Table of Contents

Background — Why UDN Needed a Change

Traditional publishers around the world have faced steep pressures on advertising revenue as major digital platforms (like Google and Meta) capture a growing share of ad spend and online traffic. This has caused publishers’ ad revenue to stagnate or decline because:

AI-driven search tools and platforms increasingly keep users engaged without sending them to external publisher sites, reducing ad impressions from search traffic. (FinancialContent)
Publishers have historically relied on broad, shallow audience segments and direct sales models, which struggle to compete with data- and algorithm-driven platforms. (inma.org)

In this environment, UDN chose to rebuild its advertising model rather than retreat from the market. (inma.org)

UDN’s Strategy: From Inventory-Driven to Intelligence-Driven Advertising

UDN’s success came from transforming its entire advertising ecosystem by using data and AI at every stage of the ad lifecycle — not just as a tool but as a foundational business strategy. (inma.org)

📌 1. Building a Robust Data Infrastructure

UDN created systems to capture and organise:

first-party data about device use, behavioural signals, and context, and
compliant tracking that respects user privacy.

This infrastructure made data usable for both internal analytics and advertising partners. (inma.org)

2. Turning Data Into Audience Intelligence

Rather than selling generic audience segments, UDN:

analysed raw behaviour to define 110+ distinct audience categories
used AI to interpret patterns in user behaviour, interests, and conversion signals. (inma.org)

Result: advertisers gained deeper insight into who their ads would truly reach — and why. (inma.org)

3. Applying AI to Targeting and Prediction

UDN deployed advanced AI models for several purposes:

Semantic analysis to understand content and context,
Predictive modelling to estimate click and engagement likelihood,
Lookalike modelling to find new users similar to high-value audiences. (inma.org)

This AI layer enabled smarter campaign planning and reduced waste in ad delivery — especially compared with blunt demographic targeting alone. (inma.org)

4. Integrating Creative Intelligence

UDN didn’t stop at data science; it linked insights to creative execution by:

using AI to test and match ad creative to audience segments
optimising cross-channel delivery (display, native, mobile, etc.)
ensuring messages resonated with predicted behavioural patterns. (inma.org)

This creative-intelligence loop helped campaigns perform better AND avoid irrelevant placements. (inma.org)

5. Continuous Optimisation

Instead of one-off campaigns, UDN embedded:

ongoing measurement
feedback loops
dynamic refinement of targeting and creative

This meant campaigns continuously learned from performance data — improving results over time. (inma.org)

Real Results — Evidence of Revenue Recovery

Under this systemic transformation:

UDN executed 2,000+ data-driven campaigns with the new platform.
Click-through rates increased by 200 % in high-value categories.
The company achieved over 45 % penetration across its advertiser portfolio, meaning many advertisers expanded their campaign footprints with UDN. (inma.org)

Importantly, advertisers shifted from asking “How many people did we reach?” to “Who exactly did we reach and why does it matter?” — a sign of higher-value advertising relationships. (inma.org)

Comments and Reflections from UDN Leadership

Carrie Wong — Director, Data Development Department, UDN:

“Our breakthrough was not about AI as a standalone tool, but about structurally integrating data and AI across the entire advertising lifecycle. By uniting data, insights, creativity, and optimisation, we shifted from inventory-driven to intelligence-driven advertising.” (inma.org)

Wong emphasises that the transformation required cross-functional alignment, not just technology — data became a shared language across editorial, product, and sales teams, fostering trust and practical adoption. (inma.org)

Industry Reactions and Wider Context

1. Publisher Pressure and the Need for Innovation

Many publishers globally struggle to reclaim advertising revenue lost to dominant digital platforms, with only a small share confident about meaningful recovery without strategic change. (comunicacionmarketing.es)

UDN’s results contrast sharply with the broader industry uncertainty — showcasing how data and AI strategies can unlock value even as traffic and traditional ad revenue models face disruption. (comunicacionmarketing.es)

2. AI as Revenue Growth Engine

The broader ad tech ecosystem — from big platforms like Meta to agency groups — demonstrates that AI can drive better ad performance and pricing by improving targeting accuracy, increasing conversion rates, and enabling real-time optimisation. (AInvest)

UDN’s work mirrors this trend in a publisher context — moving beyond traditional display inventory to data-enhanced, performance-driven campaigns. (inma.org)

Why UDN’s Approach Matters

UDN’s transformation illustrates a blueprint for modern media companies seeking to reclaim lost advertising revenue:

From broad reach to nuanced relevance
Advertisers today demand precise targeting and measurable ROI — not just ad placements. (inma.org)

From reactive campaigns to predictive optimisation
AI enables campaign systems that learn and adapt, creating long-term value rather than one-off buys. (inma.org)

