First-Party Data Is Now the Only Data That Matters

The advertising industry spent six years and considerable fortune bracing for a cookieless future. The cookies stayed. The competitive moat built on borrowed data did not.

That is the sharper argument Tealium’s Nick Albertini makes in a recent piece worth reading slowly: the real disruption was never a technical deprecation. It was the quiet compounding of an ownership gap — between brands that have been systematically building direct relationships with their audiences, and those still renting reach from platforms they do not control.

The Pharmacy Test: Why Media Creativity Loses to Data Quality

Albertini’s case study from a national pharmacy retailer illustrates the gap cleanly. The campaign that actually drove vaccination uptake was not the most inventive in terms of media planning. It was the one grounded in the retailer’s own customer data — purchase history, loyalty behaviour, location signals — that allowed them to identify who was genuinely at risk and reachable through a channel they trusted.

This is not a story about programmatic sophistication. It is a story about data provenance. Owned data carries context that rented data structurally cannot: behavioural sequences, consent signals, relationship history. In Southeast Asia, where Shopee and Lazada loyalty programmes generate some of the richest transactional datasets in the world, brands that are capturing and activating their own slice of that behaviour have a compounding advantage that platform fees cannot simply replicate.

The implementation question is less glamorous than the strategic one: it starts with tagging architecture, customer data platform selection, and consent management that is genuinely granular — not the checkbox theatre that still passes for compliance in too many markets.

Agent-Ready Data: The Infrastructure Bet Worth Making Now

The announcement from Fivetran and dbt Labs at dbt Summit 2026 deserves attention from marketing data teams, not just engineering. The joint debut of Fivetran’s Context Layer — which enriches pipeline metadata so AI agents can reason about data provenance and freshness — alongside dbt Charts and the broader open lakehouse vision represents a meaningful shift in what enterprise data infrastructure can do.

The strategic implication: first-party data that is well-modelled and semantically enriched is not just more useful for human analysts. It is dramatically more useful for the AI agents that are increasingly doing the activation work — audience segmentation, bid optimisation, personalisation sequencing. Garbage-in still produces garbage-out, but the velocity at which that garbage compounds has accelerated.

For Southeast Asian brands managing multilingual catalogues across platforms like LINE, TikTok Shop, and Shopee simultaneously, the ability to have a single, agent-readable data model that understands context across those channels is no longer a nice-to-have. It is the difference between a marketing team that can act on signals in hours and one that is still waiting on a data pull by end of week.

The practical starting point: audit whether your current data models carry enough metadata for an AI agent to understand what a field means, when it was last updated, and how confident the system should be in it. Most do not.

---

Pricing Intelligence as a First-Party Data Proof Point

Denmark’s FTZ selecting Zilliant for real-time pricing — after a live benchmark confirmed superior performance at scale — is a useful illustration of what first-party data looks like when it is genuinely activated at the sharp end of revenue.

FTZ, operating across high-volume digital channels for automotive parts, is deploying Zilliant’s Price Manager and Real-Time Pricing Engine to generate customer-specific prices dynamically. The underlying fuel for that engine is transactional history, customer segment behaviour, and margin data that only FTZ owns. No data co-op, no modelled lookalike audience — just the signal quality that comes from years of direct commercial relationships, properly structured.

This is a model that Southeast Asian distributors and platform-adjacent retailers should be watching. Brands on Lazada or Shopee that have negotiated data-sharing arrangements with platform partners, or that have built sufficient owned-channel volume to see genuine behavioural patterns, are sitting on pricing intelligence they are almost certainly underusing. The infrastructure to act on it — real-time engines, governed pricing workflows — is now available at a scale that does not require the budget of a Tier 1 automotive supplier.

The failure mode to avoid: building a first-party data programme that is rich in collection but thin in activation. Data that sits in a warehouse and feeds a monthly report is not a competitive asset. Data that feeds a pricing engine, a personalisation layer, or a retention trigger is.

Consent Is Not a Legal Checkbox — It Is a Data Quality Signal

Here is the argument that still does not get enough airtime in regional marketing conversations: consented first-party data is not just safer data. It is better data.

When someone opts into a loyalty programme, downloads a brand’s app, or registers for a webinar, they are signalling intent and category relevance in a way that a modelled third-party segment cannot approximate. In markets like Thailand and the Philippines, where personal data protection legislation is still finding its enforcement stride, brands that build consent infrastructure now — not because they are forced to, but because it produces better signal — will inherit a structural advantage when regulatory expectations tighten.

The practical architecture here involves three layers: a consent management platform that captures and versions permissions at a granular level; a CDP or data warehouse that tags every downstream record with its consent lineage; and activation tooling that respects those tags at the point of use, not as an afterthought. The brands that have built this are not just compliant. They are operating with higher-confidence data that converts better, segments more accurately, and ages more gracefully than anything purchased from a third-party broker.

The open question for marketing leaders in Southeast Asia: if your current data programme disappeared tomorrow, how much of your audience understanding would survive? If the honest answer is “not much,” you are not building an asset — you are renting one.

---

At grzzly, first-party data strategy is where we spend a disproportionate amount of our thinking — because it is where the longest-lasting competitive advantages get built, and where most brands in Southeast Asia are still leaving significant value on the table. If you are ready to move from data collection to genuine data ownership, Let’s talk.