Why I love big surveys

The 2026 “AI and the Writing Profession” survey, sponsored by Bernoff.com and Gotham Ghostwriters, will be closed to new respondents in two weeks. Now would be a good time to take it if you haven’t already.

So far 1,224 writers and editors have completed it. I’d love to get to 2,000. But you have a right to ask: why does it matter how many people complete the survey? What can you learn from 2,000 people that you couldn’t learn from, say, just 1,000?

Anyone with a passing knowledge of statistics knows that the bigger the sample, the smaller the margin of error. While this is not a probability survey (that is, we have not balanced the respondents to match a particular population, as pollsters often do), we can still estimate the margin of sampling error. That’s the amount by which answers are likely to vary from a theoretical population result with a 95% probability.

Assuming a population of 250,000 writers and editors (an estimate from the US Bureau of Labor Statistics), a 1,000-person survey has a 3% margin of sampling error, and a 2,000-person survey has a 2% margin of sampling error.

Why sample size really matters

As I learned in doing many large-scale surveys at Forrester Research, the real power of large-scale surveys is the ability to examine slices of the data — especially when you’ve collected data for several years and can see trends. For example, our survey will be able to answer questions like these:

• How have writers’ and editors’ use of AI changed in the last year?
• Have individual writers and editors who use AI changed how much they use it?
• Do writers use AI more or less than editors?
• How many writers of various kinds (such as authors, thought leadership writers, marketing writers) use AI to generate final text, versus those that just use it to support their writing?
• Have freelance writers and editors seen an impact on the demand for their services and their income from AI?
• How many ghostwriters are dealing with AI-generated source material from clients?
• Which kinds of writing professionals are most optimistic and pessimistic about the future of the business?

I’ll tell you a secret. I already know the approximate answers to many of those questions, just based on the data we’ve already collected. Some of the 2026 survey statistics are going to shock people. Last year we learned that writers were generally split between AI lovers and AI haters. This year, it’s clear that a major segment of the writing community has embraced AI — and we have actual data and testimony about the challenges that is creating.

I wish I could tell you what I know . . . but anything I shared would (1) be based on preliminary data, since the survey is not yet complete, and (2) might bias the perspectives of the people who have yet to take the survey.

So I have to keep my mouth shut, for now. I’m a tease, I know. All I can say is that it’s going to surprise people.

For now, the best I can do is offer this deal.

You take the survey.

And I’ll put in an unreasonable amount of time and effort analyzing the results, and share them first with the people who participated.

Thanks for being a part of this. There is far too much uninformed opinion on this topic and not nearly enough actual data. You can help me fix that.