You probably missed it, but in December of 2017, the Google Analytics team released a long-awaited new feature: Audiences in Reporting. This feature opens up a variety of new applications and uses in the Google Analytics Suite. I still believe that advanced segments, also referred to as audiences are one of Google Analytics strongest features. These audiences allow you to quickly customize the entire Google Analytics interface to only show you data related to the audience you care about. Not only can you understand everything about their user behavior, but you can also activate these audiences directly within AdWords, DoubleClick Bid Manager, and Google Optimize 360. They are by far my favorite feature available in Google Analytics.

Over the years, Google has added numerous features to these audiences in GA. We can now share them with all users, create state-based rules, and create user-scoped or sequence-based rules. One of the most common questions I can get asked when showing Advanced Segments to new users is, “if there is a way to combine different audiences together or to use one audience that excludes another?” As an example, lets say you built one audience of users who abandoned their cart and a separate audience of mobile users. What if you wanted to easily identify the users that overlapped between both of those segments. The only way to do that in the past was to create a third audience that combined both conditions. Now we have a new option to solve for this using the new audiences in reporting feature.

Audiences in Reporting

Here is what audiences in reporting looks like after you have set everything up.

What changes with audiences in reporting is that you actually publish your audience and Google Analytics will pull them out and put them in the audiences report and make them available for a number of use cases. Before we get too stacked audiences, let’s cover how to publish your audiences to this report. This support article has everything you need, but as a quick summary make sure you have edit permission for the property you want to publish to. Other important details to know is that you can also only publish 20 audiences to Google Analytics, sequence-based segments cannot be used, and currently, they do not pre-populate data, so they will only be available from date of setup moving forward. With this in mind, the process is as follows:

Step 1: Create Your Audience/Segment

Simply click the Add Segment button from any report in GA, name your segment, add your definitions, and click save.

Step 2: Select Build Audience

Click the Arrow next to the segment at the top and click build audience.

Step 3: Select Build Audience

Name your audience again(I wish Google would pre-populate this since we already named it) and then select Google Analytics as your destination.

After 24-48 hours when you come back to Google Analytics your audiences will be available in the Audiences Report.

Stacked Audiences

Now that we have published audiences available in Google Analytics, we have them available when we create other audiences. If I have audiences published for current shoppers and intense athletes, I can use both to create a new combined/stacked Audience. This is why Audience in Reporting is one of my favorite new features in Google Analytics! I recommend working with your Analytics team to create core audiences that are applicable to your vertical or use cases, so you can leverage them throughout the Google Analytics Suite. Hopefully, Google will make an update soon, so that they can be pre-populated and we do not have to wait for them to collect data.

The Data Studio team at Google has been on fire lately! I highly recommend subscribing to their blog, as they continue to release important new features. This past month two highly requested feature have been announced or previewed, which are Embedded Reports and Data Control! I want to give you a quick preview of both!

Embedded Reports in Data Studio

Adam Singer from Google uncovered this feature in a recent post announcing filter and chart updates. I took one of the embedded charts from the post and included it below! This is so exciting, because we can finally remove the restriction of having to access Data Studio reports solely from a URL. Now charts and reports can be embedded in any webpage, however you like! The embeds appear to retain all functionality of a regular report, including the interactivity. Try using the Device Category selector to filter the data! Also, note that you can change pages in the bottom left corner.

The embed feature does not appear to be publicly available yet, but it sure looks ready =).

Data Control in Data Studio

One of the most commonly requested feature that I’ve been asked about is whether you could create a Data Studio report and easily use it as a template across multiple Google Analytics properties, views or even accounts. To date, you had to physically make a copy of that report and then connect it to an additional data source, which could result in you creating hundreds of duplicate reports. Data control removes all of these limitations and takes templates to the next level!

You can play with Data Control, by changing the data source to any Google Analytics account you have access to in my Dashboard that I created in 10 minutes: https://datastudio.google.com/open/0Byw3d6m14ln-eXF0bTV0TDJPamM

Even better, Data Control can apply at any level of granularity within the report you create. That means I can scope the data control only to a specific chart or set of charts, by simply grouping the charts with the data control.

