EntryWorks
PlatformIdeapadSecurityIntegrationsWorkGuidesAboutStart a conversation

Case study · Photo competition

One city. One season. Three thousand photographs.

A capital-city photo competition, run for a household-name sponsor inside the city’s metropolitan masthead. Names withheld — the numbers and the machinery are the story.

  • ~3,000photographs entered
  • 1,500+individual entrants
  • 16,800unique visitors
  • 700entries shared on

The brief

A sponsor, a masthead, and a city with cameras

A major utility wanted its brand wrapped around something the city genuinely cared about; the metropolitan masthead wanted a campaign its readers would join, not just read. The answer was a photo competition: readers submit the city as they see it, the public votes, the sponsor judges the major prize. EntryWorks ran everything between the idea and the winners' announcement.

The build

Two design languages, one engine

The campaign shipped inside the masthead's own header and footer, pixel-matched so entrants never felt they'd left the paper. Underneath, two complete design languages — a warm, glassy look and a crisp editorial one — ran on the same components, switchable per campaign from the console. Copy, prizes, dates and eligibility were all configuration, edited live by the team without a developer in the loop.

The gates at work

Three thousand entries, zero chaos

Every photograph was screened on arrival — provenance credentials, camera-metadata forensics, trained detectors scoring authenticity and content safety — and handed to a moderator with the evidence attached. People made every publish call; the system just made those calls fast and defensible. Location data was stripped from every public image, faces without a signed release were flagged before judging, and each decision landed in the audit log.

The numbers

Reporting the partners could invoice against

Every media placement — display, email, video — was tracked to what it measurably delivered: arrivals, entries, votes, day by day. Measured results were never blended with inferred ones, small samples showed a dash instead of a misleading rate, and the campaign export, the daily table and the overview all reconciled to the same totals. The sponsor's own staff were excluded from client-facing figures automatically, so the engagement they read was genuinely the public's.