Interactive Analysis
You Put In the Work.
Here's What the Data Shows.
4,212 posts. 3,303 creators. Three editions with $300K in prizes. We scraped every submission and ran the numbers so you can see exactly what separated the winners from everyone else.
How big was this, really?
Most people have no idea how many others they were up against. Before you ask why you didn't win, start here: the field was enormous and the community was real.
Getting fewer than 10 likes? So did most people.
Before you judge your own submission by its like count, see what everyone else got. Engagement was heavily skewed. A handful of posts took most of the attention. The chart below shows what percentage of posts fell into each likes tier, per edition.
Color = edition.Sep 2025Mar 2026Config 2026
Winners didn't just build better. They posted differently.
Across all three editions, winning posts averaged 7.7 times more likes than non-winning posts. That gap isn't about luck. The scatter below shows every post in the contest window.Green stars are winners. Gray dots are everyone else. Hover any point for details.
WinnersNon-Winners
Did posting earlier give winners an edge?
In Sep 2025's 8-day sprint, winners posted at day 5.4 on average versus day 6.2 for non-winners. That half-day head start let engagement compound before the window closed. In longer editions (Mar 2026: 44 days, Config 2026: 15 days), timing was nearly identical between winners and non-winners. Early posting only matters when the window is short.
The wide range (217 vs 23) shows that high engagement correlates with winning, but is not a requirement. The Config 2026 grand prize went to the best interaction design, not the most-liked post.
Should you have posted a video? Almost certainly yes.
The single most consistent difference between winners and non-winners across all three editions was whether the post included a video demo. A working video proves the interaction is real. It reduces skepticism. It gets shared.
Winners with VideoNon-Winners with Video
In Sep 2025 and Mar 2026, cross-posting made a small difference. By Config 2026, it jumped: winners averaged 1.2 social platforms versus0.4 for non-winners. That is not a coincidence. Config 2026 added a "Build In Public" award category that explicitly rewarded public documentation of your process across platforms. If you are entering the next contest, check whether a "Build In Public" category exists before deciding how much to share.
Were the judges watching your posts? We checked.
One of the most common questions after not placing: did the judges even see my work? We loaded the full comment threads for three winner posts to find out.
We checked three winner posts directly: Cara Ellis's Web Poetry (Sep 2025, $50K, 217 likes), Lee Black's Airwwave (March 2026, Boundary Pushing, 222 likes), and Bene Brandhofer's Visual Atlas (Config 2026, $50K Grand Prize, 23 likes). Not a single Contra staff member or judge commented on any of them before the winners were announced.
This means you cannot reverse-engineer who will win by watching for judge engagement. It also means the absence of a judge comment on your post tells you nothing. Judging happens entirely off-platform. The winners were selected based on project quality reviewed externally, not on how much attention their Contra posts received from the Contra team.
What did Contra staff actually post?
Staff posts were entirely broadcasts: kick-off announcements, judging criteria clarifications, and winner reveals. Gui Seiz's March 2026 winner reveal reached 91,802 impressions, making it the most-viewed post across all three editions. Staff posts generated conversation (comments, replies) but were not engagement with individual submissions.
Staff includes: Gui Seiz (Contra CEO), Jan Mráz, Galaxia Wu, Sabrina Polanco, Daniela Muntyan. Staff posts are announcements, winner reveals, and community updates — not submissions. Reply/comment interactions are not shown (require post-level API calls).
Did winners actually read the judging criteria?
Each edition had published judging criteria. We looked at whether winner posts reflected those criteria in their language and project framing, and whether the criteria evolved between editions in ways that changed what it takes to win.
Engagement score vs posting day, across all editions
Green stars are winner posts. The cluster of high-engagement winners above the cloud shows that engagement and judging quality are correlated, but they are not the same thing. Several non-winners in the top-right (high engagement, late posting) did not place.
Every edition rewarded posts that clearly demonstrated a working interaction, explained what was novel about the approach, and showed that the creator understood what the contest was asking for. Winners who had high engagement also tended to have posts that were clear, shareable, and community-forward. Winners who had low engagement (like Halle Clark and Bene Brandhofer) were category-specific and criteria-precise. The through line: match your submission to the specific category you are entering, not just to the general idea of the contest.
Additional Charts
Supplementary comparisons for deeper exploration.
Median Metrics: Winners vs Non-Winners
Median values for posting day, engagement score, and social platform count.