Archive for the ‘web analytics’ Category

Mar 08

combine qualitative and quantitative dataA couple of weeks ago, I wrote a post about marrying qualitative and quantitative data, giving recommendations on how to conceptually combine both types of data for more useful insights. For the savvy marketers and the analysis ninjas out there, you want a tighter integration, and of course, more segmentation on the qualitative data. Visitors told you “No, I didn’t accomplish what I came to the site for” and “that pricing was too high”. Now, maybe you want to know…

  • …where these “No’s” and “Yes’s” are coming from
  • …what content they are consuming
  • …what campaigns are driving these segments so you optimize at the source

In this post, we’ll cover this – the insights you can get from a direct integration of Google Analytics and Qualaroo.

Google Analytics – Quantitative Data Tool of Choice

It’s no secret that Google Analytics is our web analytics platform of choice. It’s a “monster” when it comes to providing web usage stats. It’s the gold standard. And as far as pricing, the cost benefit is unbeatable – it’s free (and if you want more data and enterprise analytics consulting, Google Analytics Premium is there for you). When planned and implemented right, you can get a wealth of qualitative data on your website – what pages are visited, where they come from, what keyword they searched for, how long they stayed, who converted – etc etc etc.

Qualaroo – Non-Invasive Voice Of Customer Survey

As far as qualitative data, Qualaroo can be your best friend. As an internet surfer, I’m sure your feelings are consistent with the rest of the world – forms and surveys are kind of annoying. They’re long, tedious, and often a waste of time. What we like about Qualaroo is how non-invasive it is. A small question will pop up and a visitor has the option of quickly answering it, or quietly dismissing it. We feel this increases the likelihood of it being answered.
We also love its configuration flexibility. “Nudge” the question after 30 seconds. After a visitor has visited 2 pages. Publish the survey from a certain date, shut it off after a certain date. Display only on certain pages. Etc.
Note: to integrate Qualaroo data in Google Analytics, you need to sign up for Qualaroo’s enterprise edition.

Google Analytics and Qualaroo Integration

With the ability to integrate Qualaroo data into GA, you now have a very powerful surveying, measurement tool. You can see “what’s” happening on your site by looking at the Google Analytics metrics. Then, with Qualaroo integrated into Google Analytics, you get a deeper understanding about “why”. Google Analytics itself will tie the connections between your qualitative data and quantitative data, between your “what” and “why”.

ROI by Channel

Here’s a simple example. Your company sells a SaaS product targeting CMOs and Marketing Managers at medium and large companies. You have your paid search campaigns running on AdWords and Bing and you want to expand into new paid social ad campaigns. You have ads on the following social networks:

  • Facebook
  • Twitter
  • LinkedIn

Your landing page is converting, but it’s not converting like it should be, and you’re looking to really cut down your ad spend and optimize ROI. Your quantitative numbers tell you the following:

  • Your spend is about equal (budget was set by yourself).
  • There actually is a lot of traffic from Facebook and Twitter and LinkedIn.
  • However, it looks like you are getting very few conversions.
  • You wonder:
    • “Why is that?”
    • “Should you continue spending on these channels?
    • If you shut down these campaigns, what opportunity are you missing?”

Thus, we add the following survey:

Qualaroo Survey - Did you accomplish what you came for

Your goal is to understand what’s happening and why your visitors can’t find what they are looking for and thus not converting. With Qualaroo Enterprise version enabled, some cool Qualaroo event user interaction can be reported in GA (an event is fired based on user interaction with the survey). The image below shows the number of events when responders answer “No, we didn’t accomplish what we came here to do”. Here is what we see in GA (go to Content –> Events –> Top Events –> and then click on the Qualaroo Enterprise Event Category):

qualaroo google analytics events aggregate

Your first reaction is :( , so many people (and a high percentage of the total responses) can’t find what they are looking for, no wonder they are not converting. Time to dig deeper and segment.

As we know, once you have the event captured in GA, you can make it a goal or segment on it in whichever way you like. In our case, we want to understand where the No’s are coming from. Let’s take this specific event and segment the report by channel (apply secondary dimension to the Top Events Report) and voila! You have the channels responsible for driving the unengaged visitors.

Google Analytics Qualaroo exvent segment by Channel

In the snapshot above, you see LinkedIn is a potential culprit here. We love LinkedIn and we’d expect it to generate high value leads, what’s going on here? Our paid ad LinkedIn campaign appears to be bringing visitors who are not interested in the SaaS product. After verifying the statistical significance (and correlating the number of “no” events with the traffic volume from the respective source) and looking at some of the survey responses, so of the answers to tell us why you didn’t find what you came here for were very revealing, here is a sample of these answers, once you read them, you’ll know exactly why the campaign is not converting.

  •  ”I can’t find the technical product description”
  • “your product appears to be solid, but I am interested in customization features and API import/export capabilities”
  • “how technical does the system admin need to be, I can’t tell from the info provided”
  • “I am a system admin and I need to know your system up-time and technical support you provide, can’t find this information”

Most of the answers/comments were of a technical nature, and appears to come from IT folks and your LinkedIn ad campaign (including the landing page) is all designed around speaking to the CMOs!

These insights are very actionable. Two immediate actions are now taken to adjust our targeting options, and really optimize our ROI (by maybe putting our budget towards these better channels).

  1. Cost savings: Enhance the LinkedIn targeting to marketing managers
  2. New opportunities: create a new campaign/new landing page with information requested for the IT folks.


