Treasure AI Launches Personalization Studio and Engagement Pricing Model

Introduction

As a data and CRM consultant, when assisting Taiwanese enterprises in implementing CDPs, BigQuery, and AI automation tools, I frequently hear two deep pain points from marketing executives.

The first pain point is "excessive communication and waiting costs": When marketing teams want to launch a website personalization campaign, they often need data engineers to build Data Pipelines, followed by web engineering teams to set up Event tracking and UI adjustments. By the time the campaign goes live, the optimal window has long passed.

The second pain point is "disconnection between software pricing and business outcomes": Enterprises pay huge annual fees for email send volume, but this cost does not directly align with engagement or conversion outcomes, sometimes even damaging brand reputation through blast messaging.

Today, in October 2026, as the center of gravity in the martech stack shifts toward AI-powered applications and data warehouses, how businesses operationalize their data is undergoing a fundamental transformation.

Treasure AI's latest product launch and pricing model directly target these two martech headaches. This is not merely a vendor product update; it reveals the inevitable path for enterprises building CRM and personalized experiences in the future.

What Happened: Key Highlights

At the Agentic World 2026 conference, Treasure AI announced "Personalization Studio," a marketer-facing UI, alongside a new engagement-based pricing model tied directly to performance.

Personalization Studio is part of the Personalization AI Suite within Treasure AI's Agentic Experience Platform (AEP). Its main feature is providing a user-friendly interface that allows marketing teams to build, launch, and manage real-time website personalization directly.

Architecturally, the tool is built on a unified customer data foundation, combining batch profile data (including audiences, loyalty status, purchase history) with real-time behavioral signals to deliver sub-second personalized experiences.

Rafa Flores, Chief Product Officer at Treasure AI, stated: "You can't really sit in the data warehouse and do personalization in real time because the data warehouse can't do real time." He emphasized that Personalization Studio allows marketers to come in and build a campaign with personalization in 10 minutes.

Alongside the product release, Treasure AI introduced an engagement-based pricing model that ties email costs directly to customer actions, such as clicks, rather than the number of messages sent.

Flores noted that engagement pricing is essentially a bet on Treasure AI's platform intelligence and its ability to deliver better outcomes, stating that they are "betting on the click-through rate."

This movement reflects broader industry trends. As AI works its way into martech platforms, vendors are re-thinking pricing models. HubSpot recently launched a new pricing model for new customers in the EMEA region, and Salesforce's pricing model for Agentforce has evolved since its debut.

Why It Matters: Impact on Data, Marketing, and CRM Teams

This announcement provides critical architectural signals for marketing, data, and CRM teams in Taiwanese enterprises.

First, the "decoupling of technical dependencies." Adding standalone personalization tools to the martech stack typically requires separate integrations, data pipelines, and profile stores.

This fragmented stack leaves marketers dependent on data and web teams whenever they want to launch or change a campaign. Personalization Studio unifies the underlying profile data while giving operational control back to marketers, drastically reducing time-to-market.

Second, recognizing the "natural boundaries of data warehouses." Many organizations assume that deploying a cloud data warehouse (such as BigQuery) allows them to run real-time marketing directly via SQL. However, data warehouses excel at batch analytics, not sub-second real-time execution.

Treasure AI's approach illustrates a modern data architecture: storing static audiences, loyalty status, and purchase history in data warehouses or CDPs, while evaluating real-time behavioral signals at the edge. This provides a viable solution for sub-second personalization.

Third, the "shift in software pricing dynamics." Traditional volume-based pricing incentives blast messaging, creating stack bloat and noise.

Tying pricing directly to customer engagement like clicks aligns vendor incentives with enterprise goals. Vendors must ensure their AI recommendations drive clicks for payment, pushing marketing teams away from volume toward high-quality personalized engagement.

Practical Applications: What Taiwanese Enterprises Can Do

In response to this evolution in international MarTech tools, Taiwanese enterprises planning CDP, CRM, and AI deployments can adopt the following actionable strategies.

In retail and e-commerce, when marketing teams want to deliver real-time limited-time offers to high-tier members (loyalty status), they traditionally rely on engineers to code dynamic layouts and rules. (Inferred) By leveraging tools that combine batch profiles with real-time behavioral signals, marketers can set up rules based on category browsing frequency and publish sub-second dynamic banners within 10 minutes.

In insurance and finance (InsurTech), customer policy histories and risk tiers (purchase history) are strictly managed batch data. (Inferred) Financial marketers can use these static profiles as base attributes, and when a customer uses an online calculator for mortgages or auto insurance, combine real-time click signals to present personalized consultation portals without needing IT system deployments.

For B2B and technology sectors, avoiding tool bloat is an urgent priority. Before adopting new tools, teams should verify whether they introduce additional data pipelines.

Organizations should establish reusable prompt frameworks, prompt logs, and guardrails to prevent AI-generated content from going off track and to prevent AI usage costs from scaling out of control.

Marketing and CRM leaders should audit current marketing automation contracts, especially email send fees. During future vendor negotiations, teams should proactively request engagement-based pricing terms aligned with performance outcomes.

Implementation Evaluation: Costs, Risks, and Limitations

While no-code interfaces and engagement pricing are appealing, decision-makers must maintain a realistic assessment of potential risks and constraints.

First, the risk of tool bloat and data silos. Adopting standalone personalization tools without a unified data foundation creates redundant data pipelines and profile stores, increasing the burden on technical teams.

Second, real-time compute and infrastructure costs. Sub-second personalization relies on robust real-time decisioning engines. If underlying data quality is poor or API latencies exist, even Personalization Studio cannot deliver expected outcomes.

Third, budget control risks under click-based pricing. While click-based models guarantee engagement, unmonitored campaigns with high click counts but low downstream conversions can rapidly consume marketing budgets.

Fourth, the lack of AI governance and guardrails. AI automation without proper permission controls or prompt logs risks generating non-compliant personalized messaging that strays from brand guidelines.

Before adopting the latest AI tactics, leaders should demand clear evidence, experience, and replicable results to ensure the technology addresses genuine business challenges rather than short-term market noise.

Action Checklist for Decision Makers

To translate these trends into concrete operational changes, marketing, CRM, and data leaders should execute the following four steps:

• Audit technical dependencies in the MarTech stack: Evaluate whether launching or modifying website personalization requires IT/data engineering tickets, and measure average time-to-market.
• Redefine boundaries between data warehouses and real-time execution: Retain data warehouses like BigQuery for deep batch analytics and profile storage, while routing sub-second personalization to dedicated real-time decisioning layers.
• Evaluate communication channel pricing structures: Assess whether current email/messaging tools charge by volume or engagement, and recalculate ROI under an engagement-based pricing model.
• Establish guardrails for AI automation and personalization: Implement prompt logs, content review mechanisms, and spending caps to prevent runaway AI costs or compliance risks.

Suggested KPIs to Track:

• Campaign Time-to-Market: Time required from campaign concept to live personalized implementation (target: minutes).
• Click-Through Rate (CTR) and Cost per Engagement: Measure whether personalized recommendations drive customer actions and budget efficiency.
• Engineering Ticket Volume for Marketing: Reduction in data/web engineering support tickets requested by marketing teams for personalization.

Extended Reading

• A Seller Pain Point, Small Squad, and a Solution That Shipped in Weeks
• Trustworthy, Explainable, and Accountable: How to Give AI Autonomy Without Letting It Run Wild

Jason Analytics empowers enterprises to drive CRM and marketing growth through data and AI. Redistribution welcome with attribution; contact us for consulting opportunities.

Original source： Treasure AI takes on two martech headaches at once — MarTech.org