indexing on funnel analysis.

Practice segmentation while you are here. Click on + Add Segment on top of this report, choose Google (or whatever interests you)…

google traffic segment

And you can analyze acquisition performance with a unique lens (remember you can't segment the funnel that exists in the old ecommerce reports which is still in your GA account!)…

shopping behavior analysis google traffic

A little better. Still. You spend money on SEO and PPC. It should be a lot better than this. If this were your data, start with questioning your PPC landing page strategy and then move to looking at your top SEO landing pages, and then look at bounce rates and next page analysis for those that stay.

I can honestly spend hours on just this report digging using segmentation (geo, media, new and loyal customers, all kinds of traffic, product page types and so on). It has been a great way to immediately influence revenue for my ecommerce engagements.

While you are here, you can play and learn to use the new funnel report… it is called Checkout Behavior Analysis…

checkout behavior analysis google analytics

Much simpler, so much easier to understand.

You can also, FINALLY, segment this report as well. Try it when you are in the Store demo account.

Take a break. A couple days later come back and checkout the new Product Performance and Product List Performance reports. The latter is particularly useful as an aggregated view for senior executives. In case of the Store data, the first report has 500 rows of data, the second just 45. Nice.

I wanted to flag three metrics to look at in the Product Performance report.

Product Refund Amount is $0.00 in this dataset, but for your company this is a great way to track refunds you might have issued and track were more of that is happening.

I love Cart-To-Detail Rate (product adds divided by views of product details) and Buy-to-Detail Rate (unique purchases divided by views of product-detail pages). Remember I was so upset above about the poor merchandizing. Using the sorting option on these two columns I identify where the problem is worse and where I can learn lessons from. Very cool, try it.

I could keep going on about more lovely things you'll find in the Enhanced Ecommerce reports, but let me stop here and have you bump into those cool things as, and I can say this now, you have access to this data as well!

Bonus: If you are a newbie, in your interview you'll be expected to know a lot about Goals (I call the micro-outcomes). Explore that section. Look the Overview, Goal URLs and Smart Goals. Ignore the eminently useless Reverse Goal Path report (I don't even know why this is still in GA after years of uselessness) and Funnel Visualization (almost totally useless in context of almost all Goals).

2. Gain Attribution Modeling Savvy.

My profound disdain for last-click reporting/analysis is well known. If you are using last-click anything, you want your company to make bad decisions. See. Strong feelings.

Yet, many don't have access to a well set-up account to build attribution modeling savvy and take their company's analytics the year 2013. Now, you can!

I am big believer in evolution (hence my marketing and analytics ladders of awesomeness). Hence, start by looking at the Assisted Conversions report (Conversions > Multi-Channel Funnels)…

assisted conversions google analytics

Then metric you want to get your company used to first, to get them ready for savvier attribution anything, is the metric Assisted Conversions. The last column.

Here's the official definition: A value close to 0 indicates that this channel functioned primarily as the final conversion interaction. A value close to 1 indicates that this channel functioned equally in an assist role and as the final conversion interaction. The more this value exceeds 1, the more this channel functioned in an assist role .

Now scroll just a bit back up, stare at that column, what would your strategy be for Organic Search if it is at 0.46? What about Display advertising driving which plays primarily an "upper funnel" introducing your brand to prospects 1.58?

The change required based on this data is not just your marketing portfolio re-allocation, that is almost trivial, what' bigger, huger, crazy-harder is changing how your company thinks. It is painful. Largely because it quickly becomes about how people's budgets/egos/bonuses. But, hundreds of conversions are on the line as well on insights you'll get from this data. Learn how to use this metric to drive those two changes: marketing portfolio – people thinking.

Couple bonus learnings on this report.

On top of the table you'll see text called Primary Dimension. In that row click on Source/Medium. This is such a simple step, yet brings you next layer of actionable insights so quickly. You'll see some surprises there.

Second, look at the top of the report, you'll see a graph. On to top right of the graph you'll see three buttons, click on the one called Days before Conversion…

assisted conversions days before conversion

I love this report because it helps me understand the distribution of purchase behavior much better. I profoundly dislike averages, they hide insights. This report is the only place you can see distribution of days to purchase for Assisted Conversions.

If you've changed the think in your company with Assisted Conversions… You are ready for the thing that gets a lot of press… Attribution Modeling!

You'll find the report here: Conversions > Attribution > Model Comparison.

You'll see text called Select Model next to Last Interaction. Click on the drop down, ignore all the other models, they are all value deficient, click on the only one with decent-enough value, Time Decay, this is what you'll see…

attribution modeling last click vs time decay

Half of you reading this post are wondering why I don't like your bff First-Interaction (it is likely the worst one on the list btw) or your bff Linear (the laziest one on the list)… worry not, checkout this post: Multi-Channel Attribution Modeling: The Good, Bad and Ugly Models .

The column you are of course looking at is % Change in Conversions. The GA team is also helping you out by helping you understand where the results are significant, green and red arrows, and where it is directional, up or down gray arrows.

This is the data you'll use to drive discussions about a change in your marketing $$$ allocations.

Where you have CPA, it is is an even more valuable signal. And, such a blessing that the Store demo account has that data for you.

You'll need all your brain power to understand the report above (make sure you read the models post above), and then some more to drive the change in how your company thinks. Attribution model is not a software or math problem, it is an entrenched human minds problem.

