By now, you may already know that GA4 operates across platforms, uses an event-based data model to deliver user-centric measurement, and does not rely exclusively on cookies.
And you recognize that GA4 uses machine learning to generate sophisticated predictive insights about user behavior and conversions, create new audiences of users likely to purchase or churn, and automatically surface critical insights to improve your marketing.
Heck, you may have already started to move to GA4 as soon as possible to build the necessary historical data before Universal Analytics (UA) stops processing new hits on July 1, 2023, and UA 360 stops new hit processing on Oct. 1, 2023.
Many people may mistakenly think they have a good bead on things.
Well, I was at Pubcon Las Vegas on Nov. 14, 2005, when Google announced that Urchin Software, which it had acquired in April of that year, was being renamed Google Analytics. Yep, I was in the room where it happened.
I was standing next to one of my clients, John Marshall, the CEO of ClickTracks Analytics, which offered a range of competing solutions that cost $495, $1,195, or $3,495.
That’s when we both heard that the basic version of Google Analytics was free for the first time.
So, I know a little something about the impact of new versions of Google’s web analytics service.
And, I’ve learned that you don’t need to wait for machine learning to generate sophisticated predictive insights about a couple of “events” that the adoption of GA4 is likely to trigger for your organization or clients in the next 14 months.
One is a reorganization. The other is an agency review.
The Reorg
The “web analytics” team still sits in the IT department in far too many organizations.
Why?
Because the team was originally created back in 1995 when web analytics meant servers, log files, and complex handwritten code to parse the log files and pump out reports.
So, putting them in the IT department made perfect sense back then.
But, data collection, storage, and processing have all moved into the cloud (hosted by your application service provider rather than in-house).
This eliminated the need to maintain IT teams for web analytics, except perhaps to update measurement codes and related code fragments collectively known as “tags” on your website or mobile app.
In addition, your website itself has transformed from being “brochure-ware” back in the early days into an increasingly integral part of your business – both online and offline.
Nothing highlights this change more than the fact that we no longer count the number of client requests (or hits) made to the web server like they did a generation ago.
Because of these trends, the “digital analytics” team doesn’t belong in IT anymore.
Where does it belong?
Well, ask yourself three questions:
Who uses analytics?
Marketing (not IT) needs to see unified customer journeys across their websites and apps.
Marketing (not IT) needs to use Google’s machine learning technology to the surface and predict new insights.
And marketing (not IT) needs to keep up with evolving customer needs and expectations.
Who directs implementation?
Marketing (not IT) needs to decide which recommended events to add, which suggested audiences to use, and which events to mark as conversions.
Marketing (not IT) needs to decide what associate monetary values to use for micro-conversions, custom insights to create, and anomalies to act on.
And marketing (not IT) should decide which other platforms, such as Google Ads, Search Console, and Salesforce Marketing Cloud, to integrate with GA4.
Who owns reporting?
Marketing (not IT) needs to drive sales or app installs, generate leads, or connect online and offline customer engagement.
So, marketing (not IT) needs to use data-driven attribution to analyze the full impact of their latest campaigns and ongoing programs across the customer journey.
And marketing (not IT) needs to export that analysis to Google Ads and the Google Marketing Platform’s media tools to optimize those campaigns and programs.
This is why digital analytics belongs in marketing – and it has belonged there for more than 10 years.
But, inertia is a powerful force – and most people hate reorgs – which explains why far too many organizations are loath to move their analytics team out of IT and into marketing.
So, why do I think that GA4 will be the irresistible force to overcome this immovable object?
Well, one of the features that you’ve already heard about is Analytics Intelligence, which uses machine learning and conditions that you need to configure to help you understand and act on your GA4 data.
And one of the statistical techniques that Analytics Intelligence uses is Anomaly detection.
Using historical data, Analytics Intelligence “learns” to predict the value of metrics for the current time period and flags any data points as anomalies if their actual value falls outside a “credible” interval.
For detection of weekly anomalies, the training period for GA4’s machine learning is 32 weeks.
For detection of daily anomalies, the training period is 90 days. And for the detection of hourly anomalies, the training period is two weeks.
In other words, somewhere between 2 and 32 weeks after GA4 is set up and starts collecting data, Analytics Intelligence’s machine learning will be sufficiently trained to analyze your data and predict future actions that your end-users may take.
That’s when marketers will begin seeing “Insights” appear on their GA4 Home page.
These Insights will show unusual changes, emerging trends, and other anomalies about your site or app.
Seeing specific Insights can help you quickly identify data changes that warrant further analysis and action.
That’s when the marketing department will start “freaking out” if the IT department doesn’t respond to urgent requests for “help” within a week, a day, or even an hour.
And that’s when the business case for moving the analytics team from IT to Marketing will suddenly become data-driven.