Any expert of Google Analytics will likely tell you that you have to have goals setup in Google Analytics. Today, I am going to argue how goals have largely become irrelevant and share some ideas on how to better measure outcomes and conversions. To illustrate my point of view let us examine how a goal was created ~10 years ago using Google Urchin, which was the predecessor to Google Analytics:

Goal Setup in Google Urchin

Goal Setup in Google Analytics Today

The problem is while everything around how our users convert has changed with cross-channel and cross-device touchpoints, the definition of a conversion in Google Analytics has never changed.

Why Goals Do Not Work in Google Analytics

• Goals are session based NOT user based
  • If you have a user that visits four times and then converts, your conversion rate for that user is 25%. You cannot change this.
• Goals can only increment once per session
  • If a user adds four products to cart, your add to cart goal is one, not four. You cannot change this.
• Goal funnels/flows are broken
  • Try to read both of these articles about Goal Funnels from Lunametrics from 2008 and 2010. I could spend an entire blog post on why these funnels are broken. They are visually unappealing, are session-based, and just do not work. After all this time, you still cannot segment a goal funnel. Why can we not build event based funnels without Google Analytics 360? The ‘new’ goal flow reports are equally flawed.
• Goals cannot be deleted or be reused
  • Make a mistake and that data lives forever. Try to reuse that goal for something else and you are still stuck with the historical data.
• You cannot have more than 20 goals
  • Try running a Google Analytics account for 10 years with only 20 goals that cannot be deleted/reused. Google Analytics 360 also does not include more goals.

Alternative: Custom and Calculated Metrics

Custom and Calculated Metrics was one of the most important features added to Google Analytics. They are not perfect, but they give a significant amount of flexility over Goals. Calculated Metrics are retroactive, can be based on users, and can be utilized in custom reports and dashboards. The major limitation is that they are not pre-applied to any of your existing reports, so you have to build custom reports or use the API to leverage them effectively. I wrote another post showcasing their power: https://www.analyticspros.com/blog/google-analytics/25-calculated-metrics-for-google-analytics/

I wrote this blog post today with hope that Google will bring updates and new capabilities to Goals in Google Analytics and make them great once again!

Advanced Segments is by far the most useful feature in all of Google Analytics. I believe that your ability to use them are what separates you from a beginner Google Analytics user. Today, I want to share a quick tip to make them even more useful. One of the limitations that I often hear, especially from Adobe users, is that you cannot created a hit-scoped segment in Google Analytics. Let’s start with why this is even important.

Advanced Segment Scoping

Say we want to analyze information related to what happens before and after someone adds a product to the cart. We can start by creating a segment of any session/visit where an event action contains add to cart. The segment below will isolate any session where an add to cart occurred and return everything in that session.

What if we wanted to know how many of these users that added a product to the cart, also looked at our spring sale page? This is where we get into scoping and order. Do you want to look at a visit/session where both actions occurred, do you want to look at users where both actions occurred (even if they happened individually in separate visits), or do we need to establish an order and define that the user must first see our sale before adding to cart? Depending on the question you are trying to answer any of these may be appropriate. If we want to go with the last option of anyone who viewed our sales page and then ever added a product to bag, we need to change our scope and conditions. Here is how we would do that:

Hit-Scoped Advanced Segments

Now for the quick tip. What if we needed to ensure that a user who viewed our sales page, actually added a product from that sales page to their cart? The conditions in the segment above do not ensure this. They simply state the user added some product to cart and that product may or may not have been included in that sale. We can achieve this by adding conditions to our sequence like this:

Pretty simple, but something pretty important happened here. What we did, was use a new ‘hidden’ scope for hit-scoped segments and conditions. To further build on why this is important lets try and build segments to identify users who added a specific product to their cart. If you are not familiar with how to use hit-scoped you could get very inaccurate data. Lets create a segment of any session where a Metal Roller Pen was added to cart. The two segments built below show the data discrepancy. The first segment shows us anyone who every interacted with a Metal Texture Roller Pen and also added a product to the bag. That interaction with the Metal Texture Roller Pen could have been a product detail view or even a product impression. The conditions are evaluated across the session and not against each other. The second segment shows how to enforce and evaluate the conditions at the same time.

Now you know why you need hit-scoped segments and conditions and how to build them =)

Excited to share on March 27th, I will be leading a 4 week live class on Intermediate Google Analytics. Head on over to Conversion XL and sign up for the upcoming course! If you are interested in a preview or would like benchmark your expertise in Google Analytics, you can check out the webinar we just did!