And because I'm the author of the quote all data in aggregate is crap I recommend scrolling up a bit in the attribution modeling report and clicking on the down arrow under the word Conversion….

attribution modeling goals analysis

This is admittedly an advanced thing to learn because even understanding marketing dollars plus user behavior overall is hard, this just makes it a bit more complicated because you can actually understand those two things for every goal you have individually or just ecommerce all by itself.

It is incredibly awesome to be able to do that because now you are this super-data-intelligent-genius that can move every variable in a complex regression equation very finely to have max impact on your company.

If you can master this, and IF you can evolve how your company does marketing portfolio allocation and how it thinks, then you are ready for the max you can do in Google Analytics when it comes to attribution… custom attribution modeling.

On top of the table, click on Select Model, then Create New Custom Model.

To get you going, here's one of my models for a client…

custom attribution model

Custom attribution models are called custom because they are custom to every company. It requires an understanding for everything I've requested you to do above, business priorities (what the business values), and business strategy.

Creating a couple different custom attribution models, seeing how it affects the data, what decisions GA recommends, helps you have an intelligent argument with all your stake holders. Again, the decisions from this analysis will flow into changes to your marketing portfolio and how people in your company think.

Once you get into custom attribution modeling, and you spend serious amount of money on marketing online (a few million dollars at least), you are ready for the thing that actually will drive the best changes: Controlled Experiments (aka media mix modeling). Hence, it is critical that you approach your learnings in the precise steps above, don't jump steps if at one of them you have not changed how your company thinks.

Bonus 1: You might think the above is plenty advanced. It is not. For the higher order bits, when you are all grown up, read this post and internalize the implications of it: Multi-Channel Attribution: Definitions, Models and a Reality Check

Bonus 2: The Time Lag and Path Length reports in your Multi-Channel Funnels folder are extremely worth learning about. I like Path Length more, more insightful. When you analyze the data, be sure to play with the options under Conversion, Type (click AdWords), Interaction Type and Lookback Window. With each step absorb the patterns that'll emerge in the data. Priceless.

3. Learn Event Tracking's Immense Value

I'm very fond of Event Tracking for one simple reason. You have to create it from scratch. When you open GA, there is no data in these reports. It can only get there if you spend time trying to understand what's important to the business (Digital Marketing and Measurement Model FTW!), what is really worth tracking, and then through intelligent thought implementing the tracking.

I love the fact that you have to literally create data from scratch. For any beginner who is trying to get to advanced, Event Tracking will teach you a lot not just about Event Tracking but creating smart data.

Lucky for us the GA team has created some data for us to play with. Go to Behavior in the left nav, then Events, and then Top Events… This is what you'll see…

event tracking top events

The Store team is capturing four events, you can drill down into any one of them to get a deeper peek into user behavior.

I choose Contact Us to analyze the Event Labels, I get all these strategies that people…

event tracking event lables

It would be valuable if the Event Value had been populated, which would also give us Avg. Value in the table above. Still. Understand that data, how it is collected, what it implies about user behavior is incredibly valuable.

You can also create an advanced segment for any of the events above, example Email. Then, you can apply that segment to any of other reports in Google Analytics and really get deep insights. What cities originate people who call is on the phone? What sites did they come from? How many visits have they made to the site before calling? So on and so forth.

The event tracking reports have three options on top of the report. Event, Site Usage, Ecommerce.

Try the Ecommerce tab…

event tracking ecommerce drilldown

While we did not see any event values, you can tie the sessions where the events were fired with outcomes on the site. Really useful in so many cases where you invest in special content, rich media, interactive elements, outbound links, merchandizing strategies etc. This report, in those cases, will have data you need to make smarter decisions faster.

Bonus: While you are in the Behavior section of Analytics, familiarize yourself with the Site Speed report. Start with the scorecard in the overview report. Move on to Page Timings to find the pages that might be having issues. One cool and helpful visual is Map Overlap, click the link on top of the graph on the Page Timings report. Close with the Speed Suggestions report. Your IT team needs this data for getting things fixed. Your SEO team can do the begging, if required. :)

4. Obsess, Absolutely Obsess, About Content

It is a source of intense distress for me that there's an extraordinary obsession about traffic acquisition (PPC! Affiliates! Cheat Sheet for Video Ads!), and there is huge obsession with outcomes (Conversion Rate! Revenue!), there is such little attention paid to the thing that sits in the middle of those two things: Content!!

Very few people deeply look at content. Yes, there will be a top pages report or top landing pages report. But, that is barely scratching the surface.

Look. If you suck at content, the greatest acquisition strategy will deliver no outcomes.

Obsess about content dimensions and content metrics.

Since you know some of the normal reports already, let me share with you a report that works on many sites (sadly not all), that not many of you are using.

The Content Drilldown report uses the natural folder structure you are using on your website (if you are) and then aggregates content on those folders to show you performance. Here is what you'll see in the Store demo account you are using…

content drilldown report google analytics

Nice, right? You are pretty much seeing all of the content consumption behavior in the top ten rows!

A pause though. This report is sub-optimally constructed. It shows Pageviews (good), Unique Pageviews (great) and then three metrics that don't quite work as well: Average Time on Page, Bounce Rate, % Exit (worst metric in GA btw if anyone asks in an interview)…

content drilldown report google analytics 2

At a folder level these really help provide any decent insights, and might not even make any sense. Think about it. Bounce Rate for a folder?

Good time for you to learn simple custom reporting.

On top of the report, right under the report title, you'll see a button called Customize. Press it. Choose more optimal metrics, and in a few seconds you'll have a report that you like.