ROI by Keywords

Now, say you’re spending on Google AdWords. You really want to know if specific campaigns or keywords are driving the traffic and unengaged visitors.

Again, we take the specific event and segment the report by keyword (apply secondary dimension to the Top Events Report) and voila! You have the culprit keywords.

qualaroo google analytics events keywords

Is this super insightful or what?! (keywords blurred to keep our clients identity anonymous…)

Stop bidding on these keywords or direct traffic from these keywords to a lower price-point product and see your sales go up like you’ve never seen before!

Here you have it, if you feel like you are hitting a wall and not sure what to do to improve your conversion rate, go back to the basics and “listen to your customers”!

How have you used voice of customer tools to improve your site and campaign performance? Very interested to hear your stories.

Feb 27

Googel Analytics Real TimeWhen working with high volume sites, slicing and dicing the data can be challenging, and looking for insights is very much like Tom Hank’s famous quote from the movie Saving Private Ryan – it’s like “finding a needle in a stack of needles”!

At E-Nor, we’re proud to work with some of the most recognizable brands in the world. Our clients expect sophisticated measurement and deep insights, more than “ancient” basic metrics like “total visits” or “site wide conversion rates” :) .

So what do you do? One thing I’ve learned over the years (the hard way at times!) is that as a consultant, and the same applies to marketers and practitioners, you should focus on what matters most to the business. Find creative ways to measure/report/analyze and find “gems” that truly impact the business. Along the way, that will make you look like a hero (and will ensure long-term employment!).

In this post, I’ll cover the challenges and solutions for understanding user behavior and cohort analysis across multiple platforms. The business we’re using as example is a large media organization where you “Sign Up” to access content. You then have the option to upgrade and buy a “Premium Subscription” as well as access to buy exclusive products and services. And in case you are just getting used to this type of analysis, a typical cohort analysis will show you the behavior of a “class” of visitors, typically segmented by an action on a specific date or date range.

We’ll cover:

  1. The Cross-Platform Challenge
  2. The Measurement Approach
  3. Integrated Reporting & Insights

1- The Challenge – Web, Mobile and Offline

In this multiple device, multiple platform world, understanding user behavior and measuring “what matters” require planning (and a bit of process) way before jumping into tagging and JavaScript. Especially as we start shifting our thinking from measuring visits and conversion rates, to visitors and customer lifetime value. I won’t get into the technical instrumentation of cohort analysis (just Google it, you’ll find many posts on the basic “how-to”). I’ll focus on the user-centric approach and marketing insights.

“Out of the Box” Cohort  - Mixpanel Analytics
Some analytics tools have cohort analysis built in their standard reports. For example, you could be using the Mixpanel platform (and assuming you’ve planned and implemented your events properly), you’ll get very meaningful cohort reports that can shed some light on user behavior over time. Say you have a content site where a subscription to access is required, you’d want to know if your users are coming back and consuming that great content you produce, if a cohort of users whom you acquired in a specific time period, are likely to be more loyal than others.

A standard Mixpanel cohort report will look like the snapshot below:

cohort mixpanel screenshot

The first column shows the date at which the “Sign Up” occurred. The “People” column shows how many people signed up on that day (e.g. 10,324 on Feb 5th, 2013) and the percentages represent the percent of people who come back after x-amount of days (where x is 1 to 12 in this chart). So for the Feb 5th cohorts, 1.84% of them came back and consumed more content two days after signing up.

In addition and without a lot of digging, you can clearly see that the segment of visitors who signed up on February 8th are super engaged in the first seven days, and they are coming back for more every other day. On the other hand, those who signed up on February 6th, behave completely different. They are interested initially and then their interest taper off.

Campaign Segmentation in Mixpanel
You are hungry for more insights, what do you do? Jump over to Mixpanel’s Segmentation reports (you can segment on the fly by region, referral and even Google Analytics utm parameters) and dig deeper to find the source of those who signed up on the 8th. You’ll notice here that the campaign that peaked on the 8th (in green) is likely to be a very good driver of such an engaged audience.

That’s good to know, but you are still not yet satisfied. The data above is from the web channel only. You are missing other acquisition channels.

In line with our cross-platform and the 360-degree customer view approach, let’s take this analysis to the next level for a large media site. Visitors register to access and share content, and with a paid “Premium Subscription” option then have access to discounted/exclusive products and services. You’d want to track user behavior across all potential channels on which a prospect might convert. A prospect can buy a Premium Subscription:

  • On the website
  • On their mobile app
  • Dial the 800 number and sign up “offline”

2- Solution Design – Analytics & Measuring Across Platforms

This large media site uses Google Analytics Premium and also has Flurry Analytics for their mobile app. But no matter what analytics platform you are using, this multiple-way conversion presents quite a challenge to cohort analysis in any one analytics product. Time to stitch and tie!

To have the 360-degree view of the customer, and stitch all the disparate data from the web, mobile and offline, there is a technical implementation aspect and a little bit of process. So yes, work is required, but the anticipated results are definitely worth all this pre-requisite hard work.

