UX/UI - Machine Learning - Artificial Intelligence
WATSON ADS BUILDER
Designing AI-Assisted Creative Workflows for Enterprise Advertisers
Overview
Enterprise advertisers wanted to harness the power of AI to create richer consumer conversations, but existing tools were either too technical or too disconnected from everyday marketing workflows. Watson Ads Builder was IBM’s answer: a self-service platform empowering creative teams, media planners, and developers to craft and deploy AI-assisted conversational advertising experiences across digital campaigns.
This wasn’t just another ad creation tool; it was a next-generation experience layer that made sophisticated AI accessible, usable, and strategically controllable for marketing professionals.
TEAMMATES
Georgios Saliaris, January Holmes, Robert Redmond
The Challenge
AI promised speed, scale, and personalization — but it also introduced complexity:
- Conversational AI systems were typically built by specialists.
- Traditional ad workflows weren’t designed for AI generation.
- Marketers lacked transparency into how AI interpreted language and generated responses.
- There was little control over tone, brand alignment, or conversational logic.
Advertisers needed a platform that enabled:
- Rapid creation of conversational ads
- Structured human-AI collaboration
- Transparency and trust in automated content
- Brand-aligned conversational outcomes
In short: How might we design a workflow that lets marketers create meaningful, AI-assisted conversations with consumers, without needing deep technical expertise?
My Role
As UX Strategy & Interaction Designer on the Watson Ads Builder team, I focused on:
- Understanding how marketers define goals and expectations for conversational campaigns
- Designing structured input flows that guide users through complex AI configuration
- Translating AI capabilities into clear, controllable UI interactions
- Balancing automation with user control to build trust and confidence
- Defining interaction patterns that supported iterative refinement and transparency
I worked closely with product managers, AI research engineers, and front-end developers to ensure the interface aligned with real user mental models and business needs.
Research & Insights
Enterprise users often shared consistent frustrations:
- Unpredictable AI outputs — without context, outputs felt random or off-brand.
- Technical barriers — set-up often required engineers or specialists.
- Limited control — little ability to guide or constrain AI conversation generation.
- Disjointed workflows — campaign setup, content, and outcomes weren’t integrated.
By leveraging IBM’s Natural Language Understanding (NLU) capabilities and conversational insights, we identified opportunities to:
- Guide users step-by-step
- Expose relevant AI behavior clearly
- Build guardrails and preview mechanisms
- Surface editable, human-reviewable output
- Support domain-specific conversational frameworks
These insights shaped an experience that bridged technical depth with human usability.
Design Strategy & Solutions
Rather than present users with raw AI control panels, we designed structured workflows that included:
Guided Setup Flows
Users were led through context, goal definitions, audience intent, and messaging constraints, reducing cognitive load and increasing clarity.
Transparent Output Preview
Before publishing, users could preview how AI interpreted inputs, allowing them to refine or override outputs.
Human-in-the-Loop Controls
Marketers retained control over brand voice, response structure, and conditional behaviors, avoiding “AI magic” and instead empowering an AI partnership.
Reusable Conversational Templates
Modular kits and pre-configured conversational paths accelerated setup while still allowing customization.
Iterative Refinement Tools
Built-in iteration loops enabled refinement based on live performance and consumer interaction signals.
These patterns helped surface what today’s “AI-assisted” UX needs to succeed: clarity, control, and confidence.
Impact & Recognition
Watson Ads Builder was recognized as a pioneering tool that simplified the creation of conversational experiences across digital advertising channels and helped shift how brands think about AI participation in marketing. The platform enabled:
- Faster campaign ideation and iteration
- Greater adoption of AI-driven workflows across teams
- Richer consumer engagement through conversational touchpoints
- Easier integration of natural language insights into real advertising contexts
Industry coverage highlighted the innovation around AI-assisted advertising and conversational campaigns, showing broad interest in how technology can reshape advertiser–consumer interaction.