This is the one I created for my use with valu-added content metrics that work better: Average Session Duration, Cart-to-Detail Rate (as it is an ecommerce site) and Page Value (to capture both ecom and goal values at a page level)…

content drilldown custom report google analytics

Much better, right? Would you choose a different metric? Please share it via comments below.

Ok. Unpause.

Even a quick eyeballing of the report above already raises great questions related to overall content consumption (Unique Pageviews), merchandizing (Cart-to-Detail Rate) and of course money.

You can now easily drill-down to other more valuable bits of content and user experience.

I click on the first one, most content consumption, to reveal the next level of detail. I can see that Apparel is the biggest cluster of content, with pretty decent Cart-to-Detail Rate…

content drilldown 2 custom report google analytics

Depending on the business priorities I can ask questions like how come the summer olympic games stuff no one seems to want (and we spent $140 mil on an Olympics sponsorship, kidding).

At the moment the company has a huge investment in Google Maps branding, so we can look at how various brands are doing… YouTube FTW!

content drilldown 3 custom report google analytics

Maps is not doing so well. You can see how this data might make you curious if this list is what your business strategy is expecting will happen? Or, is this how we prioritize content creation? I mean, Go! People are interested in something esoteric like Go (programming language in case you are curious) rather than Nest! What a surprise.

That is what this type of content analysis is so good at.

You can continue to follow the rabbit hole by the way and get down to the individual pages in any folder, like so…

content drilldown 4 custom report google analytics

Ten percent Cart-to-Detail Rate is pretty poor, compared to some of the others above. Time to rethink if we should even be selling this combo! If not that, definitely time to look at the page and rethink copy, images, design, and other elements to improve this key metric.

The above custom report is really easy to create, for Subscribers of my newsletter I'll also email a downloadable link for this and other custom reports below.

Bonus: Most people stop at what the reports show in the default view. The GA team does a great job of adding good think and express it all over the standard reports. For example, in context of our discussion here, try the Content Grouping primary dimension. Here you see what happens to the report when I switch to Brands (Content Group)…

all pages report content groupings

Even more useful than what was there before, right? So, how does GA get this data? As in the case of Event Tracking above, the Analyst and business decision making combination are thoughtfully manufacturing data. In this case using the immensely valuable Content Groupings feature. Invest in learning how to use it in the Store demo account, learn how to create content groupings to manufacture useful data. When you interview for higher level Analytics role, or for a first time Analyst role, you'll stand out in the interview because this is hard and requires a lot of business savvy (ironic right, you stand out because of your business savvy in a Data Analyst interview!).

5. SEO & PPC, Because You Should!

Ok, you've waited long enough, time to talk about the thing you likely spend a ton of time on: Acquisition.

Since you likely already know how to report Traffic Source and how to find the Referring URLs and Sessions and… all the normal stuff. Let me focus on two things that are a bit more advanced, and will encourage you to learn things most people likely ignoring.

The first one I want you to immerse yourself in when you are in the Store data is Search Engine Optimization. You know that this is hard because when you go to Acquisition > Campaigns (what!) > Organic Keywords you will see that 95% are labeled "(not provided)". This report is completely useless.

You do have other options to analyze SEO performance. Here's the advanced, advanced, lesson: Search: Not Provided: What Remains, Keyword Data Options, the Future.

But, you also have some ability in Google Analytics itself to do keyword level analysis for Google's organic search traffic. Go to Acquisition > Search Console > Queries. This report shows you the top thousands of keywords (4,974 precisely today in the Store report today). The data is available because the team has configured the Search Console data to connect with GA.

Here's what you'll see…

organic search queries report google analytics

I sort the data by Clicks, because Impressions is a lot less valuable, and with Clicks I get something closer to Sessions (though they are very different metrics). I immediately value CTR as a metric in this context, you can see the variations above. This is perfect immediate data for SEO discussions.

Average Position is also interesting, perhaps more so for my peers in the SEO team. As a Business Analyst I value Average Position a lot less in a world of hyper-personalized search.

My next data analysis step is to take this data out of GA (click Export on top of the report) and play with it to find macro patterns in the data. I'll start with something simple as creating tag clouds, using Clicks or CTR as contextual metrics. I'll classify each keyword by intent or other clusters to look for insights.

Try these strategies, can you find weaknesses in the Google Store's SEO strategy? How do your insights compare to what you just discovered in the content analysis in terms of what site visitors actually want? Really valuable stuff.

What you cannot do with this data is tie it to the rest of the data in GA for these visitors. You cannot get conversions for example, or Page Depth etc. This is heart-breaking. But, see the not provided post I've linked to above for more strategies and meanwhile you can do some cool things in Google Analytics when it comes to SEO.

Bonus: In the Search Console reports, I also find the Landing Pages report is also helpful because you can flip the center of universe, for the same metrics as above, to landing pages rather than keywords. The insights you get will be helpful for your SEO team but more than that it will be critical for your site content team.

A quick note on the above… for the current data you'll see the Landing Pages report looks a little weird with no data in the Behavior and Conversion columns. Something weird is going on, on my other accounts there is data. The team can fix this in the very near future.

Next, spend a lot of time in the AdWords section.

Both because Paid Search if often a very important part of any company's acquisition strategy, and because at the moment there are few digital acquisition channels as sophisticated and complex as AdWords. When you are getting ready for your interviews, being good at this, really good, is a great way to blow your interviewer away because most people will know only superficial stuff about AdWords.

