For large organizations, particularly multi-channel retailers, ensuring that the data flows seamlessly from user interactions to analytics and marketing tools is critical. The challenge, however, lies in how you capture and route this data in the most efficient, scalable, and compliant way.
There are two primary ways to collect user interaction data:
- Directly through a Customer Data Platform (CDP)
- Via a Tag Manager (like GTM), which then forwards the data to the CDP
Each method has its implications, affecting engineering, marketing, compliance, and data governance across the board.
From an Enterprise Architecture standpoint, the argument for using the CDP as the primary data collection layer is often overstated. A CDP’s true strength lies in processing and activating customer data, not in replacing the specialized role of a Tag Manager. Experienced marketing teams typically recognize this distinction, but it’s still common to see businesses (and CDP vendors) advocate for CDP-led data collection.
That’s why Enterprise Architects need a clear, objective framework for evaluating these options, especially when making the case for Tag Manager-driven tracking.
In this article, we’ll explore five key architectural decision areas that should guide your data collection strategy and help you make a future-proof decision. These include
- Security and compliance
- Data quality and governance
- Deployment speed and flexibility
- Tagging non-CDP properties
- Pricing and TCO
Let’s jump right in!
Security And Compliance
Privacy and consent management are critical, especially in regulated environments like GDPR and CCPA.
CDP-approach proponents would argue that the CDP tag can also respect consent from tools like OneTrust. However, security is not just about WHO gets tracked, but also what data about them. With an independent CDP tag directly injected into the site, it is not difficult to imagine unsuspecting marketing teams passing sensitive data to the CDP. This would be a severe compliance risk, and implementing data audits would become almost impossible in this scenario.
Imagine a brand with multiple teams using the CDP. In theory, each team could have its own set of data layer variables that they might want to push into the CDP tag. Without a centralized audit facility, this would become a security nightmare.
With a centralized tag management system, the data that gets pushed out from the backend to the data layer is usually well scrutinized and has to pass through multiple governance checks.
Better data quality and governance
The Tag Manager approach is known for its strong control and consistency when it comes to ensuring data quality and governance. The assumption here is that any tags that go live are approved by the central analytics team, as opposed to local businesses (which may be the case with a Direct-to-CDP approach)
- Schema Enforcement: All data collection usually follows a well-defined schema since it is populated in the data layer using enterprise backend systems. For example, all data about a cart abandonment event will have consistent field names for various objects like product, category, and campaign. Any client-side validation rules will also be uniformly implemented regardless of data destination.
- Less Risk of Inconsistencies or Duplicate Events: Since all event tracking is centralized within the Tag Manager, there’s a lower risk of inconsistent data or duplicate events. This is particularly useful for large organizations with multiple teams tracking the same types of events (e.g., product views, purchases, etc.). For example, if a retailer sets up a unified event definition for a “purchase” event in the Tag Manager, there’s less chance that multiple teams will accidentally track “purchase” in different ways, leading to cleaner, more reliable data.
- Version Control: The Tag Manager approach provides version control and audit trails for your tags, which is crucial for governance, especially in enterprise settings. This enables teams to track who changed what, when, and why. With a Direct-to-CDP setup, you may not have the same level of visibility into changes across all tags and data collection setups unless you’re closely managing the CDP’s interface.
Deployment speed and flexibility
- With the Direct-to-CDP approach, any modifications to the CDP tag are tied to the app release cycles and deployment schedules, creating potential dependencies on IT teams. In large companies, changes to live properties are typically bundled into periodic releases rather than deployed on an ad-hoc basis.
- In contrast, the Tag Manager approach allows all changes to be made directly through the GTM interface. This puts the control in the hands of the analytics, marketing operations, or mar-tech teams, enabling quicker implementation without relying on IT.
- Tag Managers provide an easy preview and debugging interface to test tag firing conditions in real-time. With the Direct-to-CDP deployment approach, testing or making quick changes (like filtering or adjusting event logic) may require more development work (e.g., adjusting JavaScript code or CDP settings), which is slower and more resource-intensive.
Tagging non-CDP properties
In many enterprise environments, it’s common to encounter digital properties where the CDP is not yet live or cannot be easily implemented. These include:
- Campaign microsites built for short-term use
- Partner-owned domains where you don’t control the infrastructure
- Web or mobile properties belonging to specific business units not yet onboarded to the CDP
This is particularly true when CDP rollouts are done in phases — which is often the case in large organizations.