The highlights of what is needed include:

  • Web data — extract data from Google Analytics Premium
  • Mobile data — download your events data from Flurry
  • Offline — export your purchase data from your backend

And here are the details:

Web Analytics: Google Analytics Premium unsampled reports with the following metrics/dimensions:

  • Revenue
  • Date of transaction
  • Cohort date – The date the Premium Subscription took place, and is stored in a visitor-level Custom Variable 1
  • Unique customer ID, (of course, no personally identifying information), stored in visitor-level Custom Variable 2 (we will use this for the advanced user-centric tip at the end of the post)

Note: As previously stated, this is not a technical post, but here is how you’d set up Custom Variables. You’ll also need the help of your IT/dev team to coordinate the passing/managing of the unique ID in the backend system.

In Google Analytics, run an eCommerce -> Sales Report, and choose your Cohort date (CV1) as a secondary dimension, you’ll have something like this:

The table above shows the dates/revenue after the Premium Subscription date (e.g. CV1 = 20130109, January 9, 2013) during which a transaction took place, along with the revenue associated with these transactions. Now, say you want to see how the January 9th cohorts did compared to the January 12th cohorts. , Simply create an advanced segment for each of the Cohort dates and the graph will look as follows:


Here are the findings:

  • The January 9th cohorts brought in 2.84% of the revenue
  • The January 12th cohorts brought in 1.94% of the revenue
  • There were no more sales from either cohorts after January 21st

Once the “Cohort date” data has been collected, you can also plot it in Excel or Tableau in a traditional cohort chart. We will cover this later in the post.

Before we move on to more analysis, insights and actions, we still need to complete the picture. Remember, our subscribers can upgrade to a Premium Subscription using the mobile app as well. So let’s take a look at the mobile data next.

Mobile Analytics Data – Flurry Analytics come to the rescue
This client is an early adopter of mobile and has leveraged the powerful capabilities of Flurry Analytics since day one. The client has done their homework and planned out their user actions (aka Events) and did not just settle to track screens and buttons, but is also tracking outcomes like In-App Purchases (IAPs).

Similar to what we gathered from Google Analytics, in Flurry Analytics we want the following information:

  • An event when the Premium Subscription purchase takes place, date stamped for our cohort analysis (e.g. 20130109)
  • Revenue associated with purchases
  • Unique user ID (and again no PII please!). The user ID will show in your Flurry Event Log (we will use this for the advanced user-centric tip at the end of the post)

flurry event

As you see in the image above, your “time stamped” mobile app eCommerce data is now available.

We can create a segment in Flurry under Manage –> Segments and choose the Premium Purchase Date as the custom event/parameter for the segment, and then export this data (download as CSV or leverage the Flurry API). If you are downloading the CSV, your parameters (revenue, etc.) will be available in one cell in Excel, and you can easily parse the parameters you want to include in the report.

Offline – Backend Data
Well, we are almost there! Since the client’s system allows for offline Premium Subscription upgrades as well, we need this offline data integrated too. These elements are required from your backend:

  • Premium Subscription date
  • Date of purchase
  • Revenue
  • Unique customer ID

And finally, we download the offline data in a CSV file, and we are ready to rock and roll!

Note: if you are forward thinking and like to get your hand into the latest and greatest, then Universal Analytics (from Google Analytics) will be an option here. With Universal Analytics you will be able to import the offline user interactions into Google Analytics and map it to existing data.

3- Integrated Reporting – Customer 360-View, Tableau Dashboarding & Insights

If you’re using a data warehouse or a database one approach would be to present all this data in Tableau using its optimized connectors available for many data environments. If you haven’t made the investment into your infrastructure yet, or you don’t have the IT resources to leverage these connectors in Tableau, you could extract your data into CSVs from Google Analytics Premium, Flurry and your backend system and upload it into Tableau.

Now that you have your data ready from the three data sources (web, mobile and offline), it’s time to tie it all together, visualize, analyze and find some awesome actionable insights. Your common element for this analysis is going to be the Premium Subscription Date.


Aha! The cross-platform full picture is ready for you. You see the value of those who signed up on the 9th of January (in terms of sales and revenue), and you start understanding where they came from, what offer resonated with them among other factors.

Advanced Tip# 1 – Google Analytics Tableau Integration
If your business is a pure online play, and your subscribers can only buy on the website, I have a very nice surprise for you (kept this till the end to reward the loyal readers :) ). With Tableau version 8, there is a native integration between Tableau and Google Analytics (in beta but works well), meaning you can pull your dimensions and metrics directly from GA and into Tableau. You bypass the need to download files, reformat and upload. I’ll share the tips on doing so in a post by our Tableau geek extraordinaire Shiraz in the next few days. (Please note if you have a large site with high volume and you are getting sampled reports, this method will not work for you. For now, you’ll have to download the unsampled reports and then upload into Tableau).

Advanced Tip# 2 – User Centric Analytics
In-line with our emphasis on concepts such as user-centric analysis (and Universal Analytics), and in capturing data from different sources, I emphasized the need to include a unique user ID when customers buy the Premium Subscription. Once you have this common user ID (aka key) you can run all sort of sophisticated analyses in your Tableau or a BI tool of your choice. Examples include:

  • Customer Lifetime Value (LTV)
  • Recency & frequency across platforms (web, mobile, offline)
  • Segment high ticket item customers
  • Map demographics data (available in Flurry and/or your backend system) with web data
  • Plus more and more scenarios that I’ll leave for future posts

So there you have it, how to create a cohort analysis across mobile, web and offline platforms. I hope you found this post useful and I look forward to hearing your comments and input!

Feb 26


What in the….? “Google Analytics” is a browser?!

