Overview
Notification fatigue was eroding advertiser trust in Google Ads. P0 alerts were too intrusive and vague, while lower-priority content crowded critical signals — leaving advertisers unable to quickly distinguish what required immediate action.
Led a 2-day rapid cross-functional workshop to design and stress-test a tiered Hedwig notification framework. Followed by iterative designs and qualitative research with 7 experienced Google Ads advertisers to validate the framework.
Established a priority-based notification architecture with clear visual hierarchy rules for P0–P2 alerts, actionable inline patterns, and a validated content IA — ready for implementation with a defined design system.
Workshop
Introduction
Notifications are pivotal touchpoints in user experience, yet they walk a delicate line: they must demand attention without overwhelming, convey urgency without anxiety, and empower action without friction. Poorly designed alerts risk user frustration, system overload, or missed critical actions — especially for high-priority (P0) scenarios like security breaches or payment failures.
This workshop tackled these challenges head-on, focusing on crafting a notification system that balances clarity, urgency, and user control. By aligning cross-functional teams around visual hierarchy, actionable patterns, and priority-based logic, the goal was to ensure alerts enhance — not disrupt — the advertiser journey.
Goals
Establish clear visual and functional rules for P0 alerts to ensure immediate recognition and action.
Simplify in-alert decisions (confirm/cancel) to reduce cognitive load and time-to-resolution.
Differentiate components (color, placement, interactions) across priority tiers P0–P2 while maintaining system consistency.
User Needs
Users must instantly recognize priority levels — "Is this urgent?" should never be a question that requires analysis.
Options to act (confirm/cancel) or defer within the alert, minimizing workflow disruption.
Alerts should surface only relevant, actionable information tied to where the advertiser is within Google Ads.
Avoid anxiety from overused urgency — "Why is everything red?" is a failure state.
Workshop Focus
Key Scenarios
Campaign has no budget / Campaign inactive
Need: Immediate, unambiguous error highlighting with one-click resolution. Risk: Abandonment if actions are not clear or steps are too many.
Apply keywords / Increase budget
Need: Ability to implement a recommendation directly if the cost of inaction is surfaced. Requires clear budget impact and cross-campaign visibility.
P2 Recommendation vs. P0 Critical Alert
Need: P2 uses subtle cues (icon); P0 demands prominent, persistent display. Non-urgent prompts must be dismissible without blocking workflow.
Workshop Focus: Validate these scenarios cross-functionally, stress-testing how design choices address conflicting needs — urgency vs. calm, prominence vs. non-disruptive. Reference existing user pain points: "P0 alerts are intrusive and vague."
Design Explorations
Rapid Iterations Using Material Design System
Four core design principles guided all iterations: match location to urgency (P0 always prominent, P1/P2 in the Right-Hand Rail), reserve alerts for action not education, apply tiered visual hierarchy across priority levels, and limit inline actions to one-click decisions to reduce cognitive load.
RHR Framework — Design Variants
The Right-Hand Rail (RHR) framework was designed to surface P1/P2 notifications, account recommendations, education content, and contextual content within a persistent panel — without disrupting key tasks in an already complex UI. By anchoring lower-priority content to the rail, advertisers can stay focused on their primary workflow while still having timely information readily accessible. Two variants were developed: a segmented view for clearer priority grouping and a continuous scroll view.
V1 — Segmented notification panel
V2 — Continuous notification panel
P0 Critical Alerts
P0 alerts are always prominent but must not block UI functionality. Two design directions were explored: a persistent top bar and an inline card within the RHR.
P0 — Persistent top bar
P0 — Critical card in RHR
P1 — Persistent Banners & RHR
P1 alerts use persistent banners within the RHR that avoid disrupting workflows. Rather than relying on color alone, existing iconography and system colors denote the type and impact of each issue or recommendation. The RHR variant shown explores a full detail panel for surfacing actionable context without leaving the current workflow.
P1 — Full RHR panel
Inline Actions
Inline actions limit decisions to 1–2 primary actions ("Confirm," "Undo") with an optional "Learn More" link. Verb-driven labels replace generic terms. Complex inputs redirect to a task-specific page rather than attempting to fit into the alert.
