Second-Cycle Contributors Are the Scarce Resource

In sustained knowledge communities, the bottleneck is not first-cycle contribution but second-cycle continuation. Roughly one in four first-cycle contributors continues to a second cycle, and the attrition at that transition is consistently high across reported cases. The first cycle produces volume; the second cycle produces sustainability, and the scarcity of the latter is what determines whether a community sustains.

Grounds

The attrition claim sits inside a broader power-law distributional finding that supplies its context. Sustained online communities consistently show power-law distributions of contribution — a small minority produces most content, a larger minority contributes occasionally, the majority lurks. Multiple measured sources converge on the shape: Jakob Nielsen's 2006 Participation Inequality article formalized the pattern as "90-9-1" (approximately 90% lurk, 9% contribute occasionally, 1% contribute most) as a rule-of-thumb drawn across discussion forums and early social-web platforms. Christopher Allen's 2008 Power Laws post refined the ratio for many online communities to a 99-and-1 pattern after observing Nielsen's 9% tier collapses toward the 1% tier in communities with higher contribution thresholds. Wikipedia's own measured numbers are tighter still: approximately 99.8% read-only, 0.2% occasional editors, and ~0.003% heavy editors at the long tail. Absolute ratios diverge from platform to platform by more than an order of magnitude; the power-law shape is the durable finding, the specific ratios are community-specific.

Two reconstructed records then supply the specific attrition claim this Observation carries inside that distribution.

A recurring observation across sustained plural-contributor knowledge communities holds that approximately five of every twenty first-cycle contributors continue to a second cycle — a ~25% continuation rate, with most of the attrition concentrated at the first-to-second transition rather than across subsequent cycles. The observation recurs across multiple communities of comparable scale; it has served as a reality check against more optimistic sustainability narratives. The record's limit is that it is a summary across participant experience rather than a controlled measurement of any single community.

A four-tier cascade reported for a large-scale online community — approximately 300,000 members producing approximately 30,000 active users producing approximately 3,000 regular contributors producing approximately 30 sustained long-term contributors — was developed across years of direct community observation and is summarized in Allen's 2008 Power Laws post. Each tier is approximately 10% of the one above it; four tiers with a consistent ~10× drop-off is the specific shape power-law distributions take when measured with enough dynamic range. The cascade's sharpest tier transition by absolute attrition is the 3,000-to-30 step (the second-to-long-term-contributor transition); the sharpest tier transition by contributor effort is the first-to-second transition, where contributors who produced once must decide whether the community is worth returning to. The two transitions sit on top of each other in the cascade and are distinct processes: the 30:1 compound scarcity across tiers reflects the structural distribution; the first-to-second drop-off reflects the specific decision point where a contributor's initial experience becomes a returning pattern or does not.

The record's limits are substantive. Both attrition-specific sources are anecdotal across participants' direct experience; no controlled study is available. The ~25% figure is approximate and community-specific; other communities may show substantially different continuation rates. A sustained community of comparable scale showing genuinely flat distribution — where the top 10% of contributors produce only approximately 10% of the content — would revise the broader power-law claim; a methodological critique that re-interprets observed power laws as measurement artifacts would revise the distributional finding at the interpretive layer. The direction of both the distributional shape and the first-to-second-cycle attrition is the durable finding across the reviewed cases. The specific ratios are not.

What Would Revise It

A sustained community reaching a second cycle with more than 50% of first-cycle contributors continuing would revise the claim's load-bearing direction. The revision would be strongest if the community were comparable in scale, domain, and platform friction to the cases currently on record; a small-scale or highly-filtered community that retains most first-cycle contributors does not revise the general pattern.

A measured case where first-cycle attrition is higher than second-cycle attrition — where more contributors drop off during their initial engagement than at the return decision — would also revise the claim. The current record places the first-to-second-cycle transition as the load-bearing attrition point; a counter-case would need to show first-cycle attrition dominating at a comparable community's scale.

A community adopting these conventions will produce data in this record's tail. If the community retains more than 50% of first-cycle contributors to a second cycle, the Observation's claim weakens as a general pattern. If it retains less than 25% — or fails to produce a measurable second cycle — the Observation's claim is corroborated at another data point.

Sources

Relations