In nature, animals migrate. Elephants, birds, turtles â by some combination of memory, instinct and intuition, creatures in the wild know to make the same epic voyage every year.
But how did they make the first great migration?
If anyone figures that out, let Google know. The company is struggling to get marketers on board with a major analytics migration set for next year.
Google sent marketers into a tailspin earlier this year with the announcement that Universal Analytics, the ubiquitous baseline Google Analytics product used by practically any business with a website or search budget, will be entirely phased out by July 2023, a move that force-shifts their entire client base to the newer Google Analytics (GA4).
âSucker punch to the gutâ
The marketing industry as a whole experienced a similar collective freak-out in 2020, when Google gave a two-year heads-up on the end of third-party cookies. Unlike the Chrome third-party cookie deadline, however, which has been delayed, Google really is going to get rid of Universal Analytics and default all marketers to GA4 by July next year.
And, also unlike the third-party cookie phaseout, marketers are largely unaware of how seriously these analytics changes will affect their business and are far behind where they need to be in the transition.
âWhen I’m talking to clients, there’s a lot of confusion,â said Chris Tatum, head of SEO at the performance agency Within. âTheyâre asking, âWhy is this happening?ââ
Kirk Williams, founder and CEO of Zato, a paid search consultancy that works with small businesses, said he would describe the initial news of UAâs demise as âa sucker punch to the gut.â
Even so, GA4 is a far more workable solution than meets the eye, Williams said.
Itâs such a difficult transition because, unlike other Google Analytics upgrades when features change or the UI is somewhat altered, the switch to GA4 is a complete structural overhaul. There are no apples-to-apples comparisons between the Universal Analytics and GA4 reporting data sets.
âThere are a lot of companies and a lot of agencies that are not talking about this change, which is way more negative than any of the changes that actually come from GA4,â Tatum said.
Sessions to events
At the heart of the transition to GA4 is a shift from âsession-basedâ to âevent-basedâ reporting.
Sessions center on an individual, as in when a person shows up to a site organically or via an ad. A session is made up of their interactions and activity with and on the site. If they return later in the day or from a different ad campaign, thatâs a new session. Taken together over time, sessions add up to something akin to a user-level customer journey.
An events-based lens shifts the focus to particular results and types of customers prioritized by a business. Events arenât pinned to individuals by third-party cookies, ad IDs or mobile device IDs.
For example, a publisher may want to analyze first-time visitors who viewed at least half of a video on a site with the goal of, say, understanding whether those types of users are more likely to become return visitors compared to paid search. A publisher could also analyze what triggers people who visit their site to scroll down multiple pages and spend time reading rather than bouncing.
Change is hard â but necessary
Google frames the GA4 migration as a necessary upgrade for every marketer and business. And it is easy to make the case against Universal Analytics.
For one, UA only incorporates site and search information. The product that became GA4 was initially Googleâs âWeb and Appâ analytics â a version for marketers whose apps or mobile audiences were of particular importance for their business.
âFor brands running direct mail, CTV, OTT, linear TV or podcasting campaigns, none of that would show up in Universal Analytics,â said Ron Jacobson, founder and CEO of the attribution company Rockerbox. A multimillion-dollar Roku ad campaign might drive a surge of site visitors, he said, but in UA theyâd show up as organic or paid search, since people often click the ad atop search results even if they searched specifically for that business.
GA4 plugs directly into cloud-based data warehouses and incorporates those disparate channels into analytics reports.
GA4 also has a Google Ads integration, at long last. It may be hard to believe, but Universal Analytics and Google Ads arenât the same system and often disagree on key metrics, like how many people a particular ad campaign trafficked to a site. Googleâs paid search team can over-attribute itself, and if a brandâs or publisherâs UA reporting doesnât agree, Google Ads still gets paid based on its own accounting.
âThatâs been a sore spot for companies, and GA4 does strive to fix that and create consistency between those platforms,â said Withinâs Tatum.