(For the quick answer, skip to the “Conclusion” below…)
We ran into a confusing situation the other day with a client. They were getting traffic from a browser called “Google Analytics“. It threw us for a loop, because obviously, we’re familiar with Google Chrome, Firefox, the awesomeness of Internet Explorer (sarcasm), Opera, and Safari. Haven’t had the chance to use the Google Analytics browser though.

That’s cause there isn’t actually a Google Analytics browser as you might have guessed. So we were weirded out to see that in our reports. What does it mean when you see on your reports that a large number of visits are coming from “Google Analytics”?

Browser equals Google Analytics

Here’s what we figured out:

On a hunch, we decided to segment for mobile traffic only.
Google Analytics - Advanced Segment Mobile

The number of visits from the browser “Google Analytics” essentially didn’t change (the change was so minor, we could assume this was due to sampling). Thus, it looks like this traffic is pretty much mobile traffic.

Browser = Google Analytics - after mobile advanced segment

For this client in particular, they saw a spike of traffic that corresponded to this number. We asked if anything special happened around the time of the spike. They confirmed that they had recently launched a new version of their mobile app.

Need Help With Google Analytics? Click Here

Conclusion: The Answer Is…

After some more digging and testing, we concluded that when the browser says “Google Analytics”, it’s mobile app traffic! Apps using either the iOS or Android SDK for Google Analytics will report their usage under the browser “Google Analytics.” Not sure when Google will change that, but hopefully, that will help anyone trying to solve that mystery explain to their executives or clients where this traffic is coming from. One method of preventing this from happening entirely is to report mobile app traffic to an entirely different account than your web traffic.

Feb 13

Marry Google Analytics and Voice of Customer“I…i…i…i… I’m soooo in love with you… whatever you waaaant to dooo… it’s alright with meeeee…!” – lyrics from the “Reverend of Soul” himself – Al Green.

It’s Valentine’s Day, love is in the air! So, we thought we’d shoot cupid’s arrow through this next post about data and measurement. Give it some “love”, so to speak.

If you’re heavy into Google Analytics, you might be “infatuated” with quantitative data - “views”, “visits”, “bounces”, etc. – and…ahem…”RAW” numbers. That’s only one “better half” though. That is, it’ll tell you “what” is happening on your site. However, what about the other half – “why”??? One way to answer that question is by “marrying” your quantitative analytics data with qualitative, “voice-of-customer” survey data.

“Why… why do people break up? Then turn around and make up… I just can’t seeeeeeee….”

Let’s take a statistic of love. Unfortunately, it looks like approximately 50% of marriages fail :(   [Cupid, what the hec are you shooting at?] The Reverend of Soul, Al Green stated in his immortal lyrics, that he could see “what” was happening, but for the life of him, just couldn’t see “why” people break up!!!

This Valentine’s Day, let’s pretend to try to solve this urgent matter of the heart. WHY is this happening?! Is the aforementioned statistic enough for you to take action? Doesn’t look like it.

Those of you who’ve been through a tough time in your relationship know it’s not black and white – there could be a plethora of reasons:

  • You might have grown apart
  • Stresses like money could have strained the relationship
  • Infidelity
  • Communication could be missing
  • Etc.

As any counselor would tell you, they wouldn’t just diagnose “what” is going on in a couple’s troubled relationship, then right away present solutions. In order to solve the issues, they would go into several deep counseling sessions so they can get to the nitty gritty of “why” this is happening.

Similarly, diagnosing “what” the issues are with your website is only part of the process. Voice-Of-Customer qualitative surveys could potentially give you insight into “why”, insight that may not be readily apparent from quantitative data.

“Let’s Stay Together” – Putting Voice-Of-Customer Together with Quantitative Data

You have several ad campaigns leading visitors to a landing page asking them to register for your event. As the example figure below indicates, your analytics may be telling you that your landing page is getting all the traffic you’re paying for, from Facebook/Google Adwords campaigns, etc. You’re spending good money, and your micro conversion, say an agenda for your event, is getting all the downloads you want. But according to your data, the bounce rate is really high and no one is registering for your conference. Why?!

campaign metrics googel analytics

We recently ran into a similar situation with a project our consultants were assessing, and tried to figure out the issue:

  • Was there not enough sales information? We listed all the bullet points…
  • Did clients not trust the organization/service? We listed impressive certifications and testimonials…
  • Was the registration call-to-action button not clear enough? It’s big and yellow, and it looks like people are clicking on it according to our data…Just not completing registration

At this point, even if you had all the unsampled data in the world from Google Analytics Premium, you’d still be stuck. There’s always an option of implementing A/B testing, but that would take time and coding. Plus, what do you start testing? You can make educated guesses, but you’re pretty much taking shots in the dark (like cupid).

So even before we explored A/B testing, we tried implementing Qualaroo – an awesome, quick, non-invasive survey that quietly pops up based on criteria you set (ex. 30 seconds on the page, after 2 pages have been visited, etc.). You can ask any question – open-ended, multiple-choice, etc. And after each survey is complete, get reports in the native interface or download reports into .xls for deeper crunching.

So we asked – “Second thoughts about registering? Why?”

qualaroo survey voice of customer

Sample responses were as follows:
qualaroo survey results
As you can see from the responses (after ensuring we had enough replies for an acceptable confidence level/ statistical significance…), it was pretty much unanimous – pricing was the issue!