 

Can you answer these questions today around attribution, your users and your conversions?

• Do I even need to be thinking about attribution?
• How long does it take my users take to convert?
• How many visits to my site occur before converting?
• How is my marketing working together?
• Am I measuring Display optimally?

I will show you how to answer these questions in under 10 minutes using Google Analytics. First a reminder!

Default Attribution in Google Analytics

Google Analytics uses a last non-direct attribution model in all of its core reports. This means that the last traffic source gets all the credit, unless that last visit was direct and a previous traffic source existed. Most marketers will either use this default model or the default models provided by their marketing platforms such as Facebook, AdWords, or Bing. The majority of these platforms all use a greedy model, where they take all the credit, even if another channel was first or last.

I want to help you understand if these are the right models or if you need to be having a larger conversation around attribution. You will be surprised to learn how easy this conversation can be when you have the data to support it. Let’s dive in!

Prerequisites

• You have to have goals or transactions configured (A Google Analytics must have!)
• You need to campaign tag your marketing (Otherwise how can you attribute anything?)
• You need to know the basics of how to use Multi-Channel Funnels (It is easy, I’ll show you below!)

Multi-Channel Funnels 101

All of these questions we are about to answer, we will answer with Multi-Channel Funnels (MCF) in Google Analytics. MCF is by far one one of the more powerful features. In MCF, we have a linear attribution model where every channel shares credit equally. Only in MCF can we see the true impact of direct traffic and easily see in detail all of the touch points our users went through prior to converting. This allows  us to answer all the questions posed above.

As we  walk-through the reports answering these posed questions, it is important to know how to take my examples and make them relevant to your business. To do so, remember and apply this one tip!

Selecting Specific Conversions

In all the MCF reports it is easy to select an individual or even a group of conversions. You will need this to focus on your specific conversions such as purchasing, filling out a form, subscribing, etc. Click the conversion drop down and select whichever conversion is most applicable.

Now we are ready to answer those questions!

1. How Long Do My Users Take to Convert?

How long does it take your users to convert from they time they first found you? This question is essential not just for answering attribution questions, but more importantly in driving your marketing and messaging in the right places and at the right time.

Using the default Multi-Channel Funnel reports you can easily answer the question of consideration. Below you can see for the Google Merchandise store that for users who purchase, 39% do so 1 day after their initial visit. Looking deeper, we can see that 22% of those conversions and 28% of overall revenue comes from users who purchased 12 or more days after their initial visit. Now apply this to your Google Analytics Account: https://analytics.google.com/analytics/web/#report/bf-time-lag/

2. How Many Touchpoints Occur Before Converting?

Understanding how many touchpoints occur before your users convert, can shape how your organize your channel teams. Most organizations have their traffic and engagement teams siloed, meaning they rarely work together.

Why is this important? Recently, I was onsite with a client who was explaining how they incentivize their social, e-mail, and SEO teams. They paid commission to these teams based on their targets using the default attribution reports in Google Analytics. Remember the default attribution in Google Analytics is last non-direct touch. What was happening as a result is that each team was fighting to be the last touch and all were heavily offering discount codes, remarketing efforts, and as a result competing with each other.  The business was under the assumption that their customer rarely came from multiple marketing activities, but in 15 minutes Google Analytics was able to show that this assumption was wrong. As a result they reworked their commission model and reallocate a portion of their budget for new cross-channel campaigns.

To answer the question around how many touches your users go through, you go to the Path Length report in Multi-Channel Funnels. Using the Google Merchandise Store we see that 64% of transactions and 70% of revenue occurs from users that had 2 or more visits prior to purchasing. We know definitively that attribution matters, as the majority of users purchasing do so over multiple days and visits. Try it on your Google Analytics Account: https://analytics.google.com/analytics/web/#report/bf-path-length/

3. Do I Even Need to Be Thinking About Attribution?

If your data looks like the Google Merchandise’s store where the majority of conversions occur over multiple days and visits, the answer is obvious. However, I do work with some clients where the vast majority of conversions occur from users who take one day and one visit to convert. If your product or offering has a quick consideration period like that, you can use any attribution model in the world, because they will all likely be correct. If your conversions are more prolonged, then you absolutely need to be having conversations and shifts to a better attribution methodology.

