Create a Google Analytics implementation doc using R

[This article was first published on Stats raving mad » R, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.


One of most frustrating things that you can face during a day is someone to ask you to give her the definitions of the funnel that someone configured a couple of years ago for 10 product categories. The only thing that can help you in this is to have an Implementation document. Oh, you bet you have it updated! (side-note: with the spread of Google Drive people tend to create documents for fun, however they rarely get to update them or even revisit them).

God, you definitely need to have written for each account you handle

  1. What you are measuring for
  2. What the configuration for the views are
  3. What is that special thing you have opted in for this view?

How we gonna do this fast?

After you run this code, you’ll have an Excel file that has all the following configuration point noted down

  1. Accounts
  2. Web Properties
  4. Goals
  5. Segments
  6. Custom Data Sources

The utility that enables us to get 1-5 is the Google Analytics Management API.

The Management API (v3) exposes multiple Google Analytics configuration entities. Most are organized hierarchically:

  • Each user’s Account can have one or more Web Properties, each of which can have one or more Views (Profiles) or Custom Data Sources.
  • Each user’s Account can also have one or more Filters.
  • Views (Profiles) can have one or more Goals, Unsampled Reports or Experiments, and Custom Data Sources can have one or more Uploads. Unsampled Reports are available for Google Analytics Premium customers only.
  • There are also entities that allow users to manage user permissions by creating a User Link at the Account, Web Property, or View (Profile) level.
  • Similarly, AdWords Links can be constructed at the Web Property level.
  • Filters can be connected to a particular View (Profile) with a Profile Filter Link. A user can also define Segments, which are not hierarchically related to any of the other entities. This diagram represents the parent-child relationships among the entities:

Google Analytics Management API

Limitations of the Management API

We cannot get some interesting stuff with the API

  1. Channel groupings
  2. Brand keywords
  3. Enhanced eCommerce settings
  4. Events
  5. Custom dimensions/metrics

to name a few.

Another R Google Analytics API package…

Let’s go on and load a new package that enables us to connect to the Google Analytics various APIs, RGA, all capitals since there is another package with lowercase rga.

I will not go through the details on how to get authenticated on Google Console as it is well explained in the package repository.

The main functions we will need to create our implementation doc are

  1. get_accounts()
  2. get_profiles()
  3. get_segments()
  4. get_webproperties()
  5. get_custom_sources()
  6. get_goals()

Off to the real thing!

In the lines below we load the package and pass the and client.secret to get authorised in the API. Once we done this we can get the views (or profiles as we used to call them few months ago)

# Initiate RGA ------------------------------------------------------------
# load package

# Connect to API ----------------------------------------------------------
# get access token

# Get the accountIDs ------------------------------------------------------
# get the unique accountIDs
accounts > head(accounts)
[1] "183263" "431217" "4476693" "7369781" "22603640" "27175492"

Now, that we have the unique accountIds we can iterate over them (say in a for() loop) and get the elements needed.

We will create an excel file for each account we have since this is the standard file businesses are using. We could also change this to work on a website (URL) level but let’s make the assumption that we have things organised under one accountIds for each client. I will wrap it with an introductory sheet that you can edit to show more of your documentation style. It will be as plain as

This is a technical report of your Google Analytics Implementation
Updated at : D/M/YYYY h:mm

The code to produce the result of this process is below

# Start getting the information! ------------------------------------------

# Add an intro sheet

for (i in 1:length(accounts))
tmp_goals tmp_profiles tmp_webproperties tmp_segments tmp_custom_sources # Let's write the results to xls files
write.xls(c(tmp_webproperties,tmp_profiles,tmp_goals,tmp_segments,tmp_custom_sources), file=paste0("ga_doc_",websiteUrl[i],".xls"))

Is that all to document for?

Well if you have everything near your hands you can do something for yourself as well. Like getting all data together and trying to find if you have anything mis-configured (or”forget”). You can do this easily by creating a master table of all accountIds as in the following lines

# We will use gdata to read the excel files
# Now we can merge them to run some
# diagnostics. Taken from

# this list the files and filters Excel files

for (file in file_list)
# if the merged dataset doesn't exist, create it
if (!exists("dataset"))
dataset }
# if the merged dataset does exist, append to it
if (exists("dataset"))
temp_dataset <-read.xls(file, sheet = "Sheet2")
dataset<-rbind(dataset, temp_dataset)

Now, with a simple plot we can find things that are going wrong (eg missing e-commerce setting OFF where it should be ON).

The end is the beginning

This covers only the basics of a Google Analytics Documentation. There are a lot of stuff you might consider adding to this document like :

  1. The actual business questions that the implementation is aiming at
  2. KPIs you are trying to optimize for
  3. Definitions of the technical piece of tracking (passing value definitions, when an event is triggered etc)
  4. A sheet with Change history. You can get this information from the Google Analytics Admin console.

To leave a comment for the author, please follow the link and comment on their blog: Stats raving mad » R. offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)