Original research means publishing findings nobody else has, because they come from your own data collection. Common sources for SaaS companies are surveys of a target audience, anonymized and aggregated product data, and structured experiments. Because the findings are new, other writers have a reason to cite and link to them, which makes research one of the more reliable ways to earn links and press coverage.
Sources of data
Each source has strengths and limits worth stating openly in the report.
- Surveys: flexible, but results depend on who responds and how questions are worded.
- Product data: large and behavioral, but limited to your own customers and needs careful anonymization.
- Public data analysis: reproducible, but others can do it too.
- Experiments and tests: controlled, but usually narrow.
- Interviews: rich detail, small samples.
Doing it credibly
Publish the method: sample size, how respondents were recruited, when data was collected and what the limitations are. Avoid claiming more than the data supports, for example presenting a survey of your own users as representative of a whole industry.
Check privacy obligations before using customer data. Aggregate it, remove identifying details and make sure your terms allow the use.
Getting value from one study
A single dataset can feed a report, a landing page with key findings, charts for social, a webinar, press pitches and sales material. Pages that state specific findings plainly, with clear headings and dates, are also easier for writers and AI systems to quote accurately.
Example
Turning usage data into a benchmark report
Hypothetical company Driftcairn runs a customer support platform. It analyzes anonymized, aggregated response-time data across its customers, segmented by company size, and publishes an annual report with the method on the first page. Journalists writing about support teams cite the report, and the team turns the findings into a webinar and a set of comparison charts.
Common mistakes
Hiding the method
Readers and journalists distrust figures without a method. State how the data was gathered and its limits.
Leading questions
Surveys designed to produce a marketing message produce findings nobody trusts. Write neutral questions and publish what you find.
Questions
It depends on what you want to claim and how many segments you break results into. A statistician or survey specialist can help. Whatever the size, report it.
