of content professionals decided against citing research because they could not verify how it was conducted
What Makes Original Research Worth Linking To
70.5% of content professionals have dropped a study in the past six months because they could not verify how it was conducted. A survey of 863 people who write, edit or commission online content.
Original research is one of the most reliable ways to earn links, coverage and citations. But not every study gets used. To find out what separates the research people cite from the research they ignore, Cherry Data Signals surveyed 863 people who write, edit or commission online content for work, including marketers, journalists and agency professionals.
The short version: writers cite research constantly, they check it carefully, and they drop it quickly when they cannot see how it was done.
Key findings
- 82% cited or linked to original research at least three times in the past six months (see the statistic)
- 70.5% decided against citing research at least once because they could not verify how it was conducted (see the statistic)
- 65% say who conducted or commissioned a survey is one of the details that matter most (see the statistic)
- 38% are most likely to reference a surprising finding that challenges an assumption (see the statistic)
- 35% say a missing methodology is the issue most likely to stop them using a report (see the statistic)
- 44% find a webpage with findings and methodology the easiest format to use (see the statistic)
Citing research is routine work
More than four in five respondents (82%) had linked to or cited original research at least three times in the past six months, and 21% had done so more than ten times. Only 5.5% had not cited any.
| In the past six months, how often have you cited original research? | % |
|---|---|
| Never | 5.5% |
| Once or twice | 14% |
| Three to five times | 35% |
| Six to ten times | 26% |
| More than ten times | 21% |
| Not sure | 4% |
The main reason is proof. 41% include research mainly to support a claim with evidence, and a further 25% use it to add authority. Fewer (18.5%) use research to find a new story or angle.
| Main reason for including research | % |
|---|---|
| Supporting a claim with evidence | 41% |
| Adding authority to the content | 25% |
| Finding a new story or angle | 18.5% |
| Making a complex topic easier to understand | 11.5% |
| Giving readers practical benchmarks | 6% |
| Meeting a client or editor's requirements | 3% |
| I do not include research | 3% |
Surprise beats confirmation
Asked which type of finding they would be most likely to reference, 38% chose a surprising finding that challenges an assumption, well ahead of trends (22.5%) and benchmarks (16%).
| Type of finding most likely to be referenced | % |
|---|---|
| A surprising finding that challenges an assumption | 38% |
| A trend showing how behavior is changing | 22.5% |
| A benchmark showing what is typical | 16% |
| A breakdown relevant to a specific audience | 13% |
| A geographic comparison or ranking | 8.5% |
| A finding that helps readers make a decision | 5.5% |
| None of these | 2% |
This is why the questionnaire matters as much as the sample. A survey designed only to confirm what a brand already believes rarely produces the finding that gets picked up.
Writers check the source, the sample and the date
When deciding whether to cite a survey, respondents could choose up to three details that matter most. Who ran it (65%), how many people responded (60%) and who the respondents were (55%) came out on top.
| Details that matter most before citing a survey (up to three) | % |
|---|---|
| Who conducted or commissioned it | 65% |
| How many people responded | 60% |
| Who the respondents were | 55% |
| When the survey was conducted | 40% |
| How the questions were worded | 30% |
| How respondents were recruited | 25% |
| Whether the data was weighted | 15% |
| None of these | 2% |
Freshness counts too. Nearly half (48%) would generally cite a consumer survey up to six to twelve months old as evidence of current behavior, and only 7% would cite one more than two years old.
| Oldest consumer survey you would cite as current | % |
|---|---|
| Less than six months old | 11% |
| Six to twelve months old | 48% |
| More than one year, but no more than two | 24% |
| More than two years, but no more than three | 5% |
| More than three years old | 2% |
| It depends entirely on the topic | 12.5% |
| Not sure | 1% |
A missing methodology is the biggest deal-breaker
The issue most likely to stop someone using a report is a missing methodology (35%), followed by leading questions (22%) and an unclear respondent group (17%). 16.5% are put off by reports that are too promotional.
