Analytics Director, Alex Davies, shares his experience of integrating Google Analytics and Salesforce as efficiently as possible via this case study.
Jellyfish manage online paid media for the online education of a campus based school in South Florida.
Our chief KPI from our client partner is to deliver leads (potential prospects).
We acquire leads online through a website we host for the school, the sites chief function is to convert site users to leads.
We use a request for information form on the site to achieve this. Once a user submits that form, they are a lead.
We were therefore basing our online media strategy on the affiliate partners, channels, keywords, etc. that demonstrated the highest conversion rates to LEADS.
Leads are a good KPI, but of course not every lead is a good lead. Only a percentage of those who convert online to leads ultimately convert offline by becoming students and therefore generate actual revenue for the client.
So we thought, wouldn't it be nice if we could allocate our paid media cost for our partner more effectively toward the channels, affiliates and keywords that demonstrated the highest conversion rates to students?
As it stood, our client partner had sole visibility into offline prospect behaviour.
Their Marketing department and trained Enrolment team qualified, nurtured and converted a percentage of the leads we delivered to actual students. Great, but which ones? Which channels, campaigns and keywords were effectively delivering students?
These are the areas we should allocate our marketing spend toward as they're more likely to convert down-funnel, where it is most important to our partner's business goals. While lead qualification and student statuses were being tracked and recorded within the University's CRM, there was no way to seamlessly connect this offline information with online data to do a comprehensive analysis.
We wanted to attribute real live sales conversions (in this case student enrollment) to their respective online origin (channel, campaign, keyword, etc.), all in one place. We wanted to be able to be more deliberate with paid media for our partner to deliver optimal results. We wanted to optimise for students rather than leads.
Measurement Protocol – what is it?
In mid-2014, Google Analytics released an upgraded version of itself: Universal Analytics. Amongst other improvements to the code are more ways to collect and integrate different types of data using the Measurement Protocol. Simply, Measurement Protocol is a way to send data from any source or device (offline or online) back into GA.
We correlated data to a cookie ID (CID) value. The CID is a unique value in the Universal Analytics cookie that is assigned to each person that enters the website (or any other platforms). Cookies for a single user persist for up to 2 years. As a CID is stored server-side, it holds all of the acquisition and user behaviour data. This makes it perfect to accurately tally with other systems.
For this University, when a user submitting their information online and became a lead, all of their relevant information is automatically sent into the University's lead bin. This is a system that qualifies a lead and determines whether the criteria required to become a student is met. Once qualifying, the information is sent on to Salesforce (CRM-offline source), where the prospect entry is stored.
How did we use the Measurement Protocol to integrate?
To give you an overview of the system hierarchy - upon a user submitting a lead, all of the relevant information is sent to a lead bin. This is a system that qualifies a lead and determines whether the criteria is met for a potential student. Upon qualifying, the information is sent on to Salesforce and a prospect entry is stored. Any subsequent changes to student profile and status – for example: a candidate successfully completes the interview process – are made in Salesforce. While all of the user behaviour data sat outside of the CRM in Google Analytics Premium.
In order to send the data into the lead bin, we required a unique key that ties the user journey to an entry in the lead bin / Salesforce. As mentioned, this is a CID or UID value. For this university we used CID.
Capturing the CID
We used Google Tag Manager to set a custom jQuery script to dynamically collect the user's CID value and pass this alongside the corresponding user's values into Lead Bin every time a user submits a request for information form. It was a complex script that read the values from the _ga cookie prior to sending the user's personal information. In order to add an additional level of security, this was encrypted using base64 upon submission and decoded within Salesforce.
The University's lead bin was then configured to accept incoming datasets from a CID value and send this directly into Salesforce as soon as the lead had been qualified. Salesforce was also amended to accept an additional column of data (CID) from lead bin.
When a prospective student is qualified in different steps of the funnel (for this University: incomplete app, complete app and student) over the phone, the Enrollment Advisor annotates this within the prospects entry in Salesforce. This was our cue. As soon as the trigger takes place, we create a string that dynamically populates an event similar to the following:
? v=1 // Version number
&tid=UA-xxxxxxxx-y // Tracking ID / Property ID / School
&cid=XXX // Google Analytics Client ID
&t=event // Event hit type
&ec=salesforce // Event Category
&ea=mm_protocol // Event Action
&el=student // Event Label to populate based on the user's status
Because Google Analytics Premium contains all of the user behavior and acquisition data, we could then marry the new user status against the original data set for analysis in Google Analytics Premium and BigQuery.
We're now able to analyse and report on student statuses and utmost conversion that directly correlate to our online media efforts. We can also analyse down to keyword level which of our media efforts are most effective for this University.
A Goal was configured within Google Analytics Premium to easily correlate with paid search traffic to gauge Conversion Rates by keyword.
What does this mean?
We can now focus all of our marketing and optimisation efforts towards students rather than unqualified leads.
In addition, as we leverage the integration between Google Analytics Premium and Doubleclick, we can feed a student audience list into DCM for prospecting similar audiences across the Doubleclick ad exchange.
To find out how Jellyfish can assist with your Google Analytics and CRM integration or anything else analytics challenge click here.