2 builders who never spoke to each other

Run the survey nobody in your field has run

When peers keep asking a question the existing data cannot answer, the answer is a study rather than a blog post: one founder built an anonymised benchmark of agency pay, tenure, retention and pricing because the salary sites bucketed too broadly to compare against. Keep it to something people will actually finish, around thirty five mostly multiple choice questions inside fifteen minutes. Promising up front that the pooled results will be published openly, your own included, is what gives contributors a reason to fill it in.

The method

1. 01

   Run the survey and publish the benchmark when peers keep asking you a question the existing data cannot answer.

2. 02

   Cap a research survey at something people will actually finish, and lean on multiple choice to get there.

3. 03

   Promise before you collect anything that the pooled results, your own included, will be published openly under a permissive license for anyone to remix, so contributors have a reason to take part.

What it returned

• 35

  Rand Fishkin's agency benchmark ran to 35 questions, mostly multiple choice, aimed at under 15 minutes, targeting agencies of 5 to 250 staff, collecting for a couple of weeks and publishing averages and distributions per country wherever there was critical mass.

• An agency founder pressed Rand Fishkin on entry-level pay, staff tenure, client retention and pricing, and found the salary sites too broadly bucketed to compare against other search agencies, so he built a survey to anonymise and aggregate the answers from agency owners.

• Hillman committed to releasing every submitted result plus Indy Hall's own under Creative Commons for reuse in other coworking material.

Sources

• Coworking Churn RatesAlex Hillman · dangerouslyawesome.com · Then what? · 2010-01-01
• Making Agency Numbers More Transparent - SparkToroRand Fishkin · sparktoro.com · undated