You can answer this question by simply using the two techniques above. How long do your users take to convert and how many touches occur prior to converting?

4. How is My Marketing Working Together?

Does your marketing work together? Can you answer these questions?

• How many of your conversions are from users that came from E-Mail and Social?
• How many of your conversions are from users who only ever came from Google?
• Is social more of an upper funnel or lower funnel?
• For Paid traffic are my brand and remarketing campaigns working together?

You can answer all of these questions easily using the Top Conversion Paths report in Multi-Channel Funnels. Here is a quick start guide for how to use it. Most important item is to change the path length to all from 2 or more, so you look at all conversions, not just assisted conversions.

The magic of the top conversion paths reports comes from when you use the secondary dimension option to add in campaign, keyword, source or any other traffic source attribute you need. The image below shows how easy it is to do a deep-dive into paid conversions as an example. Using the secondary dimension button to add the campaign dimension, while filtering for only conversion paths that involved paid gives you the exact campaigns involved and when they occurred. We can see insights right away by looking through this list. Of the top paths shown, all of them resulted from paid being the first touch, proving that paid traffic is bring net-new convertors to my site.

5. Am I Measuring Display Optimally?

Display can be one of the most challenging channels to understand. Display is typically a mix of upper funnel and lower funnel activities. Upper funnel being the branding campaigns, while lower funnel is usually the remarketing campaigns. What makes this analyzing display so challenging? View-through conversions do. View-through conversions are users who saw your ad, did not interact with them, but later somehow came back to your site and converted.

The question that always comes up is what percentage of those view-through conversions, actually converted because the saw an ad? Had the display ad impression not been present, would they have converted anyways? This is especially important for branding campaigns, where the intent is to bring in net-new users. Google Analytics is one of the only platforms that can unlock the mystery of Display and view-through conversions. If you are running display through AdWords or DoubleClick you can reach out to your account manager and have them enable view-through conversion in Multi-Channel Funnels. Once you do that you can come to this report. The eyeball icon indicates a view-through impression and anything missing the eye-ball is a click-through. This is one of the most powerful integrations Google provides!

This post is just the start of the powerful attribution features that Google Analytics has to offer. With an Attribution Modeling Tool, Data-Driven Attribution, and access to raw data in BigQuery you can leverage out of the box feature or build your own models. In a future post, I will detail where to go from here, but with this post I wanted to provide you everything you need to start having intelligent conversations around attribution and its importance to your business!

Google Surveys

Google Surveys which was recently added to the Google Analytics 360 Suite, allows marketers to create, launch, and analyze surveys and perform market research. Google Surveys offers the ability to survey your own website or their panel of more than 10 million  online respondents.  This post will walk through the creation, deployment and analysis of both survey types. After reading, you can even launch a survey through Google Surveys on your website at no cost.

How Google Surveys Work

Google Surveys uses a network of online publishers and its popular Google Opinion Reward app on Android to build is pool of respondents. You have probably seen Google Surveys embedded on your favorite publisher sites embedded in the content or know a friend who uses the Android Opinion Reward app to get Google Play credits. Google Surveys is able to use their respondents demographic & geographic information to build representative samples. By using stratified sampling, Google Surveys is able to perform equal or better to existing probability and non-probability based Internet survey polls. Much more detail can be found on the Google Surveys Help Center.

Get Started with Google Surveys

In minutes you can create and launch a Google Survey. Start by going to surveys.google.com and log in to your Google Account.

Step 1: Select Audience & Targets

The audience types available for Google Surveys include the general population, Android users, your website and Audience Panels. Within the general population and Android users you can target specific age ranges, genders, countries and sub-regions. For our first survey we will select the general population.

Step 2: Write Questions

The questions formats available are single answer, multiple answers or a rating scale. My favorite feature of Google Surveys is the ability for you to screen respondents. Say we wanted to find out what percentage of voters for the 2016 election voted for both a Democrat and a Republican or Independent. We could create a two question survey and use the first question to screen out users who did not vote.

Here is the screening question, which allows for a single answer. Note that screen in checkbox for yes is checked.

Here is second question, which allows for multiple selections.