As if those reasons were not enough, in Google Analytics AdWords is a great place to get used to the complexity that naturally arises from mixing two data sources. In almost all GA AdWords reports the first cluster of data (pink below) will come from AdWords and the second cluster (green brace) is the normal collection of metrics you see in GA…

adwords plus google analyitcs

This will naturally prod you into trying to understand why are Clicks different from Sessions? After-all it is a click that kicks off a session in GA when the person arrives. It is internalizing these subtle nuances that separate a Reporting Squirrel from an Analysis Ninja.

Above view is from the Campaigns report. I usually start there as it gives me great insights into the overall PPC strategy for the company.

While you are learning from this report, here's a little smart tip… Click on the Clicks link on top of the graph you see (you'll see it along with Summary, Site Usage, Goal Set 1, and Ecommerce), you'll get a different set of metrics you should know intimately as well…

camapign clicks deeper outcomes

The combination of CPC and RPC is very important. It is nice that they are right next to each other in this view.

When you look at Store data I also want you to live-see why ROAS not even remotely a useful metric. It looks alluring. Return On Ad Spend. That sounds so awesome, surely it is in some holy books! No. It is not.

For now, invest in understanding what is is measuring, what the data shows, is that good or bad, and what's missing. When you already to move to advanced-advanced stage, read this post: Excellent Analytics Tip #24: Obsess About Real Business Profitability

Once I've exhausted the value in Campaign reports (drilling-up, drilling-down, drilling-around), it is time to shift into detail. While it might seem that the very next step will be the AdWords Keyword report, it is not. I like going to the Search Query report first.

In AdWords context, Keyword is what you buy from Google. Search Query on the other hand is what people are actually typing into Google when your search ad shows up (triggered by the Keyword of course).

Here are the two reports from the Store account, you can clearly see why I like starting with the Search Query report….

search keywords vs search queires reports

I would much rather learn to anchor on what people are typing and then go into the Keyword view to see what I can learn there. The Search Query performance report helps me re-think my AdGroups, Match Types, bidding strategies and more. It also helps me optimize the landing pages, both from a content they contain and what ads I recommend send traffic there.

You could spend three months in these reports just learning and finessing your PPS savvy, so I'll leave you to that. :)

Bonus: Shopping Campaigns are incredibly successful for most ecommerce properties. Spend time in that report in the AdWords section, drilling-down and segmenting, to learn what makes these campaigns distinct and if you were tasked to identify insights how would you go about it.

6. Develop a Smarter Understanding of Your Audiences

Having grown up on cookies, we have typically have had a finite understanding of our audiences. This has slowly changed over time, most recently with the awesomeness of User-ID override empowering us to understand a person. Still, most of the time we are not great at digging into Audiences, and their associated behavior.

Hence, to assist with your evolution from beginner to advanced, three often hidden areas of Google Analytics for you to explore now that you have access to real data.

Go to Audience > Interests > In-Market Segments.

Here's the official definition of what you are looking at: Users in these segments are more likely to be ready to purchase products or services in the specified category. These are users lower in the purchase funnel, near the end of the process.

I've developed an appreciation of this report as I think of my performance marketing strategies, especially the ones tied to Display advertising. Far too often we rely on just PPC or email and don't use Display in all of the clever ways possible. This repor, leveraging insights from my users, help me understand how to do smarter Display.

in-market segements google analytics

You can drill down to Age by clicking on the in-market segment you are interested in, and from there for each Age group you can drill-down to gender.

Per normal your goal is to identify the most valuable ones using micro and macro-outcomes for your business.

After I've mastered in-market segments by adding near term revenue to my company and helping shift the thinking about Display in my company, I move to leverage the data in the Affinity Categories. Also a report in this section. Affinity categories are great for any display or video advertising strategies you have to build audiences around See Intent (See – Think – Do – Care Business Framework). A bit more advanced from a marketing perspective (you would have had to master strategy #2, attribution, above).

For the second hidden area, go to Audience > User Explorer.

This lovely beast shows something you think you are dying to see. It is also something I really don't want you to obsess about (except if you are a tech support representative). But you want it. So. Here it is…

user explorer report google analytics

What you are looking at is a report that shows you the behavior of an individual user on your website, as identified by an anonymous Client-ID. You can loosely think of it as a person, though it is more complicated that. If you have implemented User-ID override (congratulations, you deserve a gold star!), then you areas close to a person as you'll ever be.

Because this is everyone on your website, there is no wrong place to start and a hundred thousand terrible places to waste time. You can literally watch each person! See, what I mean when I say I don't want you to get obsessed about this?

On the rarest of rare occasions I look at this report, my strategy is to understand the behavior of "Whales", people who spend loads of money on our website (why!). I sort the above report by Revenue, and then look over the users who form the first few rows. The data, fi you do it in the Store account for the person who's at the top at the moment, looks like this…

user explorer report google analytics detail

The report is sorted from the last hit (08:16 above) to the first hit (which you don't see above, the person browsed a lot!). You can quite literally watch the behavior, over just five minutes, that lead to an order of $2,211.38! You surely want to know what this person purchased (Men's Cotton Shifts FTW!), what pages did they see, where did they come from, how did they go back and forth (this person did) and so on and so forth.

Looking at the top few of these Whales might help know something about a product merchandizing strategy, a unique source, or how to change your influence with your acquisition strategy to get a few more of these people. There will always only be handful of folks.

The higher order bit is that the best analytical strategy is to analyze micro-segments rather than individuals. Small groups with shared attributes. You can action these, at scale. Nothing in your marketing, site content delivery, servicing at the moment has the capacity to react to an individual's behavior in real time. And if you can, you don't have enough visitors. Hence, obsess about micro-segments. That is a profitable strategy.

