How to Set Up Google Analytics Account for Your Clients in 2020
Is there a better time to start a new habit than the beginning of the year? If we’re being honest, probably, but it’s definitely less catchy. Whenever it is that you’re reading this, now is the perfect time to get organized and start setting up your clients’ Google Analytics accounts the right way.
Make your life easier by following these simple guidelines. They work for us, improve our workflow and eliminate most of the common mistakes.
The Client Owns the Account
When you develop a website for your client, the client owns it. Why would a Google Analytics account be any different? It can seem easier to just go ahead and create the account, then let the client access it, but this is never a good idea, for a few reasons.
First, being the account owner also makes you the owner of all the data in it. That’s a headache you definitely don’t want to have. Second, even though you never want to plan for it, contracts do end, sometimes on not so great terms. If that happens, transferring account ownership will likely be one of the last things you want to think about.
Instead of you creating the account, make sure it is the client who does it. If they are not tech-savvy enough to do it, help them, guide them through it, do whatever it takes.
Always Have a Master Google Analytics User Account
Now that the client has created the account they are asking you which email you want to use to access it. firstname.lastname@example.org sounds like a good answer here, doesn’t it? She is the one who will be doing all the work after all. While this isn’t a terrible way to do it, it’s far from optimal.
What happens if Mrs. Analytics Superstar is on vacation and the client urgently needs to see a report? There are two options – either her vacation gets interrupted or you need to ask the client to grant access to another one of your employees. An unhappy superstar employee or a client who thinks you’re unprofessional. Tough choice.
What you should do instead is have a “master account” that will be the only one the client will ever give access to. All this account needs to be able to do is “Manage Users”. When that happens the person controlling it can add “real accounts”, ones that have a first and a last name behind it, always giving them exactly the access level they need, never higher.
Sidenote: make sure no one on your team uses a @gmail.com account to access the client’s Google Analytics. If they ever leave the company, all you need to do is kill their @yourcompany.com account and that removes their Google Analytics access as well (and Google Ads, and Google Tag Manager…). With a @gmail.com account, you need to make a list of all the Google Analytics properties they have access to and remove them from each one of them. One by one.
Always Create at Least Three Google Analytics Views
If you do website development this might sound obvious but it’s shocking how often you’ll find a Google Analytics property with only one view, the default one.
Here are some of the issues with having only one view:
- What happens if the data in that one view gets corrupted?
- How do you test new filters?
- How do you edit existing goal configurations?
You don’t get either one of these correct on the first try and the data in Google Analytics is wrong. This is why you should never make changes to your default view without testing them somewhere else first.
Our preferred way is having at least three views configured for each Google Analytics property we work with:
The way we work with them is very similar to how we work with our clients’ websites codebases. You never want to deploy the changes to a live website without testing them in staging first. And if this goes really bad, you always want to have a backup.
Master view is similar to your live site, Test is your staging site and Unfiltered is the backup.
Never make any changes to Master view without testing them in, well, Test. Only when you’re 100% certain everything works as intended should you make it live. The thing with Google Analytics data is that once it’s in, it’s final. Segments can be applied to reports, but filters are applied before data is recorded. What this means is any config error you make is final. Always use Test view first.
If something does corrupt the data in Master, you have Unfiltered to fall back to. This view should never have any custom config. Reporting from Unfiltered will require some extra work, but at least there is some raw data to work with.
Use Google Tag Manager to Add Google Analytics to the Site
To keep it simple here, it is almost never about having a reason to use Google Tag Management (or a similar tag management solution), it is about finding a reason not to use it.
Going through all the benefits of Google Tag Manager would require a separate post, so I’ll stick to the one full-service web agencies might find the most useful: You will not need a developer’s assistance to deploy tags. Adding most tags and setting triggers for when those tags fire will take no more than a few minutes in most cases. There is even a preview mode to make sure everything works as intended before you make it live.
If the only tag you deploy to your website is Google Analytics, Google Tag Manager might be overkill, an unnecessary layer of complexity, but who are we kidding, that never happens. Even if you launch with just the Google Analytics tag, you will add more tags later, so just make sure you get started the right way.
There you have it, four easy to follow tips that will save you time, eliminate most of the problems you could face and make it so much easier to figure what went wrong when it does. Do you already follow any of them? What are your best practices?