It would have been extremely hard to conclude that with just the quantitative data. We may have stumbled upon this after some A/B testing, however, after marrying the qualitative, it pointed us towards a pretty reliable direction in a quick and efficient manner. Even if we wanted to verify using A/B testing, we now have an idea of where to start.
They lowered the pricing and registrations started rolling in!


So when you’re having dinner by candlelight tonight, gazing into your lover’s eyes… slowly move in close, and whisper ever so gently… “My quantitative is incomplete without your qualitative…” <3

Feb 12

Allaedin Ezzedin Top 5 Percent LinkedInOkay.  I have to say if I were wearing my ego (bragging) hat, and if no one in our office had a higher number of LinkedIn profile views than mine (ahem… Feras Alhlou), I might be more excited about the latest brilliant LinkedIn marketing email blast. Recently, they sent a blast about their 200 million members milestone.

While I appreciate the fact that my LinkedIn friends made the effort to update me about the state of their network in 2012, the message I got today about my profile, Allaedin Ezzedin, being “one of the top 5% most viewed LinkedIn profiles for 2012” is more misleading than informing.

Here’s why.

As an analyst…

As an analyst scrutinizing the data, the first question that came to my mind was, “I wonder how many of these profile viewers were…

  • …random profile stalkers?”
  • …job recruiters?”
  • …peers from the Analytics community (that is you if you are reading my post now)?”
  • …prospects who are considering hiring my firm; E-Nor?”
  • …existing clients?”
  • …blog readers?”
  • …Jasmines searching for Aladdin?” :)

target audience is worth more than profile viewers

The number LinkedIn provided doesn’t explain any of this!  In the analytics world, we call this metric “page views”, which we give an extremely low value in understanding user behavior and engagement. It doesn’t tell you “who” is viewing your page or “why” they are viewing it, which is the real actionable insight you need.

Segmentation is always essential. Each one of us has different social networking goals, objectives, and interests. Someone using social media for branding has a different target audience than someone who is searching for a job or someone who is using social media to advocate their ideologies or methodologies.

Need Help With Google Analytics? Click Here

As a marketer…

Now, as a marketer, my first reaction to the email campaign was, “Wow, no one is viewing profiles on LinkedIn!” If my profile, which is getting an average of X views per month, made it to the top 5%, then how many views are the bottom 95% profiles getting? Also, if most LinkedIn users are not socially active, then what is real value of LinkedIn as a marketing tool? What does 200 Million users mean to me? How will the new numbers impact my social marketing strategy? Shall I continue to invest on my paid campaigns in LinkedIn? Are my potential prospects on LinkedIn? Are they active? How can I increase engagement with my personal or corporate profiles? What metrics should I track in LinkedIn?

If there is one thing we learn from the latest LinkedIn email campaign is not to run after bold numbers and to have a clear objective for every marketing channel we invest in. Don’t be fooled by the marketing numbers because most of the time they are tweaked/formatted to serve marketing interests, not yours.


1) Disclaimer: I love LinkedIn, as it is by far my number one social network of choice when it comes to connecting to my professional circles (ex-classmates, ex-workers, the analytics community, volunteering community, partners, clients, vendors, etc.).  My critique here is solely limited to their latest marketing email blast.

2) For my friends who didn’t make it to the top 1%, 5%, or 10% profiles, I would say don’t sweat it. Your profile’s success is about how far you are from reaching 100% of your target audiences, not just any audience! :)

3) Let’s all hope that next year, the annual update from LinkedIn looks something like this…

linkedin 2013 suggested email campaign

Feb 04

When Google Analytics makes changes to their platform, it’s usually a good thing, and helps users navigate through the dashboard easier, offers a more powerful data experience and overall is seen as beneficial to the user.

Google Analytics did just that this time around, as they rolled out some improved features earlier this month. We at E-Nor dove right into the changes and have been loving them.

There was a blog post on the Google Analytics blog as well as a great one by, but a lot of our clients are still unaware of the spiffy new changes, so we wanted to pass the information along!

Check it out:

Improved Navigation

The first change you notice when logged into Google Analytics is there are less tabs at the top of the screen. Just the simple Reporting, Customization, Admin and Help tabs are now available here. This area had been freshened up a bit, and some of the reports here have been moved to the left sidebar. We’ve noticed this definitely makes more sense, and offers up a more unified placement for the tools. Plus, the top orange navigation bar floats as the user scrolls down the page. Pretty cool! ;)

Create Your Own Custom Dashboard

We all have our own preferences, right? So too is the case when working with Google Analytics, and the gurus at Google have figured that out! The platform now offers new enhanced personalized dashboards. Users can choose from a variety of layouts, giving them a handful of options available. This feature expands the way the dashboard is laid out, and how users can see their data.

google analytics layout options

Dashboard Additions

1. Advanced Segments
Anyone who works regularly in Google Analytic is aware of the Advanced Segments feature. Well, thanks to this most recent round of changes, this tool has been added to the dashboard. (Yay!) Users can find this button in the upper left section of the dashboard near the Audience Overview header.

google analytics advanced segments

2. New Widgets
The Geo Maps and Bars widgets are both new additions to the Google Analytics dashboards. You will also find the Geo Maps have been added for custom reports. These are available on the improved dashboard, and provide yet another set of data gathering tools that help users. The Geo Maps widget allows users to color code data by country, state and so on. The Bars widget give more advanced graphic data abilities.