The spirit above is also the reason why I don't mention real-time reporting in this guide. Simply not worth it. (For more, see #4: A Big Data Imperative: Driving Big Action)

For the third hidden area, ok, not so hidden but to expose all your analytical talent, go to Audience > Mobile > Devices.

With greater than 50% of your site traffic coming from mobile platforms, this audience report obviously deserves a lot of attention (in addition to segmenting every single report for Mobile, Desktop, Tablet).

The problem is that the report actually looks like this…

mobile analytics

It is poorly constructed with repetitive metrics, and an under-appreciation for mobile user behavior (why the emphasis on Do outcomes when Mobile has primarily a See-Think intent clusters?). It makes for poor decision making.

So. Time to practice your custom reporting skills. (Oh, if you as an Analyst only use custom reports, you are closer to being an Analysis Ninja.)

Scroll back to the top of the Mobile Devices report and click on the Customize button. On the subsequent page, pick the metrics you best feel will give you insights into Acquisition, Behavior and Outcomes. While you are at it, you'll see just one dimension in this report, Mobile Device Info, you can add other drill-down dimensions you might find to be of value. I added Screen Resolution (matters so much) and then Page (to analyze each Page's performance by resolution).

Here's what the report's Summary view looks like for me…

smart mobile analytics

Nice, right? Smarter, tighter, more powerful.

My obsession is with people on mobile devices and not just the visits. Hence Users come first. Then, paying homage to See and Think intent, my focus is on Pages/Session. For the same reason, my choice for success is goals and Per Session Value (ideally I would use Per Session Goal Value, but as you saw in the opening this account does not have Goal Values). I would delete the Revenue, it is there mostly in case your boss harassed you. Delete it later.

Depending on the role, Acquisition, Behavior or Outcomes, I have everything I need to start my mobile analysis journey.

As I recommended with AdWords analysis above, the tabs on top of the report hold more analytical insights for you…

smart mobile analytics site usage

You will discover that you'll have to go and practice your custom reporting skills on all these tabs as there are sub-optimal elements on all three of them. For example with Site Usage, I added Think intent metrics. For Goals and Ecommerce tabs there are fewer and more focused metrics. Now almost all of the stuff I need to make smarter decisions from my mobile data is in one place.

This exercise requires a lot of introspection and understanding business needs as well as what analysis makes sense. That is how we all move from Reporting Squirrels to Analysis Ninjas! :)

As with the above custom report, I'll email a downloadable link to the Subscribers of my newsletter The Marketing – analytics Intersect. You can contrast your choices with my choice of metrics and dimensions.

Bonus: If you present screenshots from GA to your management team, make sure you take advantage of the option to show two BFF trends. In my case above you can see I choose to pair mobile Sessions with Goal Completions (again to put the stress on See – Think intent).

7. Icing on the Cake: Benchmarking!

One final beginner's advanced recommendation.

You just finished looking at a whole bunch of mobile metrics. How do you know if the performance of the Google Merchandizing Store is good or bad? Yes, you do see trends of past performance. But, how about with others in your industry? Others who have your type and size of website?

I've convinced that most of the time without that competitive / ecosystem context, Analysis Ninjas are making incomplete decisions.

The cool thing is, you can get benchmarking data in Google Analytics.

Audience > Benchmarking > Devices.

And now you have a really strong sense for what is good performance and what is non-good performance…

benchmarketing report device category

You might have come to one set of conclusions doing the analysis in the mobile section above, and I suspect that now you have very different priorities with the lens pulled back to how the ecosystem is doing.

And, that's the beauty.

There's a lot more you can do with benchmarking. You can explore the advanced-advanced version here when you are ready: Benchmarking Performance: Your Options, Dos, Don'ts and To-Die-Fors!

I hope you have fun.

That is it. A beginner's advanced guide that hopefully accelerates your journey to become an Analysis Ninja.

As always, it is your turn now.

Have already gotten access to the Store demo account? What elements recommended above had you not explored yet? Which ones do you find most easy/frustrating to get actionable insights from? Are there strategies that you use as an Analysis Ninja that are not covered above?

Please share your recommendations, frustrations, :), joyous strategies and guidance with all of us via comments below.

Thank you.


  1. 1

    What do you think about used custom dimensions? For example User Category, that distinguish retail users from Google Employees (after authorization on website)

    Imagine the following product recommendation widget "what Googlers buy to themselves". It's kinda different from all users.

    • 2

      Krill: An excellent idea!

      It is not implemented on the site yet, but I hope the GA team is reading it and will follow your advice because both the custom dimension and custom widget are wonderful. :)


  2. 3

    Hi Avinash,

    Great article. The section on event tracking prompted me to figure out and setup event data on Shopify's shipping calculator often on the cart page. By seeing the addresses people select and the delivery options available, can then segment this data with enhanced ecommerce analytics for invaluable things like cart abandonments or average order value. Will be able to learn how delivery options impact the checkout process, review shipping configurations, etc.

    In your example, why would you use content grouping when the data is available in the content drilldown report? I'm guessing because it gives you quicker access in using the branding dimension for all standard reports?

    The bug you wrote about with the Landing Pages report happens for all my client accounts.

    • 4

      Joshua: That is a very clever use of Event Tracking, thank you for sharing the detail and inspiring other readers of the blog as well!