Step 3: Confirm Pricing

If you used a survey that screens respondents like the example above, Google Surveys will run a free trial run to measure your audience size and determine pricing. The trial run is free and takes 1 to 24 hours to complete. I got pricing with 4 hours! If you use a non-screened survey you can immediately proceed through to pricing & billing. Pricing for 1 question surveys is between 10-30 cents and 2+ question surveys are between 1-3 dollars. You only pay for complete responses, meaning the respondent must complete all questions. For 1000 responses to my survey below the cost is $100.00. Payments accepted are credit & debit cards.

Step 4: Verification

All surveys require verification and approval from Google to ensure you are in compliance. I actually found this to be extremely useful as I received a reply from Mike on the Google Surveys Team within 30 minutes. I incorporated his suggestions into my survey and my survey was live within an hour of creation.

Step 5: Analyze & Share

My survey was live within 2 hours of creation and already starting to collect respondents. There are two ways of viewing the results. The default shows the confidence intervals and uses weighting to remove bias and closely represent the target population. This is why the 9 results below do not divide evenly into the answers. You can always select the raw counts in the top corner to access the raw results.

You can do a variety of analysis directly in the tool. To interact with the data you can view my survey results: https://surveys.google.com/reporting/survey?survey=c63foe5uhh66eezuxam5qxvnum. Google Surveys allows you to make your results publicly available, private or shared with specific users. In addition you can easily export the results to a spreadsheet for deeper data mining or visualization.

Onsite Surveys

You can also use Google Surveys for targeted surveys on your site. Google Surveys even provides a free website satisfaction survey that you can use to gather voice of customer research from your own site. Get started here! You can see from the image below that you can control the timing of the first survey and configure the advanced settings to ensure you are not showing your survey too frequently.

To install simply place the code snippet in the </head> section or your site or better yet if you use Google Tag Manager you can install the tag in minutes. Using Google Tag Manager you can also easily control the scoping of the tag to only target sections of your site such as the blog, support or contact pages. You can also deploy custom website surveys at a cost of 1 cent per response.

Google Surveys 360

There is also an enterprise version available as part of the Google Analytics 360 Suite. Google Surveys 360 provides advanced targeting options that include zip codes, occupation and industry, more custom panels, and most importantly remarketing. Remarketing allows you to reengage with your users to gain additional insights. A retail store could target their users who left the site on a product detail page and ask them if they bought their product in the last 30 days, which provides rich insights into campaign performance. Also included is invoiced billing, enterprise pricing, and SLAs.

Google Surveys 360 versus Google Surveys

Source: https://support.google.com/360suite/surveys/answer/6244101?hl=en&ref_topic=7171368

Google Analytics has advanced video tracking capabilities that even the biggest media companies can leverage, however, almost no one knows this. There are limited resources on video-centric solutions. Worse, almost all current implementations are outdated and do not take advantage of core features like custom metrics and calculated metrics. Another problem is with the so called integrations provided by companies like Brightcove. Brightcove’s ‘integration‘ has this great callout:

If you already have a account in Google Analytics created for your web site to track page views, we highly recommend that you create a new account for the same web site that you will use for tracking Video Cloud data only. If you don’t, then you need to be sure that you understand the effect that video player interactions can have on items like the bounce rate.

By separating video data into a new account you are unable to understand something as simple as how many users who watch a video convert. My intent with this post is to showcase Google Analytics’ capabilities and provide a reference point for getting video platforms to update their integrations. If you have code access to your video player on your site, you will be able to do everything in this post!

Questions You Should Be Able To Answer

Using your current tracking can you you answer any of these questions today?

• What is the average amount of playtime watched per user or per video?
• Do you track a video start differently from a play after a pause?
• Do different countries/regions/devices engage with video differently?

These are all questions that you should be able to answer. Working at Analytics Pros, we have completely rebuilt vendor video analytics integrations for a number of media-driven companies. We have tracked media on almost every device including Xbox, Apple TV’s, and LG Smart TVs. The solutions below have been implemented across web, mobile apps, and connected devices and provide sophisticated analytical capabilities.

Current Video Tracking Reports

Lets start by taking a look at what your current video tracking likely looks like in Google Analytics.

