Universal Analytics - GACP Summit Day 1 Announcements


We’re GACP and proud!

We here at High Position have been GACP partners for the last 4 years. Once again, two intrepid members of  our team are braving the wonderous sunrise view of the Golden Gate Bridge, and struggling through the legendary West Coast hospitality, to attend the GACP Summit in Montain View California where there have been a whole reel of new product announcements and great feature improvements!

Here is the report from day 1:

“We started the day with a warm welcome from the Google Analytics support teams at The Computer History Museum, Mountain View. We were promised big announcements and it was only minutes before we had the biggest of the day.

Universal Analytics

The biggest issue facing marketers on the web today is that of tracking users across multiple platforms and multiple session visits, web analytics tools have for too long been stuck in the mindset of tracking visitor sessions to websites, today Google announced they are turning this on it’s head and completely restructuring their architecture and the way GA runs by turning it into a User-centric system, where tracking a user becomes the most important part of the system.

“The Biggest Change to Google Analytics since the original Urchin Model” - Paul Muret

How are they doing this? The GA cookie and javascript tags haven’t changed for a long time, but soon they will be replaced by a far simpler tracking code that identifies users by a unique visitor key, this could also be a unique identifier from your CRM system.

What this enables is for Google to build up a server side profile of your users and persist the state of their interactions, wheras in the current system the state is held within the users web browser, and can therefore be wiped out by a reset of your cookies or by simply switching to a different device / browser.

When they switch to using a mobile device, you send the same unique identifier and google add the new interactions in to the pile of the previous interactions building up a large pool of information about your customers. Google have architected their system to take this new user centred approach.

the full path to conversion

This also opened the door to a whole flood of new interactions including those offline such as a change in state your internal to your CRM (maybe a car salesman closes a deal offline), you can now track this against your users within GA and see the full path to conversion.

This all comes with a new API and libraries for many many platforms and SDK’s. Read the full Google Blog post. There is also a great summary post on the Analytics blog


Dimension Widening

The next big announcement was that of Dimension Widening, to keep it brief, this allows you to upload your own CRM data in CSV format to Google Analytics, merge your information with the information google have on your users, then segment your users using this information.

You can then create first-class custom dimensions and metrics to create custom reports based on anything you have uploaded. Examples included Gender and demographic information merged in with your GA data to track ecommerce sales across different audiences.

Avinash on measurement

The icon of Web Analytics has to be Avinash Kaushik, he came on stage to a massive applause and gave a great talk on many of the subjects he covers on his blog, including how to measure social media (the right way) how to deal with companies obsessed with capturing information rather than analysing information and how analytics is like a woman, “deep”.

“Great Analysts are Great Marketers” - Avinash

Also addressed were the key analytical tasks of Data capture, Data Reporting and Data Analysis and the amount of resources allocated to each of these. He criticised that in most instances Data Capture and Data Reporting receive the larger part of our attention when more investment should be made for Data Analysis, the bit that tells you what to do!

Avinash suggested the following allocation of analytic resources be made: 15% Data Capture, 20% Data Reporting, 65% Data Analysis

Attribution Modelling comes to standard GA

For the last year or so the fantastic Attribution modelling tool has only been available to GA Premium members, and today we were in for a treat when the Google team announced they would be making the tool available to all GA members, the release is currently in a limited release to developers only.

This is a great tool for measuring the marketing impact across different channels and working out the complex calculation of where to attribute ROI to each channel. The team also announced some automatic data-driven models to help with the manually created models currently available within the tool.

Another new feature being added to attribution modelling is the ability to upload your own cost information to Google so you can work out the real ROI of your system, not just the ROI in the Google related services such as Adwords. This will give real power and speed to doing calculations that we’d otherwise have to do in spreadsheets.

Channel groupings within Attribution modelling allow you to create higher level groups of traffic sources to get a view of “all paid search” or “all offline” grouped together.

More granular control over Data sampling

For some time Google have been sampling the data that they report on when you get close to more than 1 million visits within a time frame. Todays announcement shows us that we can now control how the data is sampled, so you can attribute your 1m rows to different Goals or e-commerce to ensure you always have the most accurate data you can.

Campaign Management

For quite some time google have had the URL builder tool that allows you to build a series of tags to append to your urls to make it possible to record information within Google Analytics, today Google announced that they will be merging the URL builder into a new Campaign Management function within the GA interface.

This lets you do the same functionality but now instead of having 5 or 6 parameters on the end of your url you get a single ID that you append instead, which not only looks much tidier but also allows you to rename campaigns and addresses issues with formatting. It looks to be similar to how the GCLID works when you use Adwords auto tagging.

Unified Segments

The last big announcement of the day was that of Unified Segments. These are an improvement to the current Advanced Segment model. Rather than creating an advanced segment of a few rules, you will now be able to use all the new functinoality of Universal Analytics, such as the total number of times a user has visited your site over the lifetime of their account, along with the session level features like time on site or campaign medium.

The new tool also allows us to create a segment based off of a series of actions a user takes, and lets us enforce that these actions must have happened in a certain order, similar to how the goal funnels currently work. By creating these more comprehensive segments we can create better models to track our users over longer periods of time and larger numbers of interactions.


At the end of the day, the whole purpose of taking a user centric approach to web analytics is to build up a profile of your users and their behaviour over long periods of time, taking this deeper approach to analytics enables us to create very advanced models of how customer interactions work in the real world.

This is where remarketing come in, we will soon be able to create remarketing lists based on Unified Segments directly from GA visits and within seconds users could be seeing remarketing ads on other sites customised to what they were liooking at on your site, all with the aim of having that user return to your site and convert.


In this new world of tracking all your user interactions online and offline within GA the obvious privacy questions started to arise. Google took their standard approach of “do no evil” and told us that we have to obey the laws in our own lands and they would do everything they could to prevent any issues. It was stressed that no Personally Identifiable Information should ever be uploaded to Google, including content contained in the new CSV upload feature.

All in all the first day of the summit had some exciting releases and we’re really happy that we’ve been able to share them with you. We have another full day tomorrow of geeky-ness but unfortunately everything tomorrow is confidential so we wont be able to talk about it publicly. If you have any questions that you would like to ask the Google Team, or ourselves, send a tweet using the hashtag #askGACP with your question and we’ll do our best to answer them.”


Leave a Reply

Your email address will not be published. Required fields are marked *