      Content Grouping is hand crafted in this case (jump to the Admin section of GA to create your own Content Groupings). In the case of the Store account, Brands, Product Categories and Clothing by Gender are all content groups created by the Analyst.

      For some websites Brands or Products might be how the hierarchy of the site is set up. In these cases you are right that it might be in the content drilldown report.


  3. 5

    Hi Avinash – very cool…will this also give access to Google Analytics Premium or 360?

    Been wanting to check those out.

    • 6

      Jim: With GA Premium/Analytics 360,, you get things like unsampled reports, more custom dimensions, scheduled reports, some savvy things in the Admin interface, technical support directly from Google etc, but for the most part the core collection of things and analytical horsepower, for now, is pretty much the same as you'll get through using the account in this post.

      Other parts of 360, Tag Manager, Optimize, Attribution, Data Studio, all have free versions where there is some uniqueness in the paid versions but usually you have a limited version for free. For example you have access to full horsepower of Data Studio for free, you can just use it for five dashboards.

      So, you should be all set. :)


  4. 8
    Sushant Singhal says

    Marvelous post, as always !!!

    Google and Googlers are really changing the world. They are giving so many services for FREE !! Now, the real-data is free. I can't believe it !!!

    Hats off :)

  5. 9

    Thanks for this, I never knew about this public data set.

    This account will be very useful for me in another way. I regularly give seminars on Google Analytics, and it's always a challenge to show demos. I don't want to use my sites, and I'm certainly not going to use a client's site!

    Now I'll have an account that I can use for demos, saving me from creating fake static images for demonstrations.

  6. 10
    Italo Quispe says

    Thanks for this typically detailed article Avinash. It will serve as a template for learning for many people looking to improve their analytical skills.

    I have signed up for the newsletter and look forward to getting the custom reports you'd created for this post.

  7. 11

    You managed to do it yet again.

    Another absolutely must read article from the GA Master. As a beginner, I've found your articles to be enlightening but best of all, practical.

    Please continue doing what you're doing. It's make a big difference. Thank you.

  8. 12

    Hey Avinash !

    Firstly, you are FAB and so are all your posts !


    Please excuse me in pointing this out ( considering your vast experience and the amount of time,thoughts, patience and energy you invest in each and every article of yours) that why are you not directing your subscribers to your website/blog instead to read the complete article ?

    If you are sharing the entire content with us over mail, people like me hardly find any reason to visit your website as we are getting all that we require over mail itself ! In doing so are you not losing out on visitors to your website thus affecting your organic ranking ?

    Again please excuse me if you had thought this over and have some other strategy in mind, if not my suggestion is that, just give us a teaser to your amazing work and allow us or direct us to your website to explore more. Its a win-win situation.

    Anyways thanks for this great article and as always looking forward to more posts from you.

    Cheers !

    • 13

      Anoop: You are making a very good point, but my objective is to simply teach. And, as you can see there is nothing here connected to me asking you for money or showing you ads or other such elements.

      Hence, I don't care where you read it. I'm just happy that you do and that it has a material impact on the way you think.

      One day I'll be tired of being poor and then I will implement aggressive strategies that lead to making money! :)


      • 14
        Just Some guy says

        "One day I'll be tired of being poor and then I will implement aggressive strategies that lead to making money! :)"

        This certainly is something we all need to think about.

  9. 15
    Aurelien says

    Hi Avinash!

    Many thanks, it is very useful. About the content analysis, I am wonder if there is any chance to know the pageviews per user inside each folder / or content group.

    From my understanding of google analytics, the tool give us the performance from a website viewpoint. But to better understand the visitors behavior, it would be interesting to know with accuracy how they consume content.

    For example, if a folder is considered as a topic, knowing how much articles (pages) are viewed per visitor inside this topic would be very valuable. But I do not know if it is conceivable…

    What do you think about that ?


  10. 17

    Your posts are easy to consume and invaluable, Web Analytics 2.0 is still an absolute must read.

    I have been fortunate to have had access to massive data sets, courtesy of my professional career – now you and Google have unlocked this final key chapter.

    Fantastic work – long live Avinash!

  11. 18

    Hey Avinash,

    This is like a treasure trove.

    Especially someone like me who is a "hobbyist" analyst for now, any data set is welcome. This is the jackpot though!

    I am going to read, re read and share this post several times.

    May your tribe increase!

  12. 19
    Shweta Sharma says


    Very nice article and the way of explaining is best. But I want to know how to filter our own traffic as it counts my own visits also. Can anyone help me on this issue?

  13. 22

    Hi Avinash,

    Why use Avg. Session Duration with Page Path? What does that dimension/metric combination mean?


    • 23

      Mirissa: This is where you bump into the limitations of what you can do with GA.

      The Average Session Duration metric works pretty well at higher levels, but when you get to Page Path 4, it makes little sense. But you are locked into it with your custom report. That is ok, there are other metrics at level 4 that still provide value.


  14. 24

    Hi Avinash!

    Thank you for a nice post!

    I have setup content grouping for a client in order to categorize content into meaningful groups. Everything went smooth and the implementation with Datalayer + GTM worked like a charm. So far so good :)

    However I did this implementation so I could tell my client what content was of high value based on goal completions of particular goals. So i was hoping that I could attach goal completions to contents groups in a custom report but so far I have not been able to do this. The goals do simply display 0 when they are combined with content groups. Do you know if content grouping does not work with goals? The particular goals that I am using are setup in GA through events.