This is the most common type of integration provided by most video platforms. It provides data on what videos are being watched and provides insight into what percentages of videos are being watched. The problem is this provides no insight into how much time of video is being consumed and often times has many underlying implementation issues. For example, most solutions I have reviewed are unable to distinguish a video start from a play that occurs after a pause.

Proper Video Tracking Reports

Below is a series of reports of what your video tracking should look like. At the very end, I include examples of dataLayer pushes for Google Tag Manager that drive all of these reports. I am going to use Major League Soccer (who I would love to work with) as my example. Major League Soccer provides their users the ability to livestream games, as well as watch videos that are hosted on their site and apps.

Here is what their dashboard for video should look like in Google Analytics:

What makes this dashboard different is that it features video playtime as metric. Using this we can produce reports that center on video playtime. We can add video playtime to provide insights into which traffic sources, countries, or devices consume the most video in aggregate or average. We can also provide insight into how often videos are stuck buffering and what bitrates users are being served.

Livestream is also something that is often viewed as challenging to track in Google Analytics. Again, the video playtime metric makes it look easy. Video playtime is a custom metric that is passed with how many seconds of the stream a user watched. The most important addition here though is the stream minute that we showcase in the reports below. Using the stream minute we can report for any livestream what minute during that stream had the most active users.  I pulled the data from the Google Analytics API through the Google Sheets connector and created a visualization of the video by this stream minute to provide a minute by minute active user view. The visual below shows an example of what a soccer game being livestreamed might look like. You can see that active users peaked right before half-time and again at the end of the match.

Implementation Specifics

To make all this happen you need to implement solid eventing in your video platform. Here are some examples of our dataLayer pushes that power everything in Google Analytics:

Video OnDemand – Start

dataLayer.push({
     'event': 'video-on-demand-start', //do not fire for play after pause
     'video-id': '<video ID>',
     'video-name': '<video name>',
     'video-bitrate': '<video bitrate>',
     'video-autoplay': '<true or false>',
     'video-platform': '<Flash or HTML5>'
});

Video OnDemand – Compete

dataLayer.push({
     'event': 'video-on-demand-complete',
     'video-id': '<video ID>',
     'video-name': '<video name>',
     'video-bitrate': '<video bitrate>',
     'video-autoplay': '<true or false>',
     'video-platform': '<Flash or HTML5>'
});

Video OnDemand – Duration

dataLayer.push({
     'event': 'video-on-demand-duration', //pass for every 15 seconds of playtime consumed
     'on-demand-seconds-watched': '<seconds watched>',
     'video-id': '<video ID>',
     'video-name': '<video name>',
     'video-bitrate': '<video bitrate>',
     'video-autoplay': '<true or false>',
     'video-platform': '<Flash or HTML5>'
});

Video Livestream – Duration

dataLayer.push({
     'event': 'video-livestream-duration', //pass for every 15 seconds of playtime consumed
     'minutes-from-start': '<minute>', //what minute into the livestream from start are they at
     'livestream-seconds-watched': '<seconds watched>',
     'video-id': '<video ID>',
     'video-name': '<video name>',
     'video-bitrate': '<video bitrate>',
     'video-autoplay': '<true or false>',
     'video-platform': '<Flash or HTML5>'
});

These dataLayer pushes need to be combined with Custom Metrics, Custom Dimensions, and Calculated Metrics in Google Analytics to power all of the reports built above. Below are what the foundation of these should look like:

Custom Dimensions in Google Analytics

Custom Metrics in Google Analytics

Calculated Metrics in Google Analytics

This post provides a solid foundation to build on. My hope is that this can be referenced to push providers like Brightcove to create better integrations. If you need any help with your video tracking needs, please contact me over at Analytics Pros.

Custom Funnels are the best new feature in Google Analytics, since Multi-Channel Funnels came out. In this blog post I will share with you everything you need to know about Custom Funnels for Google Analytics. Lets say you have implemented YouTube tracking using readily available scripts from LunaMetrics or Cardinal Path. For the first time you can visualize how users consume video!

