Apple's Safari Breaks Conversion & User Tracking
Apple Safari Breaks Conversion and User Tracking by removing query parameters from URLs.

Apple's Safari Breaks Conversion & User Tracking

TLDR:?Apple's Safari is removing query parameters (gclid, fbclid, ...) from URLs in iOS 17, iPadOS 17, and macOS Sonoma, breaking conversion and user tracking for various platforms like Google Ads, FB Ads, HubSpot , Adobe Marketo , Drip , Intuit Mailchimp , Twitter , and more).

Advanced Tracking and Fingerprinting Protection options: Off, Private Browsing (default), or All Browsing.

Advanced Tracking and Fingerprint Protection in iOS 17

Impact on Digital Marketing

Click IDs and user IDs in query parameters (gclid, fbclid, mkt_tok, and others - See Cory Underwood 's list of parameters and vendors) are unique visitor identifiers that can identify visitors across sites, clicks on a marketing campaign or individuals, depending on the platform.

ClickIDs and user IDs are effectively breadcrumbs used by sites/platforms to facilitate the identification of a specific device or individual.

These unique IDs help the advertising platforms improve campaign performance, identify users cross-site, calculate ROAS, and segment customers based on their behavior by associating the device with the click ID or user ID.

Without click IDs, advertising platforms are blind to where the user originated, making it impossible to optimize campaigns and calculate return on ad spend.

Three examples of tracking that will stop working:

? Conversion tracking with Google/FB Ads

? Tracking of a visitor who clicks a Marketo campaign

? Cross-side tracking using query parameters

What Should Marketers and Data Leaders Do?

There are two options:?buy?or?build.?Buy into an attribution platform (make sure you track events as first-party) OR build a Modern Marketing Data Stack.

For now, independently, if you buy or build, you should:

  1. ?? Add UTMs (or equivalent) query parameters that identify the origin of paid traffic for all your campaigns. UTMs will associate the visitor identifiers provided by your analytics tool (first-party cookies and or a fingerprint) with the campaign that brought them to the site.
  2. ?? Purposefully add identifiers across your digital properties to facilitate identity resolution. See Article.
  3. ?? On conversion, collect in your backend all your visitor identifiers (e.g., cookies, user ID, fingerprint, and others).

In a simplistic approach correlating the data from your analytics tool with your conversions will give you the traffic origin for each conversion.

Note attribution is a complex data science problem influenced by several factors, including but not limited to:?

  • Duration of the sales cycle
  • Cross-device tracking?
  • Overlapping campaigns
  • Multiple Channels

In a build scenario, attribution is a challenge for data teams.

Final Thoughts

With each iteration of privacy-enabling features, Apple is making small businesses' lives increasingly difficult. Larger organizations have the financial and technical resources to solve attribution regardless.

Fortunately, we see new SaaS companies and Marketing agencies implementing a?Modern Marketing Analytics Stack?on behalf of their customers and integrating it into their products and Ad Management services.

At Snowcatcloud, we provide customer data infrastructure to help organizations tackle, among other things, customer attribution, cost-efficiently, securely and at scale.

Follow me on Twitter and LinkedIn for more #analytics #data #


Jim Gianoglio

Helping marketers measure performance in a privacy-focused era.

1 年

I love that Apple keeps helping me grow my business ??

Michael Marquez

Analytics Engineer at PlayStation ??

1 年

Welll that sux. so cross domain tracking for purchasing sights will lose tracking.

回复
Mickael Tanious

Growth Marketing Consultant | Expert in PPC, Paid Social, GA4, GTM, CDPs, and more.

1 年

Few thoughts: 1 - How do they determine what vendors will be targeted? Because TikTok, Snap Inc., Pinterest, & Reddit, Inc. are currently safe according to the current list. My guess is that it's based on Apple's analytics on what the most common identifiers are; ironic that they're tracking users to prevent users from being tracked. 2 - The natural solution platforms will take is having websites try to collect more PII and send that data for customer matching. Feels counterintuitive to privacy since now companies will want way more personal data vs. a random identifier. 3 - Feels like this mainly hurts SMBs. Large companies will hire experts who can work around it. E.g. using path names instead of query params, using a custom key with S2S tracking (e.g. randid={gclid}), etc. Meanwhile, small businesses won't be able to improve their ad performance because of difficult tracking AND they'll be hit by the collective algorithmic data loss that enables companies with 0 sales to be successful on ad platforms. No specific opinion on whether or not this change is good or bad. User privacy is a complicated issue. Either way, great article Joao Correia!

要查看或添加评论,请登录

社区洞察

其他会员也浏览了