GWI · UX Research Case
0. GWI
While at GWI in London, I worked as a Senior Product Designer on the Integrations team, developing connections between GWI and advertising platforms including Google Ads, Meta Ads, and TikTok Ads. GWI is a market research company built on survey-based data. In this case study, I'll focus on the User Research Pilot Project we implemented.I kicked off with an introductory workshop for the new team. This provided an excellent opportunity to assess our collective understanding of our users and their workflows.
We identified key roles and named them for the first time.
1. Context & Challenge
Users were increasingly relying on GWI for insights and audience definition, while simultaneously managing campaigns on platforms like Facebook and Google Ads. However, the integration between insights and activation was fragmented and inconsistent:- Media planners (Researchers) used GWI to define audiences.
- Activators rebuilt those audiences manually on Meta/Google.
- Communication gaps, misinterpretations, and platform limitations resulted in inefficient targeting and wasted budget.
We needed to answer a critical question:
I mapped the current user flows together with the devs and saw more gaps and inconsistencies.
The question became even broader:
What is the user problem we’re solving with our tool?
The existing tool was simply pushing audiences directly from GWI into Facebook with minimal attribute mapping capabilities, adding no extra value to the user, but rather an extra layer of complexity.
Solely design case focused on platform usability challenges:
2. Research Set Up and goals
Since UX Research wasn't established as a discipline at the company, we decided to develop and test the process by addressing our domain's specific need for research insights. We handled legal paperwork, created knowledge base documents and templates in Google Drive, and established a structured planning process. By We I mean me + Head of UR.Our goals included:
- Understand core workflow;
- Uncover frictions in the workflow across strategy and activation;
- Evaluate if the integration solves real user pains;
- What is the main user problem/need in reality.
3. UR Methods
- Contextual UR with in-depth interviews with planners, strategists, and activators conducted Worldwide (from South Africa to U.S and Australia);
- Observation of campaign setup workflows (Smartly, Ads Manager);
- Usability testing of the existing product;
- Synthesis of artefacts: briefs, audience spreadsheets, mock campaigns, GWI dashboards;
- Pain-point mapping across journey from insight to launch.
See Figma file
4. Key Insights
1. Translation gap: Audience definitions from GWI rarely translate 1:1 to ad platforms. Activators often ignore or reinterpret segments due to technical constraints.
2. Trust breakdown: Strategists don’t trust activators to "get it right"; activators don’t have access to GWI and rely on vague summaries.
3. Manual inefficiency: Audience setup is often done manually in Facebook/Google Ads using Saved Audiences or shell campaigns. It's slow, error-prone, and subjective.
4. Data credibility vs. actionability: GWI data helps win arguments with clients, but doesn’t always help run better campaigns. Activation teams optimize based on performance data, not insights.
5. UI changes & taxonomy removals on Meta make audience targeting unstable. (E.g. "iPhone" interest removed; healthcare targeting deprecated.)
I created UJM and Persona artefacts based on our knowledge:
So it was no longer a whiteboard with a few stickers on it, but a proper document used for decision making.
5. Strategic Outcome
We reframed the problem: this is not an integration issue, it’s a broken workflow issue.Instead of refining the existing approach, we proposed a 180-degree pivot:
- Build a prototype workflow where GWI data flows directly into platform audiences via exportable segments.
- Bring activators into the research process and co-create audience definitions (Seat Growth Opportunity).
- Focus on communication and translation gaps.
The core problem we identified was that researchers and activators approached audience definition differently. This created a translation gap that reduced the effectiveness of campaigns. We called it the “Lost in translation” case.
This reframing based on heavy research convinced leadership to approve a pivot.
6. Impact
- Clear visibility into cross-functional pain points
- Executive alignment around the need for pivot
- Budget and resources allocated to explore structured integration (via Smartly, clean handoff docs, or platform partnerships)
- Established shared language between planners and activators