Polls published without clear methodology erode reader trust.
Breaking stories need answers faster than traditional fieldwork allows.
Graphics teams need clean, exportable data rather than raw tables.
Mapped platform capabilities
What Closer does for this use case.
Each card shows the kind of analytic you get, with illustrative sample data.
Hours to close, by sample sizeTypical fieldwork duration
n=500 county pulse36hrs
n=1,000 national48hrs
n=1,500 national60hrs
n=2,500 national72hrs
Sample view · illustrative data
56 hrsMedian commission to results
Rapid Pulse Polls
Rapid-turnaround national and county pulse polls on breaking national issues.
A questionnaire is approved, fielded across SMS, USSD, WhatsApp and web, and closed within 48 to 72 hours, so a newsroom can publish while the story is still live. Quotas by region, age and gender are monitored during fieldwork. It gives editors an answer on breaking issues, not a stale one.
Fieldwork in 48 to 72 hours
Live quota monitoring by region, age and gender
National or single-county pulse options
Sample quality scoreRepresentativeness vs census, 95% band
Sample view · illustrative data
2.5 ptsMargin of error at n=1,500
Methodological Disclosure
Clear methodological disclosures (sample size, channel mix, margin of error) suitable for editorial publication.
Every poll ships with a disclosure block covering sample size, dates, channel mix, weighting and margin of error, written in plain language for a footnote or on-air read. Readers can see exactly how the numbers were produced. It protects the newsroom when a result is challenged.
Plain-language disclosure ready to publish
Margin of error by subgroup, not only total
Observed versus surveyed labels on every figure
Support for a policy by countyPercent in favour, illustrative sample
Nairobi58
Mombasa54
Kisumu49
Nakuru52
Eldoret46
Garissa41
Nyeri60
Kitui44
Turkana38
Sample view · illustrative data
4Export formats per poll
Newsroom Data Exports
Interactive visual data exports for newsroom graphics teams.
Results are delivered as clean CSV and JSON, plus chart-ready tables and an interactive embed, with county names matched to standard boundaries. Graphics teams skip re-keying numbers and get confidence bands in the file. It saves hours before a broadcast or front page deadline.
CSV, JSON and embeddable interactive charts
County names matched to map boundaries
Confidence bands included in every export
Sample by channelShare of completed interviews
WhatsApp31%
Web panel18%
SMS24%
USSD27%
Sample view · illustrative data
38%Interviews on feature-phone channels
Channel-Balanced Samples
Blend digital and feature-phone channels so polls are not skewed toward urban smartphone users.
Smartphone panels on WhatsApp and the web are blended with USSD and SMS respondents on feature phones, then weighted to census age, gender, region and urban or rural split. Without the feature-phone share, results lean towards Nairobi and younger smartphone users. Blending keeps rural voices in the number.
Channel mix reported alongside the topline
Weights for age, gender, region and urban or rural
Comparison of raw versus weighted result
Approval of a county serviceSurveyed, monthly waves
Sample view · illustrative data
+6 ptsChange since first wave
Trend Trackers
Repeat the same questions over time to show how opinion moves.
The same questions are asked on a schedule, such as monthly, with a stable method so changes are real movement and not method noise. Each point shows its confidence band, so editors know when a shift is within error. It turns a one-off poll into a story that develops.
Fixed wording and sampling across waves
Confidence band on every wave
Flag when a change exceeds the margin of error
Sample deliverables
What you receive.
Open any report to see how it reads, or save it as a PDF. All figures are illustrative.
A publish-ready package with headline results, chart files and plain-language method for one national poll.
This pack gives editors the topline, key subgroup cuts and approved wording for a story or broadcast. Producers and data journalists use it to publish quickly without questions about method.
County-level results for a national poll with sample sizes and margins of error for every cell.
This workbook gives full crosstabs by county, age and gender, flagged where samples are too small to publish. Data teams and regional editors use it to build local stories without overreading thin cells.
A one-page, plain-language statement of how a poll was run, ready to publish with the story.
This sheet sets out who was asked, how, when and with what error, in wording editors can drop under a chart or read on air. Standards editors and readers use it to judge the poll fairly.