Examples of Widgets

google analytics geo map
google analytics bar chart

new widgets google analytics

So if you haven’t already noticed the improvements to the Google Analytics interface, now is a good time to jump in and check them out.

We at E-Nor have found these new and/or revised features and tools highly beneficial, and love the fact that they allow us to continue to provide top-notch analytic services to our clients. We hope you enjoy the changes too!

Feb 01

49er-mobileSuperbowl Sunday! GO NINERS!!!! Is mobile phone traffic on your site going to be the same during the game? During commercials? During halftime? In this post, we’ll show you how to figure that out!

Mobile is taking over! The ability to access the Internet from everywhere is so convenient – when you’re at the supermarket, gas station, even when you’re driving (don’t do that). Sometimes, I’m too lazy even open my laptop at home, it’s simpler to just pull out my mobile phone in front of my TV and connect.

And mobile use is only growing:

  • In 2013, more people use mobile than PC’s (Gartner 2011)
  • 50% of U.S. cell phone users have smartphones
  • 47% of consumers look up local information (for example, stores they want to visit)
  • 46% look up prices on a store’s mobile site
  • Etc.

(Statistics from

Marketers who don’t start accounting for this trend will surely be left behind. The design and structure of your site, how your visitors use it, how visitors buy, etc. – is completely different on mobile devices vs. desktop – even vs. tablets. A mobile visitor is on the go, the screen is significantly smaller than a PC and tablet, it’s touch screen, etc. As a marketer, you need to be able to do dive deep and figure out exactly what’s going on to get insights unique to each.

Ideally, the following insights is what you want to see. You can see in this case that mobile phone and tablet behavior is different – tablet visitors are much more engaged, spend more time on the site, and view an average of one more page than phones:

Mobile Traffic Only Report - Google Analytics

“Mobile” (including Tablet)

It’s tricky. Google Analytics lumps tablets into the “mobile category”. But what if your design is responsive, and you have a different design for your mobile phone site vs. your tablet site? What if that’s impacting your traffic differently? You’ll need to separate the data.

Advanced Segment: Mobile Phones Only

Google already has a default segment to analyze ONLY “Tablet” traffic. But where’s the “Mobile Phone only” default segment? We love Google, but hint hint, cmon guys…

Have no fear – our resident analyst genius, @charlesfarina created a simple advanced segment to do this. Here’s how:

Step 1. Create a New Custom Advanced Segment
At the dashboard, choose Advanced Segment and click on “New Custom Segment”
Create advanced segment in google analytics

Step 2. Name Your Custom Segment
Name it something useful like “Phones”

Step 3. Include Mobile Traffic
Get all of mobile traffic, including Tablets and Mobile by selecting “Include” > “Mobile (including Tablet)” > containing “Yes”.

Step 4. Get rid of Tablet traffic
In order to separate mobile phones from tablets – you want [Mobile (which is mobile + tablets)] – [Tablet]
Add an “and” statement and “Exclude” > “Tablet” > containing “Yes”. This will get rid of tablet traffic.

Your advanced segment should look like this:
Mobile Phones Only Advanced Segment - Google Analytics

You are ready to slice and dice only mobile phone traffic!
Just in case, here’s a quick link for the segment (so you can just save it to your profile automatically).

Jan 25
Google Analytics is a powerful tool.

There are so many screens, features, tools, filters, searches, etc. For the heavy data cruncher, it’s always nice to have a set of shortcuts for quick execution of common tasks. In case you couldn’t find it on the Google Analytics Blog, we laid them out for you.

Here are a list of keyboard shortcuts in Google Analytics:

Google Analytics Keyboard Shortcuts

Date Range Shortcuts

d t Set date range to TODAY
d y Set date range to YESTERDAY
d w Set date range to LAST WEEK
d m Set date range to LAST MONTH
d c Toggle date comparison mode (to the previous period of whatever you are looking at.
Example, if you’re looking at 6 days, this will compare it to the 6 days before it)
d x Toggle date comparison mode (to the previous year of the period you are looking at)

Application Shortcuts

? Open keyboard shortcut help
h Search help center
a open account panel
shift + a Go to account list
s / Search reports
shift + d Go to the default dashboard of the current profile

Need Help With Google Analytics? Click Here
Some of them not working? If you’re a genius like me, it’s probably because you pressed one of the keys on accident and it took you to the search box. Make sure you are out of the search box when you try these.

Jan 21

Do You See What I See?
As an analytics consultant, it’s important – strike that – mission critical, to make sure you understand how your customer uses data. Sometimes we have a tendency to assume our clients look at and interpret data the same way we do. Nothing could be further from the truth. All customers look at data differently, and uncover insights that we might consider surprising or unexpected. We owe it to our customers to ask the right questions and best understand not only what’s being measured and reported on, but how that data is being consumed and interpreted.

reCAPTCHA = evil
recapcha failA recent experience highlighted the importance of understanding how a customer looks at data.

I received a request from a customer asking me to QA some Google Analytics tracking code that had been deployed to a page. In order to test this scenario, I needed to go to the page, fill out a form and submit. If all went well in my testing, I would see a hit sent to Google Analytics via my HTTP header monitoring tool.

Seems straightforward and easy enough right? Sure… except that I wasn’t able to submit the form. Huh??? Yeah you read that right – I wasn’t able to complete and submit the form. Not once, not twice, but 30 times I tried and failed! What in the darkest depths of Middle Earth (Yes, I loved The Hobbit, and watched it recently, as you might guess), would cause this issue, you ask? The culprit was that most dreaded of all online phenomena – the reCAPTCHA!