But despite certain improvements, the market clearly prefers UA. After all, GA4 has existed for a decade and hasnât earned a large share of the market because it doesnât work that well, not because it missed it a fair chance.
Purpose-built for privacy
Which is where web privacy comes into it.
Why is Google removing all historical data from UA, rather than continuing to house the data in its servers? Customers must download their raw UA data within six months after the July changeover and then can sift through the data at their own discretion, because Google wonât accumulate or store it anymore.
Some small business would be willing to pay a monthly fee for Google to continue storing that data and making it available in GA4, Williams said. But thatâs been a nonstarter.
âPutting on my foil hat for a second, Iâd say theyâre going to wash their hands of all of that data and move forward without worrying about legislation,â Williams said.
Worrying about legislation is not a theoretical issue for Google.
Last week, Italy became the latest EU nation whose data protection regulator issued a notice to all site operators banning the use of Google Analytics products. The same is true in Germany, France, Luxembourg and Austria. The rationale for these cases is that no American company may collect server-level data on European citizens in Europe, because that data is inherently subject to US surveillance and can be subpoenaed by the FBI even if the data never leaves European servers.
When Google announced the expiration date on Universal Analytics earlier this year, it said it would also cease logging or storing IP address data â a crucial datapoint that qualifies as personally identifiable and thereby subjects Google Analytics to tough interpretations of GDPR.
Removing IP addresses may not be enough for Google Analytics to stem the tide of GDPR suits. But privacy concerns do explain why Google will force a change across its customer base, rather than continue to offer multiple services.
Google is deliberately rejecting the old system as a privacy risk, not on grounds of added efficacy.
âThe marketing community is willing to make the concession that these are things that make sense â we get what Google is doing,â Williams said. âBut Google needs to make the concession that some things really are a problem and not just force people to a product that is not ready.â
The bad and the ugly
Marketers may understand that Google must change its analytics to something like GA4, rather than the primarily cookie-based Universal Analytics that had tunnel vision on websites and search.
But itâd be nice if the product actually worked.
Enterprise marketers often use Adobe Analytics or are already on GA360, Googleâs enterprise analytics product â the one-time Adometry attribution solution â and so arenât much affected by the move to GA4 â though GA360 customers will be folded into GA4 as well.
But small businesses are among the heaviest users of Universal Analytics, and for SMBs, the supposed cross-device selling points of GA4 arenât persuasive. Local pizza shops, electricians and barbers, et al., usually have a site and a Google account but donât have their own apps and arenât fussed about the missing podcast attribution in UA.
Another important ingredient that isnât widely available with GA4 is granular location data, said Carrie Shaw, CMO of a CRM provider called Copper that specializes in small businesses that use the Google suite. GA4 doesnât rule out location data, but IP addresses, which will disappear, are the mechanism that underpins most location-based segmentation.
Still, once SMBs make the migration and GA4 is running at full capacity, the new platform is âactually going to be a big winâ for small businesses, Tatum said.
But that doesnât mean SMBs are happy about losing control.
GA4 and the new Google Ads integration make it easy â inescapable, really â for marketers to hand more control over campaigns to Googleâs black-box system. Some savvy marketers and tech practitioners deeply begrudge the new Google Analytics, because it consumes a brandâs valuable first-party data and returns no data itself, other than aggregate campaign reporting. (We did great!)
âYou are going to have to give Google the benefit of the doubt that they’re modeling your data accurately based on what any of us is allowed to look at in GA4,â said Shaw, who just oversaw Copperâs own transition from Universal Analytics to a GA4 account.
Giving Google the benefit of the doubt, however, is ⌠not a given.
âThe very reason Google Analytics exists in the first place and is made free is so that Google can exploit this data commercially to sell ads at higher value,â said Thomas Petit, a mobile growth marketing consultant. âGoogle is subsidizing a free analytics tool to increase the worth of their ads.â
Petit said that, whenever possible, he chooses to use a non-Google ad product, particularly for analytics.
