Ad platforms are phasing out manual optimization tools, leaving advertisers with one key lever: the signals they send back. As automation takes over, the quality of those signals increasingly determines how well campaigns perform.
In this webinar, we’ll take a practical look at conversion signals - what they are and why they matter. We’ll also show how the Stape platform supports the implementation of both online and offline conversion signals. Advertisers still sending weak data to ad platforms will continue to fall behind those who’ve learned how to show platforms what “good” looks like. With signal quality emerging as a key trend for 2026, now is a great time to get clear on the fundamentals.
If you had to choose, would you prioritize increasing signal volume or improving signal quality—and what specific data enrichment has delivered the biggest performance lift in your real campaigns?
We wouldn’t choose one over the other. First, make sure your signal volume is as close to 100% as possible compared to your source of truth (CRM/CMS). Then, enrich events with as many relevant parameters as you can send to the platform. In practice, ad platforms value user data the most (in addition to basics like IP address and user agent). Better user data helps platforms build stronger audiences and optimize toward people who are more likely to convert.
When it comes to tracking data from campaigns where conversions start on Instagram (message conversions), with no pixel setup or website—can your tool help with that?
From a consent perspective, how do we label offline conversions if the client never saw the cookie banner?
Google Ads Offline Conversions includes an “unknown” consent option for cases like this. If you’re in a GDPR jurisdiction and you have not received valid consent, you may not be able to legally transfer those conversions (and there can be local nuances). We recommend checking with your solicitor or DPO for the correct approach in your jurisdiction.
How should agencies prioritize signal improvement versus increasing media spend?
Both matter, but signal quality directly impacts how effectively your ad spend can work—so improving signals often makes your media budget perform better.
Do you have recommendations for implementing server-side tracking when conversions happen on a separate website/domain, and the third-party vendor doesn’t allow third-party tracking?
If you have no control over the third-party domain where the conversion happens, server-side tracking typically won’t work there, because you need a validated custom domain to operate in a first-party context. That said, many third-party platforms allow you to host part of their service on your subdomain, which solves the issue. Another common option is a redirect to a custom thank-you page on your domain, where you can track the conversion as usual.
How do you validate that the signals you’re sending are actually improving model learning—not just increasing reported conversions?
In practice, you usually validate this through improvements in performance metrics—more attributed conversions and better ROAS (and/or lower CPA), assuming everything else is held reasonably consistent.
Will low ad spend negatively impact EMQ?
EMQ is heavily influenced by the user data you provide. In general, the more click identifiers (and other relevant parameters) you send, the higher the score can be. So it’s not that low spend “hurts” EMQ—rather, higher spend can sometimes increase the amount of identifiable traffic, which may improve EMQ.
We run B2B automotive campaigns on Meta. Most traffic is mobile, but users often switch to desktop before converting. When they switch devices, we lose fbp/fbc and can’t attribute conversions to Meta (user data matching is weak in B2B). How can we solve this mobile-to-desktop attribution gap?
The main option is cross-device/cross-browser tracking. However, it’s complex and can introduce drawbacks—especially the risk of incorrect/false attribution.
What’s the difference between cost of goods sold (COGS) in Google Merchant Center and profit margins sent via a tracking server?
COGS is a native parameter, so you’re effectively delegating the calculations to Google. If you send profit directly, you have more flexibility in how you build reporting and performance dashboards.
When a dispensary sends Google Ads traffic to a compliant landing page and then to the main site, what’s the best way to structure tracking so we don’t lose attribution between domains?
Use a landing page on a subdomain of the main site. That way, you’re less likely to lose cookies and other attribution attributes.
How would you track consent when using webhooks and collecting them with a Data Client? It seems like a workaround.
Store the user’s consent status in your CMS/CRM. Then include that consent status in the webhook payload sent to your server-side GTM endpoint. Based on that field, you either send conversions or suppress them. In some cases, you can also enforce consent at the webhook-sending stage (depending on the CMS/CRM).
What’s the best way to estimate monthly cost for an sGTM subscription?
How long would it take to test and validate the implementation you just created?
It depends on your experience and the scope (number of platforms, events, and complexity). It can range from a few minutes (simple ecommerce + one platform) to several days (complex flows with many platforms and events).
Why should I use the add_to_cart_stape event instead of the normal add_to_cart?
Sometimes there’s another data layer implementation (or a “rogue” setup) firing the same event somewhere in the code. Adding _stape helps ensure you don’t double-count events if another add_to_cart is firing.
What is the recommended webhook URL path format when sending delayed events (via Stape Power-Up) from the CRM to sGTM? I’ve seen conflicting documentation—what’s correct?
[Stape webinar] Consent explained: practical implementation for web and server-side tracking
Learn how consent works in practice and how to implement it correctly for analytics and advertising. Clear explanations, real demos, and practical guidance.
Comments