From siloed departments to integrated intelligence
Cross-team data adoption ensures insights inform creative, sales, and product decisions holistically. (inma.org)

From revenue decline to strategic growth
By embedding AI across the ad lifecycle, UDN didn’t just stabilise revenue — it helped advertisers see value in working more deeply with the publisher. (inma.org)

Here’s a detailed, case-study-focused look at how United Daily News Group (UDN) — one of Taiwan’s largest media organisations — leveraged data and AI to recover and grow advertising revenue in a tough media market, plus real results and commentary from industry insiders: (inma.org)

The Challenge: Traditional Advertising Under Pressure

Like many legacy news publishers, UDN experienced declining ad performance and weak audience targeting as digital platforms captured more ad spend and commoditised inventory. More than 70 % of UDN’s ad revenue came from direct sales, but these were largely inventory-driven rather than intelligence-based, leading to shallow segmentation and inefficient campaign outcomes. (inma.org)

UDN’s Solution: Systemic Data + AI Transformation

Instead of incremental tweaks, UDN launched a comprehensive transformation that re-oriented its ad business around data intelligence and AI, turning raw user signals into actionable insights across the advertising lifecycle. (inma.org)

1. Advertising Data Infrastructure

UDN built a compliance-ready data platform that captures rich device, behavioural, and contextual signals — the foundation needed to make advanced AI analytics possible. (inma.org)

2. Audience Research & Insight Segmentation

Rather than broad categories like “male/female” or “age bracket,” UDN used data to define 110+ industry and lifestyle-oriented audience groups. These refined categories let advertisers target real behavioural patterns (e.g., tech-interested families vs. urban professionals), not just demographic boxes. (inma.org)

3. AI-Powered Targeting and Predictive Modelling

UDN applied multiple AI techniques:

Semantic analysis — understanding how content resonates
Predictive models — estimating likelihood of engagement
Lookalike modelling — finding new audiences similar to high-value segments

This predictive layer helped move from generic reach to audience affinity and intent targeting. (inma.org)

4. Creative & Omni-Channel Intelligence

Rather than treating creative messaging separately, UDN integrated AI insights into creative testing and optimisation, improving relevance across digital platforms and ad formats — strengthening the connection between message and audience. (inma.org)

5. Measurement & Performance Feedback Loops

UDN embedded continuous measurement and optimisation into the workflow, so both sales teams and advertisers could see performance signals in near real-time and refine campaigns dynamically. (inma.org)

Real Case Outcomes — Data-Driven Growth in Advertising

Once UDN shifted from inventory-driven to intelligence-driven advertising, the results were striking:

200 % increase in click-through rates for high-value ad categories, indicating much stronger ad relevance and engagement.
45 %+ penetration across advertiser portfolios, meaning advertisers expanded use of UDN’s data-powered insights across multiple campaigns and objectives.
Advertisers started shifting from “How many did we reach?” to “Who exactly did we speak to, and why does that matter?” — a sign of value perception beyond raw impressions. (inma.org)

These metrics show that UDN didn’t just stabilise ad revenue — it redefined how value is created and sold to advertisers. (inma.org)

Comments & Strategic Perspective

Carrie Wong — Director, Data Development Department, UDN:

“Our breakthrough was not about AI as a standalone tool, but about structurally integrating data and AI across the entire advertising lifecycle… turning data into a shared language across editorial, product, and sales teams.” (inma.org)

This insight underscores a key theme: the success wasn’t just technology — it was organisational alignment around data-driven decision-making. (inma.org)

Industry Reaction & Context

UDN’s work has been highlighted in WAN-IFRA innovation reports as an example of publisher transformation through data and AI, particularly in how data insights feed both audience engagement and monetisation strategies. (WAN-IFRA)

UDN also received industry awards recognising AI-driven innovation in revenue and audience strategy — a sign that peers are noting its impact. (LinkedIn)

Key Takeaways — Why UDN’s Approach Worked

Dimension	Traditional Model	UDN’s Data-AI Model
Ad Strategy	Inventory-led, shallow segmentation	Intelligence-driven, behavioural segmentation
Targeting	Broad audiences	110+ refined actionable segments
Creative & Delivery	Static and channel-specific	AI-optimised, cross-channel
Measurement	Completion metrics	Continuous performance feedback loops

By fundamentally reframing advertising as a data-driven product rather than inventory to be sold, UDN successfully recaptured advertiser interest and delivered measurable return on ad spend that drove broader revenue engagement. (inma.org)

Final Thought

UDN’s experience illustrates how traditional media companies can innovate through AI not by simply adding tools, but by embedding data intelligence into the core of the advertising value chain — from audience understanding to creative alignment and performance optimisation. This model doesn’t just recover lost ad revenue — it positions publishers as strategic partners for advertisers seeking meaningful engagement in a crowded digital marketplace. (inma.org)