Open
Influence
The analytics were excellent — and trapped in the shape of a spreadsheet.
Open Influence’s Go Prism platform had genuinely strong influencer data. My remit was the visual language for its reports, dashboards, and decks — so a non-technical brand manager could read them in about three seconds, not decode them.
Good data a decision-maker couldn’t read.
The numbers were right; the presentation asked too much. Rows and columns are fine for an analyst and hostile to a brand manager who needs a verdict now. The design work wasn’t new metrics — it was a visual language that let the existing data land instantly.
Design the visual language of the data.
Complex social-media data sets, made to read clearly to a non-technical brand audience — three requirements stacked on top of each other.
Accurate & readable
Individual charts, graphs, and infographics that were correct and immediately legible — no data-analyst decoder required.
One product, not many
A consistent visual system so a report looked like Go Prism whether it was a brand-safety audit or a campaign recap.
Holds up on stage
The same data had to work in a live client deck — presentation-ready output, not just an in-app view.
The whole Go Prism surface.
Three connected surfaces, one visual grammar: the emphasized headline number, the tidy metric row, the card as the unit of a story.
Creator discovery
A creator-discovery interface with a deep filter rail and a grid of scannable creator cards.
Brand Safety report
A Brand Safety report that compresses “how risky is this creator” into a three-second read.
Data-viz reports
Data-visualization screens that tell a campaign’s results back to the client.
Lead with the verdict. Keep the proof one glance away.
The clearest example was the Brand Safety report. Instead of a wall of flagged posts, it opens with a single Overall Risk verdict — the thing the brand manager actually came for — with the creator-post evidence cards sitting beneath it for anyone who wants to check the work.
Resolve the tension between clarity and fidelity with hierarchy, not omission — nothing removed, everything ranked.
The verdict a brand manager came for, first.
Brand Safety was one of the biggest pieces of the work. The shipped report opens on a single Overall Risk reading and a ranked category breakdown — alcohol, profanity, offensive language — so the call is legible in one glance, across every platform a creator posts on.
The flagged-post evidence then sits one scroll beneath: each post with its tags, follower count, and engagement, so a brand manager can trust the verdict and check the work — the answer-first principle from the wireframe, shipped.
Brand Safety Results
Five equal-weight metrics became three scannable ones.
The old creator card gave five numbers equal visual weight — which is the same as giving none of them weight. The redesign cut to three scannable metrics, added a Brand Safe badge, and a one-tap link straight to the full safety report. Same underlying data, a tenth of the effort to read.
The report screens carried the language to the storytelling end.
The same grammar ran out to the campaign reports — the screens that tell a client how their spend performed. Every number found its place in the reading order rather than landing as one more cell in a grid.
Translation, not invention.
The data already existed. My job was to find its right visual form — which made stakeholder collaboration the core of the process, not a step in it. I sat with the people who owned the data and the client relationships to learn what each number meant, and what decision it was meant to drive. Then I ranked it.
What brands argue about — given the most weight.
The shape of the trend, one glance in.
Quiet, present, available when wanted.
Off the page entirely.
A structured read of 15+ platforms.
The visual language wasn’t taste — it was argued. I ran a structured competitive analysis, scored against a psychological-usability heuristics framework, so every choice could point to a reason rather than a preference. Paired with a features-and-pricing teardown and an A/B testing plan, so genuinely contested layouts got settled by evidence, not argument.
Hierarchy
What the eye lands on first — traced to how people actually read a dashboard, not to layout habit.
Chart type
Each encoding chosen for the question it answers, not for variety.
Density
How much fits before a glance turns into a study — set by a legibility threshold, not by taste.
Clarity versus fidelity.
Social data is genuinely multi-dimensional — reach, authenticity, engagement, sentiment, risk, all at once. Stay too faithful and you rebuild the spreadsheet the customer came to escape.
Push too far toward simple and you flatter the data into saying something it doesn’t. The useful presentation is simple — but simplicity can lie.
One product, and a package to keep going.
Redesigned surface
Discovery page & card, Brand Safety report & detail views, campaign data-viz screens.
Visual style doc
A documented style system so every report and deck read as one product.
Competitive analysis
15+ platforms scored, teardowns, three prioritized reports + buyer personas.
A/B plan & package
A testing plan and handoff so the team could keep iterating after I left.
The through-line: complex social-media data sets a non-technical brand stakeholder could read at a glance — with the Brand Safety report as the representative example.
The visualization is the product.For a data company, information design isn’t decoration — it’s where the intelligence either lands or it doesn’t.
A good chart is downstream of a good question.The fastest route wasn’t studying chart types — it was answering “what is the brand actually trying to decide here?”
When the data is the product, legibility is the feature.
The work delivered a report-and-dashboard visual system that put the answer first and the evidence underneath — turning excellent-but-unreadable analytics into something a brand manager could act on at a glance.
No KPI-movement or adoption numbers were captured in the source, so I don’t quote any.