    • 25

      Klaus: The best way to look more deeply into this and solve any issues might be to work with a GACP to get their expert guidance on your account. Here's a list:

      Goals should work though with content groupings. A public post with some screenshots of this is this one by Paul:


      • 26

        Should you combine goals and content groupings? (session and behaviour metrics)

        I think you should combine content groupings with page value!

        • 27

          Looks like the Online Metrics blog uses content groupings in the landing pages report (dimension is 'landing content group') to show goal completions.

          Which makes sense because then you're using a session metric (goal completions) with a session dimension (landing content group). You can use this dimension in custom reports also.

          Otherwise I think Gerard is right, when using content groups rather than landing content groups, page value makes more sense than goal completions

  15. 28
    maharcin says

    Hi Avinash, in the Shopping Behavior Analysis screenshot, sessions with checkout (16.014) are higher than session with add to cart (14.744). How is this possible, am I missing something? Could it be because there are sessions which starts directly from the checkout, such as users coming from newsletter, cart-abandonment mails, etc.?

    • 29

      Maharcin: You are right, it shows that.

      It is a result, as you guessed from people directly dropping into the checkout. The use cases include what you mention, and additionally on sites like this were employees can also directly drop into the checkout to order stuff plays a role.

      You won't normally see it on other sites where the behavior is "normal". :)


  16. 30
    Ray Vinciente says

    Nice article avinash. It has become my new go to guide for future training sessions that our agency will be doing for our clients.

    Thanks for sharing such a useful article.

  17. 31

    Very interesting article.

    It just incredible how versatile google analytics has become and in such a small span of time.

  18. 32
    Robert Souza says

    Thank you for creating something for both new and advanced users Avinash!

    I fit somewhere in between. I can see the things I did right and those that I skipped over when I was a new analyst. And now I don't have to guess the next set of things I should focus on to continue my path to becoming an advanced user.

    I can only imagine the amount of effort that goes into writing a post like this, surely just the images take a very long time to create. Please know that you are having a huge impact on the knowledge and careers of many, many people around the world.

  19. 33
    Vasil Bekyarov says

    Great article ! Thanks.

    I like how you provide options that anyone can implement. I’m definitely interested in adopting this practice myself, thanks for working to improve your own footprint and others’ as well!

    Passing this on to friends for sure.

  20. 34
    Arjun Menon says

    Great read.

    Thanks a lot for the robust collection of insights Avinash.

  21. 35
    Sambhaji says

    Thanks for sharing useful article, it is very useful article to guide me in my upcoming future session

  22. 36
    Badra Malik says

    Google analytics gives you a vast amount of information. Knowing what to do with it (i.e. creating actionable insights) is key: For example understanding which market segments to concentrate your marketing on.

    Anybody looking for some further reading/tutorials on using Google Analytics might want to check out Google's eCommerce Analytics course.

  23. 37

    As always great insights Avinash. The real world data-set is really useful for people who are interested in the fascinating world of digital analytics though might not be professionally directly involved in it.Count me as one of those.

    I should have asked this question before but any how…..Till date I have not been able to understand why a site i put as one of my properties ( has viewer data coming to my account even though I did nothing except for putting this URL.

    From January 2016 I have had 564 sessions by 545 users. I have no complaints :-) as it has given me some data to play with this awesome tool. But I can't understand how is this possible.


    • 38

      Pravin: It is very difficult to answer your question with the provided information. But, your referring url's report, geo report and other areas of GA should be able to help you get closer to the answer.

      Alternatively, you can also work with a GACP to dig in and identify what is happening. You'll find a list here:


  24. 39

    There are so many, many things that I take away from this article.

    Thanks for creating such a handy resource Avinash. I feel like I have one years worth of learning to do.

  25. 40
    Abhinandan says

    Nice information shared and very helpful.

    I want to know that i have filtered my own traffic in google analytics but after few days then it again started tracking my visit also. Can anyone help me on this?

    Anyways thanks for your kind support.

    • 41

      Abhinandan: One contributing element to this occurring is that you are on DHCP and the shifting IP address causes you to show up again.

      I do want to share that I have never been a believer in filtering self-data/company-employees data. If your self data is so big that it is causing your performance to be skewed, you have bigger problems. Stop analytics, go do marketing – owned, earned, paid – to get more traffic. :)

      A GACP could also help look at your data and be more specific in their guidance. A list is here:


  26. 42

    Hi Avinash, and thanks for a hugely interesting and useful article. As always, this was great!

    I know we analysts shouldn't really be focusing on averages in data but I've recently become increasingly worried about an issue regarding data shown in Google Analytics in comparison to GA exported data.

    For example, I'm looking at the Acquisition -> Search console -> Queries report in GA and see an average CTR of 12.94 % and average Position of 9.7 (worrying, yes, but that's beside my point here). However, when I export the same data to Excel (or Google Sheets) and calculate the same averages, I get completely different numbers. Why is this? Is there something I don't understand in how GA calculates the averages? Which numbers should I trust? This is web property and unsampled data, and I went through the rows one by one to make sure that the data itself matches between GA and Excel – just the averages are different.

    • 43

      Simo: The best option for complex GA requests is to hire a GAAC to go through the requirements and validate and recommend the right path. You'll find a list here:

      It is difficult to know what might be up here. Two thoughts though:

      Sometimes when this happens to me, I find I don't have all the rows that GA has to compute it's numbers in the front end vs what I download via excel or the api.

      The second nuance in this case is that GA is simply pulling into the Search Console the data from your webmaster tools account and reporting it to you. This is not GA captured data, just GA reported. Did you check your Webmaster Tools account? Maybe a lead worth exploring.