Tatumâs point is that SMBs like mom-and-pop pizza chains and barber shops donât have sophisticated growth marketing experts at the helm, and so they can do just fine by handing over their complete campaign apparatus to Google.
But retailers and ecommerce sellers do have savvy ad tech and data experts analyzing their media, and theyâre struggling with the changes and loss of data that comes along with the migration to GA4.
With Universal Analytics, an ecommerce or retail owner could pull up average online site visitor conversion rates, lifetime customer value and average session values. But all those data points â which are based on accumulated sessions and not events â disappear in GA4. The same ecommerce marketers are also big users of UTM parameters (the code tagged to the end of a site URL that indicates if traffic is coming from a certain publisher, newsletter, social platform or affiliate partner), and they, too, are not tracked in GA4.
And the toughest hit: A factory reset on historical sales data.
Sellers that are just now beginning a GA4 migration will have an instance of GA4 and of Universal Analytics running simultaneously. By the time the training wheels are removed in July 2023, theyâll have just one yearâs worth of look-back data.
Retailers donât gauge growth based on whether sales grow or fall between June and September, say. Itâs about whether sales have grown or fallen compared to the same June-to-September period last year and the year before that.
Having only one year of historical data â and data thatâs coming from a transitional test year to boot â is a major issue for commerce companies.
The third-party opportunity
Historically, Google Analytics made it practically impossible for another baseline analytics service to take root â the notable exception being Adobe Analytics, which has a services-based approach and focuses on large global companies.
Google Analytics is ubiquitous, free and boasts a veritable army of top-end engineers. What startup could win? Which VCs would invest? None and none.
But vendors are pivoting into analytics services specifically because GA4 and the whole Google Analytics machinery is a mess right now.
Amplitude began running a B2B marketing campaign this year explicitly targeting potentially ticked-off Google Analytics customers who may be open to a wholesale vendor shake-up rather than adopting GA4.
Edgemesh, an ecommerce site and server operator that took its first steps into advertising and analytics tech this year, likewise said customers are asking for GA4 solutions. If commerce metrics like average session value and lifetime customer value are disappearing with GA4, an outside vendor that can recreate those metrics can easily attract test budgets, Jacob Loveless, Edgemeshâs co-founder and CEO, told AdExchanger.
Commerce-specific tech vendors are being called in as well, Rockerboxâs Jacobson said. Many Shopify merchants see widening disparities between the new GA4 reporting and the analytics they get through Shopify, he said. For CDPs or attribution companies like Rockerbox, thereâs a chance to help retailers or merchants to rationalize their new Google reporting data compared to what theyâre used to.
âI fully expect there to now be a small market for third-party SaaS data and visualization software that charges a low monthly fee,â Zatoâs Williams said.
What to do?
For a marketer or business owner thinking about next steps with Google Analytics, the advice from vendors and tech intermediaries is consistent: Begin yesterday.
âI’m telling every client that as soon as you can get a GA4 integration on your site, do so,â Tatum said. âYou want to start gathering that information as quickly as possible, both for historical references and the machine-learning pieces that are being built into GA4.â
Marketers will have to learn how to build and use a pivot table in Excel and an entirely new platform interface, with none of the familiar Google Analytics click-and-learn site reporting, Copperâs Shaw said. But itâs going to be much more difficult for businesses that make the effort come mid-2023, she said, when the familiar tools are yanked away.
Although vendors will make their pitch for data migration services, they might not be able to woo marketers, who have largely resisted Google Analytics updates in the past and will get over this one, too, Tatum said.
âI donât anticipate this really drives people away from GA4 as the mainstay solution for analytics,â he said.
But even right now, with less than a year to go, it feels like Google Analytics engineers and support staff are too removed from the broader marketing ecosystem, according to Williams.
Cooler heads need to prevail, he said, because each side has valuable feedback thatâs getting lost in the messy middle ground.
âSometimes what happens in the Google Ads landscape is advertisers complain about everything around a product update,â he said. âAnd then Google struggles to hear the important things because all theyâre getting is an overwhelming amount of complaints.â