Inline applying — design framework
Content Architecture & Page Mapping
A key challenge: the proposed framework showed global content that didn't change based on where the advertiser was in Google Ads. No design cues indicated the RHR content was tied to the current page — a critical gap for contextual relevance.
Design mapping — notification-to-page context alignment
Research
Validating the Framework with Advertisers
Following the workshop and iterative design phase, qualitative research was conducted with 7 external Google Ads advertisers to validate the proposed Hedwig framework. Three focus areas: Hedwig Identity (content organization and prioritization), High Priority P0 Alerts (prominence, persistence, and actionable steps), and In-line Recommendations (contextual application preferences).
Methodology
45-minute individual contextual interviews, conducted remotely via GVC video conference with screen recording (QuickTime).
Static design mocks across two primary sections, each with two variants. Common design elements shown across sections and consolidated for analysis.
n=7 external Google Ads users (LCS & GSM). Business areas: Education, Creative Studio, Health & Beauty, Hotels & Travel. Experience: 3–15 years. Monthly spend: $3,000–$1M. Scope: single city/region accounts up to 700+ client global managers.
V1: Segmented notification panel + education panel with dedicated resolution page. V2: Continuous notification panel + education panel to new page. Feedback across both variants was consolidated and analyzed.
Research Focus Areas
Content Organization & Prioritization
User needs assessment to identify advertiser content requirements. Analyze optimal IA strategies (prioritization, categorization, segmentation) aligned with advertiser mental models.
Visibility, Persistence & Actionability
Investigate expectations and contextual requirements for balancing alert visibility (visual hierarchy/salience) with duration (persistence/dismissal) — identifying thresholds where effectiveness meets minimal cognitive disruption.
Mental Models & Workflow Fit
Map advertiser mental models against recommendation implementation patterns to identify expectation gaps, trust calibration factors, and actionability friction.
Results
Key Findings
High Priority Alerts (P0)
Advertisers want high priority alerts to be prominent both outside and inside the proposed RHR framework. Alerts should be visible at-a-glance regardless of the advertiser's current location within Google Ads.
Advertisers generally wanted P0 alerts to be visually persistent and undismissable — while not interfering with their ability to navigate Google Ads. Alerts should remain readily accessible in the RHR.
Advertisers want P0 alerts to link directly to content ("Go to page…") with clear steps on how to resolve the alert. Note: actual page content is out of scope for this research, but findings direct what content should exist.
Hedwig Identity
Overall, Education content within the proposed Hedwig framework confused advertisers. This ties into the relationship between organization and actionability — Education was not seen as high priority vs. content that was hidden.
Across both design variants, advertisers frequently mentioned the need for content within Hedwig to be organized based on priority. Pending Requests were prioritized above recommendations as a secondary example of this principle.
The proposed framework suggested global content that did not change based on where advertisers were within Google Ads. When navigating to "Keywords," there were no design cues to suggest the RHR content was tied to the current page.
In-line Applying
Actionability is context-dependent. Advertisers liked the ability to apply a recommendation directly within Hedwig — but this depended on the type of recommendation. Low-risk recommendations (perceived as simple or low-cost) were applied inline. High-risk recommendations (complex or high-cost, e.g., increase budget) prompted advertisers to navigate to a dedicated page before deciding. This aligns directly with the P0 priority findings.
Next Steps
Recommendations
Prominence + Clear CTAs
Ensure P0 alerts are visually prominent while providing advertisers clear calls to action that meet expectations. Validate persistence behavior with interactive prototypes rather than static mocks.
Determine What Advertisers Want
Validate what content advertisers want within Hedwig and how they prioritize it. Ideate on visual organization focusing on glanceability and actionability — avoid hiding high-priority notifications behind lower-priority content.
Strengthen Page-Context Signals
Visually strengthen the mapping of notifications shown to current page content. Balance organization of content based on previous findings — the RHR should reflect where advertisers are, not a global view.
Complexity-Based Apply Patterns
Ensure inline applying aligns with advertiser expectations based on perceived complexity and cost. Surface notifications with higher probability of inline acceptance as "quick wins" to facilitate higher adoption.