Direct-to-CDP Approach: What You Should Know
If your data strategy depends entirely on direct CDP tagging, you may run into challenges in these scenarios:
- The CDP’s JavaScript SDK must be installed — something that may not be feasible on third-party or externally managed sites.
- Implementation typically requires developer support for setup and data mapping. Future changes will also depend on development cycles.
- In theory, server-side event forwarding can help, but it depends on having backend access and integrations, not always available for short-lived or external sites.
Tag Manager-Based Approach: A More Flexible Option
By contrast, using a tag manager like Google Tag Manager (GTM) offers significant flexibility and speed:
- GTM can often be independently installed, even on third-party or campaign-specific sites.
- It supports rapid tag deployment and event tracking, often without writing new code.
- Through GTM, you can build templates or custom tags to forward data to multiple destinations — including the CDP — without being locked into a single pipeline.
Example:
A large retail brand partners with an agency to launch a holiday campaign microsite. Instead of waiting for CDP enablement, the marketing ops team quickly installs GTM. Within hours, they’re capturing key events like “Add to Wishlist” and “Email Signup” — and forwarding that data to downstream platforms like Klaviyo, InMoment, and the CDP. Because GTM already supports the CDP as a destination, no extra code is needed — it’s a configuration step, not a deployment one.
Pricing Considerations
Pricing is a critical — and often overlooked — factor when choosing between a Direct-to-CDP data collection approach versus one that uses a Tag Manager like Google Tag Manager (GTM) as the collection layer.
How CDP Vendors Typically Charge
Most Customer Data Platforms (CDPs) — like Segment, BlueConic, or Tealium AudienceStream — follow event-based pricing models. This means:
- You’re billed based on the number of events ingested per month or year.
- Some CDPs charge per Monthly Tracked User (MTU), which counts how many distinct users generate events.
- In some cases, fees increase based on data volume, number of integrations, or API calls.
Implication: If you send every click, scroll, or view directly to the CDP (even those not relevant for activation), your bill can quickly escalate. There’s also a risk of overage charges if volume limits are exceeded.
How Tag Managers Typically Charge
Most Tag Managers — like Google Tag Manager, Adobe Launch, or Tealium iQ — operate under very different models:
- Google Tag Manager is free for most use cases (unless you’re using GTM 360, the enterprise version).
- Even paid Tag Managers are usually priced on flat licensing models, often not tied to event volume.
- GTMs act as a filtering layer, allowing teams to control which events are sent to downstream tools like CDPs, reducing unnecessary data transfer.
Implication: Tag Managers don’t add marginal cost per event. You can use them to refine and throttle what actually gets sent to tools like the CDP, controlling costs downstream.
Example: Retailer with 10 Million Monthly Events
Let’s say a large retailer generates around 10 million events per month from web and mobile user interactions.
Scenario A: Direct-to-CDP
- All 10 million events are sent to the CDP.
- CDP charges $x per 1,000 events.
- Monthly cost = 10,000 x $x
- Additional costs: More ingestion = more processing, more integration setups, more data audit complexity.
Scenario B: GTM → CDP
- GTM collects all 10 million events.
- Only 3 million are deemed useful and forwarded to the CDP after filtering.
- CDP cost = 3,000 x $x – which is almost 1/3rd the cost of direct integration
Even if the team spends more upfront to configure GTM filters, the long-term Total Cost of Ownership (TCO) is significantly lower.
Final Takeaway: Use the Tag Manager for Data Collection — Not the CDP
In enterprise environments, architectural decisions must serve the long-term interests of the entire organization, not just the needs of individual teams or projects. While CDPs are powerful for unifying and activating customer data, they are not built to handle the complexity, governance, and agility required for client-side data collection at scale.
Relying on the CDP for direct tracking may seem like a quick win, especially for business-led initiatives, but it often results in fragmented governance, increased costs, and unnecessary IT dependency down the road.
The smart move? Let the Tag Manager own the data collection layer. It provides centralized control, faster iteration, stronger compliance alignment, and significantly lower total cost of ownership. Reserve the CDP for what it does best: processing, enriching, and activating customer data across channels.
Enterprise architects should champion this distinction because getting it right at the foundation layer sets the stage for scalable, secure, and sustainable data operations.
Want expert help making the case for Tag Manager at your organization? Get a detailed recommendation report based on your company scenario. Your own data, systems and team dynamics!
Book a free discovery call with our team today.