| Issue most likely to stop you using a report | % |
|---|---|
| The methodology is missing | 35% |
| The questions appear leading | 22% |
| The respondent group is unclear | 17% |
| The report is too promotional | 16.5% |
| The headline overstates the findings | 11% |
| The findings are too similar to existing research | 3.5% |
| Nothing in particular | 2% |
This is not a theoretical concern. 70.5% had decided against citing research in the past six months because they could not verify how it was conducted, and 45% had done so more than once.
| Decided against citing research you could not verify (past six months) | % |
|---|---|
| Yes, once | 25.5% |
| Yes, more than once | 45% |
| No | 22% |
| I have not considered citing research during this period | 6% |
| Not sure | 7% |
A webpage beats a PDF, and a press release comes last
44% say a webpage with findings and methodology is the easiest format to use, ahead of a downloadable PDF (28%). Only 2.5% prefer a press release summarizing the findings.
| Format that makes research easiest to use | % |
|---|---|
| A webpage with findings and methodology | 44% |
| A downloadable PDF report | 28% |
| A collection of ready-to-use charts | 15.5% |
| An interactive dashboard | 8.5% |
| A spreadsheet containing the results | 4% |
| A press release summarizing the findings | 2.5% |
| No preference | 1% |
Methodology can beat a famous name
Faced with equally relevant research from an unfamiliar company and an established organization, 40% would choose whichever had the stronger methodology. Only 15% would prefer the established organization regardless.
| Unfamiliar company vs established organization | % |
|---|---|
| Choose whichever had the stronger methodology | 40.1% |
| Prefer the established organization if methodologies were comparable | 30.5% |
| Prefer the established organization regardless of methodology | 15% |
| Choose whichever had the more recent findings | 12% |
| Choose whichever was easiest to access and reference | 2% |
| Not sure | 2% |
What earns the link
The biggest single driver of linking to the original research page is access to detailed findings not available elsewhere (35%), followed by a clear explanation of the methodology (22%).
| What would most increase your likelihood of linking to the original page | % |
|---|---|
| Access to detailed findings not available elsewhere | 35% |
| A clear explanation of the methodology | 22% |
| An easy-to-reference summary of key statistics | 15.5% |
| Breakdowns relevant to my audience | 10.5% |
| Downloadable charts or tables | 8.5% |
| Permission to reproduce material with attribution | 5% |
| None of these | 1% |
What this means for brands publishing research
Taken together, the findings point to a simple checklist for research that gets cited and linked:
- Lead with a finding that challenges an assumption. Surprise is the most referenced type of result.
- Put the methodology on the page. Name who ran the survey, how many people responded and who they were.
- Publish on a webpage, not only a PDF. Keep the full detail on the original page so there is a reason to link to it.
- Keep it fresh. Most writers treat consumer data as current for about a year.
- Hold back on promotion. An overly promotional report is a reason not to cite it.
Two of our own projects follow this approach. For Grand Canyon University, we surveyed 1,518 US adults and published the findings, charts and methodology on the university's own website. For AMFM Healthcare, a survey of 1,032 supporters of people with bipolar disorder became a website-native white paper. The same survey data also powered press coverage for AMFM.
If you want research built this way, see our original research content and survey research services, or browse more citable figures in our original research statistics.
Methodology
This study is based on a survey of 863 people in the United States who write, edit or commission online content for work, including marketers, journalists and agency professionals, conducted by Cherry Data Signals in October 2026.
At a sample size of 863, the overall results have an approximate margin of error of ±3.3 percentage points at the 95% confidence level.
Respondents were sourced from a double opt-in online research panel and screened for relevant professional status. Eligibility was confirmed through both respondent profile information and in-survey qualification questions.
Responses were subject to quality controls including duplicate and re-entry detection, bot and automation detection, geo-verification, completion-time checks, attention checks and consistency review.
Results are based on self-reported opinions and behavior. Where respondents could select up to three answers, figures do not sum to 100%. Percentages are rounded and may therefore not always total exactly 100%.
Citation and reuse
Journalists and researchers are welcome to cite these findings with attribution to Cherry Data Signals, with a link to this page. For the underlying data tables or an interview, get in touch.
Want insights like these for your brand?
Cherry Data Signals runs custom surveys and delivers data-ready results. See how our survey research works, what a study costs in our survey cost guide, or how we turn results into original research content for PR, SEO and AI search.
Request a Survey