First, let me state my complete and utter disdain for reCAPTCHA. I hate it, and consider it amongst the darkest of evils on this planet. Why do I hate reCAPTCHA so much? That’s easy :) I certainly understand the benefits gained by reCAPTCHA. It blocks spam form submissions and only allows legitimate human submissions. Yadda yadda yadda :) I’ve heard it all. Still hate it! Some of those alphanumeric combinations are so out of this world that no human could ever read them.

recapcha fail reCAPTCHA is a marketer’s worst nightmare. Web forms are sort of like Cookie Monster “More cookies…more cookies!”. All they want is for users to fill them out. Cookie Monster is cute, and simply wants more cookies. Sure, he leaves a mess of crumbs behind, but I’ve never heard him complain about the quality of the cookies he consumes.

recapcha fail reCAPTCHA takes an innocent, unassuming, cute and cuddly puppet like Cookie Monster and turns him into…the Soup Nazi! For those of you who didn’t catch the reference, or were deprived of the wonders of Seinfeld, check out this link to learn more. The Soup Nazi makes you stand in line (quietly), and only gives you a very measured amount of soup. Don’t you dare look at him the wrong way, or “No Soup for you” will echo in your ear drums. The mere presence of the Soup Nazi strikes fear into the hearts of the most brave of people, and renders them the likeness of jello – jittery and paranoid. reCAPTCHA has the same impact on web forms by making conversion significantly more complicated, and frustrating the user beyond compare.

reCAPTCHA sits at the forefront of the classic battle between IT and Marketing. Both sides have valid arguments as to it’s usefulness, and as much as I dislike it, reCAPTCHA obviously does filter out spam. At what cost though? Are legitimate customers jumping ship out of sheer frustration? Let’s see what the data has to say…

Data’s Turn to Talk
When I was unable to complete the form, I started thinking how this must be affecting other users. Surely, this must be giving other users the same headache it gave me. Luckily we were tracking all sorts of details about the form, so the answer was in the data. Time to dive in :)

In this type of analysis, it would be good to look at:

  • Percentage of users running into reCAPTCHA errors
  • How many users are receiving multiple reCAPTCHA errors?

Here’s what I found: (all data referenced below is for a period of one month)

# of times the form was viewed: 2,174,325
# of times the form was submitted: 241,803
Unique # of times a recaptcha error occurred: 184,318
# of times multiple recaptcha’s were encountered: 270,969

What stands out from looking at the data above?
1. There were more reCAPTCHA errors (270,969) than submissions (241,803)!!! That means the problem is very widespread, and means that a majority of people are running into this problem. In fact many are hitting multiple reCAPTCHA errors.
2. 76% of form submissions resulted in reCAPTCHA errors!!! This was nothing short of shocking for me. 76%??? That’s insanely high! I happily wrote an email to my customer, full of excitement at what I’d found. I received a prompt reply thanking me for the analysis. The marketing team was also flabbergasted by these results and understandably wanted the reCAPTCHA removed.

Wait…A Different Perspective?
A few days later I received a call from an IT manager in the same organization. First he asked to verify how I had come up with this data. Upon confirming its validity, he also thanked me and said this data was immensely useful for them. He then went on to point out that their reCAPTCHA error rates are inline with industry standards and he wanted to make this a monthly report that he could trend with the intent of making sure the rate isn’t going down. Not going down??? What? You see he was in IT, and his focus was to make sure the reCAPTCHA was doing what it’s designed to do. The harder it is to read those ridiculous patterns, the more effectively it must be working. I thought they would be motivated to remove the reCAPTCHA, and instead they want to make it harder!!!

Always understand how your client will use data. It’s of critical importance and helps you provide meaningful insight, and ensure that data is actually being used to impact the business.

Thoughts, comments? I’d love to hear your perspective!

Dec 28

Update 1/23/2012 – As the commenters have pointed out you will need to delete the analytics.dat file from the program download in order to authenticate successfully.  Will update blog post/files soon.

Update 1/24/2012 – Link to github -–Python-


As the year closes, I wanted to make you aware of a feature in Google Analytics that has amazing potential.  This will tie in with all of your New Year resolutions around making better marketing decisions.   In October at the Google Analytics Certified Partner Summit we attended, Google announced the new cost data upload feature.  For the first time we can measure the return on investment for all of our paid advertising campaigns.  You can segment and dive into this data to find what aspects of each paid traffic source or the overall campaign are under or over performing and optimize accordingly.  This feature allows you to upload non-AdWords cost data for Bing, Facebook, LinkedIn and any of your other paid traffic sources.  What we are trying to create is the report below:

You can now make decisions on your marketing campaigns based on ROI inside of Google Analytics!  It is very apparent in the above report that we need to reevaluate or do optimization for the campaign, since we are making less than what we are spending with an ROI of -1.43%.  This is the type of insight I want to help you find.

Use This Program or an API?

The first thing you should do is check to see whether the traffic source you want to upload cost data from has an API. Many popular services like  Facebook and Bing have an API that you can leverage to save you a significant amount of time and effort.  E-Nor has leveraged this with many of our clients like OEMPCWorld.  If your source has an API, you can use a tool like ShufflePoint In2GA and work with them to establish the connection.  This connection will upload all of your historical cost data as well as upload your new cost data each day.  This means you don’t have to do anything on your end outside of creating/paying for an account with ShufflePoint.

