Monthly Digital Insights – August 2026

This month, we look at new AI capabilities in analytics and Customer Data Platforms (CDPs), why Content Security Policy (CSP) deserves more attention from analytics teams, and where you can meet us at Data Expo in Utrecht this September.

AI and CDPs are moving from reporting towards action

Analytics and tagging tools such as Tealium, Hightouch, Piano Analytics, Google Tag Manager (GTM) and Google Analytics 4 (GA4) were busy this summer introducing new features. These updates align with the industry's shift from traditional reporting towards AI-assisted decision-making, activation and automation.

Hightouch, for example, introduced its concept of an Agentic CDP, where AI agents can use customer data and business context to support activities such as audience creation and campaign execution.

Google is also expanding its Gemini-powered "chat with your data" capabilities. Ask Advisor brings natural-language analysis to GA4, while the Conversational Analytics API enables similar interactions with data from sources such as BigQuery, Looker and Looker Studio.

Tealium has also made its Audience Fill functionality generally available, making it possible to use historical visitor data to populate audiences for activation.

Why this matters

These developments can reduce the time your teams spend manually querying data, building audiences and moving insights between tools. But they also make the quality of the underlying data more important. AI can only make useful recommendations or take reliable action when customer identities, consent and measurement are in order.

A strong data foundation, identity resolution and clear data strategy therefore remain essential. As more repetitive tasks become automated, analytics teams can spend more time on implementation, orchestration, governance and quality assurance.

Content Security Policy deserves more attention from analytics teams

Content Security Policy (CSP) helps protect websites against attacks such as Cross-Site Scripting (XSS). While CSP is primarily a security mechanism, it can directly affect analytics and marketing implementations.

Stricter policies can block JavaScript, analytics tags, marketing pixels and consent management technologies when they are not configured correctly. At the same time, overly permissive policies can leave unnecessary security risks in place.

MDN recommends a strict CSP based on nonces or hashes instead of relying solely on domain allowlists. Trusted Types API can provide an additional layer of protection against DOM-based XSS.

Why this matters

At our clients, we increasingly see a growing reliance on JavaScript and third-party scripts for measurement and activation. This makes CSP changes particularly impactful: a single blocked tag can already affect reporting, consent flows and audience activation. We therefore recommend strict nonce-based CSP implementations combined with Trusted Types, rather than traditional allowlist-based approaches. Analytics and security teams must work closely together to ensure CSP policies protect users without disrupting measurement, consent management or marketing technology integrations.

When implementing or tightening CSP, involve your analytics specialists early.

Meet us at Data Expo in Utrecht

On 9 and 10 September, we will be at Data Expo 2026 in Utrecht, where we will share several customer cases:

• How Glaspoort broke its dependency on BI reporting with Databricks
• How Vodafone Germany uses its Customer Data Platform as a Context Engine for AI and Activation
• How data helps Wakibi grow from €3M to €10M in annual lending
• From flows to agents: scalable AI for customer service at Univé

If you are attending Data Expo, you can find us at booth #10. Come and talk to us about web analytics, CDPs, AI, measurement strategy, privacy or how to build a stronger data foundation.

Key takeaway

The developments we have highlighted this month point in a similar direction. Analytics and marketing technology are making it easier to move from data to action. This creates opportunities to reduce manual work and speed up activation, but it also increases your dependency on reliable data, measurement and governance.

AI readiness starts with data readiness. Before giving AI more autonomy, make sure it has the reliable data, context and guardrails it needs.

Want to continue the conversation? Get a free ticket for Data Expo and meet us in Utrecht on 9 or 10 September.

Receive data insights, practical cases and a look behind the scenes once a month?

Sign up to our mailing list and stay 'up to data':