In This Article
Ask who was counted
The OECD's Generative AI and the SME Workforce report is a useful example of why scope belongs beside a statistic. Published in November 2025, it draws on a 2024 survey of more than 5,000 SMEs in Austria, Canada, Germany, Ireland, Japan, Korea and the United Kingdom. Its findings should not be relabelled as Middle East adoption data.
Before using a percentage in a presentation or article, identify the population behind it. Is it all respondents, users of a service or a smaller group that answered a particular question? The answer can change the meaning even when the number has been copied perfectly.
Keep the condition attached
The OECD reports that 65% of surveyed SMEs using generative AI said it helped increase employee performance. Removing the condition about use would change the claim. So would describing a reported perception as an independently measured productivity gain.
Write the condition into the sentence rather than hiding it in a source note. If the sentence becomes too long, simplify the presentation without dropping the population. A chart title can state the subject, while a subtitle supplies the sample, geography and period.
Separate evidence from interpretation
A survey can show what respondents report. It may not establish why the result occurred or what another company should expect. Label your interpretation and explain the additional evidence that would be needed to test it.
When comparing two surveys, check the question wording, sample, timing and method before describing a change. A difference between unlike groups is not automatically a trend. Avoid building a league table from numbers that answer different questions simply because they share a percent sign.
Build a small checking record
Keep the exact source, publication date, data period, population and wording used in your material. Record any calculation you made and distinguish it from a number supplied by the publisher. If a denominator remains unclear, ask the source or leave the figure out rather than guessing.
This is a proposed editorial checklist, not a statistical certification. Its benefit is modest: someone reviewing the slide can recover the claim and understand its boundary. A less dramatic sentence with the right denominator is more useful than an impressive percentage that the evidence does not support.
Sources and scope
This is an editorial working method, not legal or tax advice. The linked sources supply the context; the suggested checklists are our analysis. International evidence is not presented as Middle East survey data or universal law. Confirm requirements for the specific entity with qualified advisers where relevant.