It is very likely that you will also have sources that don’t have an API.  This could be affiliate networks, banner ads, e-mail lists, and the like.  You could also decide you don’t want to pay someone to establish an API connection, so the following instructions could apply to Facebook and Bing as well.  The instructions below will allow you to upload cost data from any .csv files on a Mac to Google Analytics.  These files can be created in Microsoft Excel, Google Docs, iWork, or any of your favorite spreadsheet tools.  I tried to make these as detailed and easy to understand as I could, so that anyone could use it.  The best part about this tool is that it is free and will be a great way to end the year or kick off the new year with new insights.

Disclaimer:  These instructions are for Mac only.  If you are familiar with programming languages, getting it to work on Windows shouldn’t be difficult.  I can also not be responsible if anything breaks or your computer explodes, this is a free tool after all =)

Here are the six steps:

1. Create Project and Obtain OAuth 2.0 Client ID

a. Go to Google API’s page and click the blue “Create project…” button

 b.  Slide “Analytics API” to On

c. Agree to the terms of service
d. Click on “API Access” in the left navigation


e. Click the blue “Create an OAuth 2.0 client ID…” button

f. Fill in “Product name” field with “Cost Data Upload Program”.  Leave rest of form blank and click grey “Next” button.

g. For Application Type select “Installed application” radio button.  Leave installed application type defaulted to “Other.”  Click grey “Create client ID” button.

  h. Find and click the “Download JSON” link in the Client ID for installed applications section on the right side of page

2. Download and Install Cost Data Program

a. Place the downloaded “clients_secrets.json” file from step 1h on your desktop.
b. Download the “e-nor cost data upload program” to your desktop.  Select file then download to download entire zip, instead of each file individually.
c. Unzip the download and make sure it is on your desktop.
d. Drag the “clients_secrets.json” file (step 1h and 2a) from your desktop into the unzipped “e-nor_costdataupload” folder (step 10-11) on your desktop.  Select “Replace” when prompted.
e. Login to Google Analytics and go to the profile you want to upload data to

f. Click “Admin” Button in top-righthand corner
g. Record the provided Property ID: UA-XXXXX-XX
h. Click “Custom Definitions” Tab
i. Click “New Custom Data Sources (Beta)” Button
j. Enter a name, description and select profiles to send data to and click “create”

k. Record the UID


l. Open the “e-nor_costdataupload” folder and open the “” file

m. Scroll to the bottom and replace the accountId, webPropertyId and CustomDataSourceID with what you recorded in steps 2g and 2k.  The accountId is the same as the webPropertyId, just with the beginning and end removed.  If you recorded UA-123456-1 in steps 2g.  You will put UA-123456-1 as the webPropertyId and 123456 as the accountId.

n. Save and close the files

3. Installing Google Analytics API files through Easy Install

a. Download the Easy Install files here and place file on desktop
b. Click spotlight in top right hand corner of your Mac and type in “Terminal” and open it.

c. In Terminal type or paste: cd Desktop

d. Type or paste: sudo sh setuptools-0.6c11-py2.7.egg

e. Enter your password and you should see the below message if successful

f. In terminal type or paste: sudo easy_install –upgrade google-api-python-client

g. If successful you will see this message

4. Creating Cost Data Files

There are 17 different dimensions and metrics you can upload, but only 5 are required: cost, impressions, clicks, source and medium.  In the e-nor_costdataupload/October-Cost-Data folder you can find examples of the fields and formatting that is required or view the image below.

You will need to create a file per day for your cost data.  If you want to upload the month of November you will have one file per day of data.  You can create this using your favorite spreadsheet software (Excel, iWork, Google Docs, etc).  The only requirement is they have to be saved as “Comma Separated Values” which are the .csv file extension.  There are templates/examples in the cost data program provided.

5. Uploading Cost Data

You will now be uploading a cost data file.  You need to first create and save the file as explained in step 4.  We will now upload one day of data using the following command: ./ {yyyy-mm-dd} ~{cost-file-location}

If we are trying to upload the example files included with the program we would enter the following in terminal:

a. Open Terminal and type or paste: cd Desktop/e-nor_costdataupload
b. ./ 2012-11-01 ~/Desktop/e-nor_costdataupload/October-Cost-Data/2012-11-01.csv

6. Viewing Cost Data in Google Analytics

This is the hardest part of all!  You will need to be patient and wait for up to 12 hours to see the cost data appear in Google Analytics.  In my experience on average it took about 30 minutes.  You will find it in the Traffic Sources Section as “Cost Analysis”.  You can also create custom reports to add in metrics such as revenue, transaction and to apply various advanced segments or filters.

Need Help With Google Analytics? Click Here

Closing Thoughts

I love this new feature, but as you can tell if you aren’t leveraging an API tool it can be quite a bit of work to get the data into Google Analytics.  Two improvements I have planned are first to have a way to batch upload buckets of files at one time and also for a script in Excel to create multiple files from a master spreadsheet based on date.  Ideally, I would like to be able to create one spreadsheet with all of the dates/data I need and have the program parse and upload it from memory.  The source code for the program is given to you, so I encourage one of you to do it and let me test it =)

Related Posts:

  1. Google Cost Data Announcement
  2. Cost Data Development Goodies
  3. E-Nor / OEMPCWorld Use Case
  4. E-Nor Cost Data Announcement