Custom Funnels: Features and Benefits

Custom Funnels is currently a Google Analytics Premium only feature. The value this feature can provide for many organization will be a huge use case for the investment in Google Analytics Premium.  Custom Funnels address a number of key limitation that have long existed in Google Analytics. Here are the best features:

1. Visualizations can be created that include sequences that include both pages and events at the same time.
2. Unlike the Funnel Visualization report, when you setup custom funnels they are retroactive and utilize historical data.
3. Unlike the Goal Flow and other Behavior Flow reports, Custom Funnels are available at a much higher sampling rate. Behavior Flow has sampling kick in at 100,000 sessions compared to Custom Funnels where I’ve seen it kick in at 50,000,000 sessions.
4. Custom Funnels can be user scoped. This means you can have the steps in the funnel occur across sessions. You can finally create true User-level checkout funnels.
5. Unlike the Funnel Visualization report you can add advanced segments to your Custom Funnels.
6. Integrate directly with Google AdWords to send Funnel Segments for powerful remarketing lists.

In summary the best part is that once these funnels are created they work retroactively with historical data, will be completely unsampled for most reports, can be user or session scoped, and integrate directly with Google AdWords for creation of remarketing lists.

How To Create Custom Funnels

Lets get started on creating your first funnel! You can create one by clicking on ‘Customization’ in the top menu bar, click the ‘New Custom Report’ button and lastly click the ‘Funnel’ tab in the type section, which is directly underneath the title and name fields.

Overview: Funnel Builder

This image covers a quick overview of the features you need to utilize and know are available.

There are some really great feature in here. Including custom dimensions and goals there are over 200 different dimensions available for you to choose to include in your funnel. No other visualization in Google Analytics allows you to get this granular. With regular expression and exact match types, in addition to the options to include or exclude values, you will be able to match a wide range of use cases and different funnel types. I am a big fan of the time saving options that Google has included, with the ability to quickly duplicate steps, copy the entire funnel, and save it to any view needed across accounts.

Advanced Features:

There are more features available in the ‘Advanced Options’ dropdown. This is where you can specify to make your funnel open or closed. Currently, open funnels will only be available if you select the ‘Analysis Type’ of ‘All stages must occur within a session’, which is above  the dropdown. The open funnel allows the user to enter into the funnel at any of the steps. The visualization will explicitly show these additional entrances within each step. Next, you have the ability to change the metric to sessions, instead of users if desired. Lastly, just like Advanced Segments that use sequences, funnels also have the ability to specify when the next action in the funnel occurs. You can specify that all next steps occur at any time after, directly after, or specify it individually per step. Selecting ‘immediately after a stage’ means that it has to be the next hit sent. If any hit such as a pageview or event occurs before the next path the user will show as an exit in your funnel.

Funnel Creation

Here is an example of a 5 step funnel. For this we want to follow a specific path to understand how many users peform the following actions in order: visit our blog, watch a video, visit our store, add an item to their cart and ultimately complete a purchase. You can see the power of this feature in the image below. We have used five different dimensions with hostname, event category, event action, page and transaction ID to achieve this.

Below is the resulting custom funnel that is presented when this report is saved. I’ve highlight some of the most important aspects. The report is available completely unsampled and in this instance pulls in over 10 million sessions. The funnel, which should look very familiar if you use Enhanced E-Commerce, shows how many users complete each step as well as abandon. By clicking on any of the blue bars or the abandonment boxes a segment editor is shown. Using this you can quickly create funnel segments and even immediately have them available in AdWords to utilize in remarketing campaigns to recapture lost revenue. The last feature pointed out is that you can apply Advanced Segments, which would allow you to compare Desktop vs. Mobile or New vs. Returning or any other segment you would like to create.

Missing Features

Just like every good consumer, I always want more. There are a few features that I hope are added to Custom Funnels to make them even more powerful. I have found that only being able to create at most 5 steps is very inconvenient and limiting. There are work arounds I have used by creating multiple funnels with the same ending and beginning paths, but it requires a lot more effort to get them to work right. It is also really annoying to not be able to have to click on each Advance Segment applied to the report, instead of simply having them shown side by side. Unsampled exports need to be added, as I’ve found with very large data ranges sampling can be a factor. It would also be nice if this was available for all Google Analytics users, not just Premium customers =)

If you are a Premium customer there is not excuse for not using these. If you are currently considering Premium, this would be a great feature to convince others in your organization of the value. Already, Custom Funnels is now an essential tool for my reporting and insights that I perform for my clients. If you have any questions about anything Google Analytics or Google Tag Manager related, leave a comment or reach out to me on Twitter.