  27. 44

    Hi Avinash,

    Thanks for this great blog post.

    Even though I'm working with Google Analytics for quite a while already I've still discovered some new things – therefore I just want to say thank you for this great and extremely detailed blog post.

  28. 45
    Roshan Padole says

    Hello Avinash,

    First Of All, Thanks for such a great in-depth Google Analytics Guide.

    This is the key to success for every new comers.

    I have a query about spam visits.

    It is possible to avoid those visits through filters, but new spam visits do come regularly.

    What the solutions for these as these increases bounce rate?

    Thanks in advance….

    Best Regards
    Roshan Padole

    • 46

      Roshan: Unless there is something unusual about your traffic, you should see massively reduced spam visits after March or so of this year. This is simply because the GA team has deployed very smart algorithms to address this problem.

      You can also see this article for manual help: Filter domain referrals

      If you need more help, no problem as you can just ping a GACP. Here's a list:

  29. 47

    Hi Avinash, new here.

    We are working on introducing Google Analytics for our company's inbound marketing strategy but presently need to do it ourselves. Thanks for this article and others like it, they help out in the process.

  30. 48

    This is incredible. I was trying to find a pratical example in how to do Google Analytics with a high volume of data and one more time you have nailed it!

    Can you please let us know if a new/updated version of your book is coming soon?

    Thanks so much,

    Pedro Menezes

    • 49

      Pedro: You are so kind to ask.

      I have been meaning to write a book for a little while now, sadly my professional commitments have gotten in the way. I am hoping that I'll get a break in the near future to write a book. I do have the content!

      Thank you again,


  31. 50

    Hi Avinash! Long time reader/ lurker, first time commentor.

    In the Benchmarking section I was unable to see the data, "there is no data for this view".
    I haven't applied any segments (industry verticals/ country or region/ size by daily sessions).

    I also set date range to 1 Jun 2016 – 10 Sep 2016 to see whether the data would change.

    What am I doing wrong here?

    Much appreciated,

  32. 52
    Sind Hubell says

    This Google Analytics Guide was very useful to my business.

    I refer to this post and find a new nuance about Google Analytics every time I am here.

    Thanks for consistently writing such insightful articles Avinash.

  33. 53

    Great article. Read it from start to end.

    My latest qualm with Google is it's new rules to make it impossible to use the keyword planner without running a campaign. Makes SEO work expensive and time-consuming.

  34. 54
    Michele Perrone says

    Great article!

    I believe that the analysis of the data is vital for any online business.

    You have to be on a constant quest to learn more, then more, then more … :)


  35. 55
    Carl Hewitt says

    Honestly never knew about this public data set.

    This can be very helpful for me in other way. I have watched many seminars on Google Analytics, and it's always a challenge to make demonstrations. No one wants to use his websites, and certainly not going to use a client's site!

    Thank you, Great!

  36. 56
    Charles Mendonça says

    Sensacional esse artigo!

    Confesso que at´´e hoje não consigo configurar 100% o analytcs!

    Muito esclarecedor.


  37. 57
    Mohd. Atif says

    Very interesting and knowledgeable article for Google Analytics Beginners.

    I learn lots of things from this post.

    I appreciate your work "Avinash Sir", you have describe each and everything deathly.

  38. 58

    Oh, wow! This was awesome!

    I try to track button clicks as events. (If the devs are not too busy to add the codes).

  39. 59
    Puneet Singhal says

    Great article Avinash. I consider myself a pro in Google Analytics but still I got so much of value from this guide. Your website is my go to place for anything related to analytics.

    Thanks for sharing your knowledge as always.

  40. 60

    Thanks for sharing Avinash :-) especially the part about event tracking :-)

  41. 61

    Very nice, extensive guide on google analytics!

  42. 62
    Dominique Clement says

    It has been a little while since you'd published this report. I still wanted to take a moment to thank you for not just helping create this invaluable dataset but also writing a post with such valuable starting points for us Analysts. Google Analytics can often feel overwhelming.

    Audiences, benchmarking and enhanced ecommerce were all new to me.

  43. 63

    I'm just learning about enhanced ecommerce thanks to this post, so thanks a lot for taking the time! I also had no idea about this dataset and it's been super useful in my learning.

    I've been wondering, how do you best recommend to set up our targets for KPI performance? Is there another post on this subject I could read? Thanks!


  1. […]
    Avinash Kaushik shared more info on the Demo account and how you might use it available via Occam’s Razor post “Be Real World Smart: A Beginner’s Advanced Google Analytics Guide.”

  2. […]
    In Google Analytics, the Content Drilldown report is fantastic, provided you’ve organized your site by different folders, for example… This report then shows you the engagement stats related to each topic. Avinash Kaushik, Google Analytics superhero has a custom report similar to the Content Drilldown that he recommends. You’re looking to see which how-to topics are getting traffic, engagement and even conversions.

  3. […]
    Recommended reading:
    Be Real-World Smart: A Beginner’s Advanced Google Analytics Guide
    A Great Analyst’s Best Friends: Skepticism & Wisdom!

  4. […]
    Each step is concrete so you can measure which activities (or transitions) lead to the greatest drop offs. Here’s an eCommerce example using the Shopping Analysis from Google Analytics.

  5. […]
    Here’s an eCommerce example using the Shopping Analysis from Google Analytics. Different use case, same steps (or more or less). Instrumenting these events gives you the ability to zero-in on where people are leaking outside your funnel. Here’s how to fix it.

Add your Perspective