Why is this scenario likely to ripple across organizations worldwide over the next 14 months?
Well, early adopters of GA4 have already reported the benefits of getting a complete view of their customer lifecycle with an event-based measurement model that isn’t fragmented by platform or organized into independent sessions.
And I’d argue that the same benefits are available to an organization that isn’t fragmented by department or organized into independent silos.
For example, Gymshark, a fitness apparel and accessories brand based in the UK, used GA4 to understand its customers across touchpoints on its website and app.
This enabled the Gymshark team to see how users moved through the purchase funnel. As a result, they reduced their user drop off by 9%, increased their product page clickthroughs by 5%, and reduced their time spent on user journey analysis by 30%.
Oh, and non-profits can benefit from seeing the user journey from end to end, too.
For example, 412 Food Rescue, a non-profit organization based in Pittsburgh, needed to recruit more volunteers to deliver food from retailers to people experiencing food insecurity.
Automated Insights in GA4 showed their team that weekends tended to be a little bit slower in terms of volunteers and engagement, so they adjusted the social media campaigns that were driving traffic to their website.
And they’ve cut their reporting time by 50%, which has freed up their already limited staff to grow their impact throughout the community and expand to new cities.
Watch “Google Analytics: 412 Food Rescue Case Study”, which was uploaded to YouTube on Mar. 24, 2021, to hear the team tell their story in their own words.
[embedded content]
This brings us to the second “event” that GA4 is likely to trigger for your organization or clients: An agency review.
Agency Review
Now, some big ad agencies were using Google Analytics with DoubleClick’s advertising services, which Google Acquired in March 2008, even before the Google Marketing Platform was launched on July 24, 2018.
So, they should weather the storm created by the move to GA4 without too much difficulty.
But, many other ad agencies will need to hold an “all hands on deck” meeting to figure out how to hang on to a client that’s just configured their GA4 property and started recording YouTube Web Engaged View Conversion (EVC) events.
To do that, the client:
- Linked their property to Google Ads to make YouTube Web EVCs available in their GA4 reports.
- Activated Google signals to see conversions from users who are signed in to their Google accounts.
Now, they expect their agency to help them do what Harmoney did.
Who is Harmoney?
They’re an online personal loan platform based in New Zealand.
What did they do? They used YouTube to build brand awareness of its target audience in Australia.
How does Harmoney know that they did that?
Well, they used GA4 to measure EVCs after their target audience watched their YouTube ads.
This enabled them to directly correlate the uplift in brand impressions to their investment by measuring the engaged-view conversions from their YouTube ads, which often occur in mobile apps.
Or, what if a client asks your agency for new ads that target one of their “Predictive audiences.”
For example, let’s say your client has built an audience of “likely 7-day purchasers,” which includes users likely to purchase in the next seven days.
Now, they assume that your agency can help them do what McDonald’s Hong Kong did.
Umm, what was that?
Well, McDonald’s Hong Kong met its goal of growing mobile orders using a predictive audience of “likely” 7-day purchasers.” They exported it to Google Ads – and increased their app orders more than six times.
They also saw a 2.3 times stronger ROI, a 5.6 times increase in revenue, and a 63% reduction in cost per action.
Or, another client may want your agency to create a remarketing campaign to re-engage users based on their behavior on their site or their app.
What will your agency do when it’s handed a remarketing list of “Suggested audiences,” which can include:
- Achievers (e.g., users reach key milestones like reading a certain number of articles).
- Billable users.
- Cart abandoners.
- Checkout starters.
- Item searchers.
- Item viewers.
- Leads.
- Registered users.
- Searchers.
- Streamers.
- Top players.
- Top scorers.
- Tutorial abandoners.
- Tutorial finishers.
- Video completers.
- Video starters.
- Wishlist users.
Hey, you can’t make this stuff up.
So, what will you do?
Well, my scientific wild-ass guess is your agency will act like a swan, gracefully gliding across a lake – while furiously paddling beneath the water’s surface.
But, if you don’t convince everyone at your agency that GA4 will fundamentally change client expectations of what ad agencies should be able to do, then you’re likely to lose those clients.
I don’t suppose you know what clients will expect your ad agency should be able to do, do you?
Aw, wait. That was on Final Jeopardy! last night.
Mayim Bialik said…clients now expect their ad agencies to be able to use the front end of the Google Marketing Platform to leverage what the back end of the platform (the part formerly known as Google Analytics) can provide…which now includes measuring YouTube Web EVC events, generating Predictive audiences, and creating Suggested audiences.
So, don’t be surprised when your client announces an agency review.
And even if your agency is invited to compete, don’t expect to hang on to this account – unless you’ve figured out how to defeat some of the big ad agencies using the Google Marketing Platform since March 2008.
