Google Analytics Health Check

First of all! We should read these 3 books!

Business have 10x more chance of making it once weu read these.

• Startup Owners Manual by Steve Blank
• Lean Startup by Eric Ries
• Lean Analytics by Alistair Croll

Ring Model

The Ring Model is a quick and effective way to identify where our biggest leaks are.

It’s similar to a funnel as it helps us see the key loss steps. The main point is to help us see the big picture involved.

The Ring Model helps us see where our flow is stuck—where the traffic is not flowing down to the next level. The main benefit here is that it helps us see which layer of our website needs the most help.

We use a standard Google Analytics report for constructing it: Behavior -> Site Content -> All pages.

What we could do with this is two-fold:

1. Map out traffic flow per layer of the site (and see where the flow is stuck)
2. Verify whether the goal funnel has been configured properly

This flow is actually the manual construction of a funnel. By doing this manually instead of using funnel reports in GA, we’re making sure the data is correct.

Especially helpful when:

• Goals are unreliable / broken / have no data.
• Flows are mixed in funnels (mid-stage joiners).
• The data we need does not exist.
• Start with a manual walkthrough of the site to map out the URL structure. If the URL structure is not specific about the type of the page (e.g. */product/*, */category/*, etc), make sure to start tracking virtual pageviews for the same types of pages. If we run an eCommerce site, layers might look something like this:
• Home
• Category + Search
• Product
• Cart
• Checkout step 1
• Checkout step 2
• Checkout step 3
• Checkout completed
• Count the unique pageviews per layer. Go to Behavior -> Site Content -> All pages report in Google Analytics and type the URL identifier of each layer, i.e. “/products/”, or “/cart/”, or “/checkout/step 2/” or whatever they may be.
• Check the numbers against what we see in Conversions -> Goals -> Funnel Visualization. If we see discrepancies, the funnel has most likely been set up incorrectly.

If we had 300k product page views, 5k add to carts, and 1k checkouts, where would our problem be?

Cart adds!

If we had 300k product page views, 100k adds and 1k checkouts, we know our problem is in a different place.

Google Analytics Health Check

Before we start working on a new Google Analytics setup, we need to make sure that everything that needs to be measured is being measured (goals, funnels, ecommerce & event tracking setup), and the data is not corrupted.

Health checks are a series of analytics and instrumentation checks that answer the following questions:

• “Does it collect what we need?”
• “Can we trust this data?”
• “Where are the holes?”
• “Is there anything that can be fixed?”
• “Is anything broken?”
• “What reports should be avoided?”

Nearly all analytics configurations are broken

90% of setups have insufficient tracking or broken configurations.

Health checks are serious. Running with bad data renders all of our work useless.

Who does the health checks?

While this is one of those things that we could outsource to an analytics expert who is familiar with the client’s goals, we as optimizers need to be able to check most of these things ourselves. 

We don’t need to be the person who sets up event tracking or ecommerce tracking (developers can and often should do that), but we do need to know how to find out what’s working and what’s missing.

What’s involved in a health check?

1. Confirm that we have full access to analytics for all the sites we’re optimizing
2. Do a walkthrough with a consultant or project lead: Get clarification on how the current setup has been created
3. Agree on areas of focus (based on our and the client’s goals)

Items to review:

• Profile setup (configuration and admin).
• Filtering of traffic (agency, office, data cleaning).
• Goal and funnel configuration (key reports).
• Code review (onpage analytics code).
• Bolt-ons : Inpage, Outbound, Scroll reach, Viewport, Other.
• Page and process instrumentation (funnels, steps, forms).
• Any issues that would prevent insight!

A health check typically takes less than a day but can take longer depending on the number of sites and the complexity of them.

Output & Benefits

A health check should result in a short summary report, showing:

• Priority items to be fixed or instrumented (e.g. add event tracking for cart adds).
• Reports or metrics that may be skewed (like bounce rate, new visits etc).
• Key things for the team to be aware of when reviewing (e.g. funnel has splits).
• Quick wins for additional insight.
• Specific instructions for developers to fix (e.g. configure 100% site speed tracking).
• Priority score and benefits for each change or fix.
• Explanation of why each change is required (some fixes may be needed for later work).

The summary report should equip the client with solid and useful recommendations, actionable changes to improve analytics reporting, and an understanding of where things are broken.

Checklist

• Account settings
• Adwords setup – ensure this is correctly configured, PPC data shown in GA.
• Filtering – filter out office IP, agencies, other 3rd parties, us.
• Property settings
• Default URL set up correctly.
• Enhanced link attribution turned on.
• Webmaster tools linked (vital for SEO, PPC).
• Subdomains and TLDs (top-level domain) set up.
• Enable demographics and interest reports (doubleclick .js code).
• Custom definitions set up if applicable (and using universal analytics).
• View settings
• Are the views set up optimally?
• Do we have PLAY (testing) and RAW profiles for all setups?
• Are we doing country filtering or lumping (LATAM) together?
• Is there a mismatch between views and intended use?
• Default page and timezone configured.
• Ecommerce tracking turned on.
• Site search tracking turned on.
• Goal configurations – checked thoroughly.
• Are filters correct for this profile?
• Basic advanced segments set up and shared.
• Common issues
• Cross domain tracking – incorrectly configured.
• More than one outcome using the same page – common!
• Missing tracking code – on some pages – check.
• Double tracking of page or event – causing 0% bounce rate.
• Events not being set as interactive/non-interactive, skewing bounce rate horribly (compare with flow reports).
• Meta refresh or redirection causing params to be lost.
• Campaign tracking wrongly set up.
• Missing tracking in emails, newsletters, RSS (Really Simple Syndication), social.
• GA tracking code not all on one version.
• Funnel or goal steps too broad a match.

Resources

1. Healthcheck-Template
2. Healthchecks Questions
3. GA site audit