Frontiers Media | Character building: personas & profiles

About:

Frontiers Media is the world’s 3rd most cited scientific publisher. With over 2.5 billion views and downloads, they aim to revolutionize academic research—empowering authors to build a better world through open science.

Frontiers needed:

A shared, evidence-based picture of who researchers actually are—so that customer-facing teams stopped writing for a generic “author” and started writing for real people at specific points in their careers.

Work included:
Defining three career-stage personas—Alvin (early), Sofia (mid), and Theodora (late)—as fictionalized researcher profiles built on observed behaviors, not assumptions.

Designing a multi-indicator segmentation model (h-index, years since PhD, institutional position) that triangulates career stage rather than relying on any single metric, with documented limitations for each indicator.



Producing a practical usage framework: a five-step method for identifying a researcher’s persona and tailoring messaging to their pain points, plus guidance on w here personas apply and where they don’t.

Writing guardrails into the resource itself—personas as “navigational aids, not straitjackets”—to prevent teams from flattening a diverse academic audience into three boxes.

Building a GPT-based persona agent that lets internal users roleplay conversations with Alvin, Sofia, or Theodora—stress-testing outreach emails before sending, rehearsing objection-handling for sales and partnerships calls, practicing onboarding scenarios for new hires in author-facing roles, prepping for editorial board recruitment conversations, and pressure-testing Research Topic invitations against a skeptical late-career reviewer’s likely reaction.

Partnering with marketing, sales, and editorial stakeholders to pressure-test the profiles against their real conversations with researchers.



Outcomes:
Gave customer-facing teams a common vocabulary for talking about researchers, replacing ad-hoc assumptions with a shared reference point.

Messaging shifted from feature-led to need-led, with teams able to name a specific pain point before drafting outreach.

The GPT-based persona agent became one of the most-used internal GPT agents at Frontiers, ranking in the top 10 by conversation volume, with a 58% repeat usage rate.

Adopted by 43% of customer-facing staff monthly, with unplanned but strong uptake in product and engineering teams, who use it to sense-check new platform features against researcher needs at different career stages.
The indicator table became a reusable qualification tool, letting teams place a researcher in seconds from publicly available data.

Established a foundation for persona-led planning across marketing, sales, and now product—well beyond the original author-communications remit.