So, for the inevitable agency review that will be triggered by GA4, I’d recommend that you organize your presentation to address the five best practices that DoubleClick once called “programmatic advertising” and Google now calls “the latest advances in machine learning for data-driven creative.”
In case you haven’t learned these five best practices yet, they are:
- Organize audience insights: Aggregate your data sources – including GA4 data, offline data, CRM data, survey data, or third-party data – to get a comprehensive view of your audience.
- Design compelling creative: Google encourages marketers to “establish a general campaign plan and align your creative, analytics, and media teams as early as possible. This allows the creative team to tailor messages appropriately for different channels and devices; it will also make it easier to ensure creative assets can work across them.”
- Execute with integrated technology: Identify a capable partner for programmatic buying. For example, you’ll find 2,424 potential partners in the Google Partners Directory.
- Reach audiences across screens: According to eMarketer, U.S. advertisers are expected to spend $62.96 billion on programmatic digital video in 2022, up from $52.17 billion in 2021. And mobile represents two-thirds of programmatic video advertising, but its share is declining as connected TV (CTV) earns more ad dollars.
- Measure the impact: Use GA4 to measure EVCs after your target audience watches your YouTube ads, grow mobile orders using a Predictive audience of “likely 7-day purchasers,” and lastly, remarket to “Suggested audiences.”
Now, some of the big ad agencies have more experience executing with integrated technology and reaching audiences across screens.
That’s why you may need to identify a capable partner for programmatic buying before the agency review.
But, even the big ad agencies are still learning about GA4 just like you are.
So, I’d argue that you should be able to hold your ground when organizing audience insights and measuring the impact.
And, the one area where you may have an advantage over even some of the big ad agencies is designing compelling creative.
So, here’s what you need to emphasize at the beginning of the agency review: Creative accounts for 56% of advertising effectiveness, and media 30%, according to Nielsen Catalina.
Okay, how do you design compelling creative for programmatic digital video?
First, I recommend that you watch “Understanding the ABCD guidelines for effective YouTube ads.”
It explains that successful ads on YouTube grab Attention, incorporate strong Branding, build a Connection, and have a strong Direction.
[embedded content]
A second approach uses emotional AI to correlate creative attributes with video performance data.
How do you do that?
Well, read my article, “What’s the Alternative to Spending $7 Million on a Super Bowl Ad?”
The first digital marketing expert to respond to my request for alternatives was Ian Forrester, the founder and CEO of DAIVID. He used his video testing tool, which uses Emotional AI to automatically predict video performance without the need to show creative to respondents.
A third option is to use YouTube Director Mix to create customized videos at scale, swapping out different elements to tailor content to specific audiences.
For example, Mondelez India designed “The Not Just a Cadbury Ad,” employing YouTube Pin Code Targeting, YouTube Director’s Mix, and Google Maps API.
This enabled them to produce thousands of customized AI-generated ads to 270 pin codes across eight cities.
This hyper-localized campaign helped nearly 1,800 local retailers grab business during Diwali during the pandemic.
It delivered incredible business results, including over 32% more business growth against what was forecasted and 2x sales for the retailers featured in the ads.
[embedded content]
The fourth way is to create a video experiment to determine which of your video ads is more effective on YouTube.
With a video experiment, you can test different video ads with the same audience and then use the experiment results to determine which ad resonates more with your audience.
For example, Grammarly used Video Experiments to test ad sequences.
To see their results, watch “Grammarly | Success Story | YouTube Advertisers.”
[embedded content]
And now for something completely different.
Instead of letting GA4 prompt an agency review, preemptively urge your clients to conduct a digital analytics review.
Matt Bailey, who teaches people how to turn marketing data into action, says:
“I’ve been talking with Adobe, and they’ve seen an incredible surge in inquiries and changeovers. With the privacy issues and Google being the world’s biggest data vacuum, I’ve decided it’s time to make a change as well. I’m loving that the analytics landscape is once again becoming a financially competitive environment!”
He adds, “I’ve been testing Matomo, Woopra, Heap, and Piwik Pro. They all have similar features as G4. The problem is that G4 still isn’t finished. They keep adding measurements and changing labels. Just two weeks ago, they added a new measurement that trashed any historical data associated with it.”
So, which of these options should you use?
Well, before David went to fight Goliath, he stopped by a brook to select five smooth stones.
And, all David needed to slay Goliath was one smooth stone.
So, here’s what you should emphasize at the end of the agency review: If an agency uses integrated technology to reach audiences across screens with creative that isn’t compelling, then the only thing you will measure is the lack of impact.
My colleagues at Search Engine Journal have already done a great job preparing you to be successful with Google Analytics 4 (GA4). Check out these resources if you haven’t yet:
Featured Image: ra2 studio/Shutterstock