Insights and guides
This section brings together practical guidance on reading polls properly and planning research that can be used with confidence. It is written for teams that need clear decisions, fair interpretation and method notes they can actually work with.

What you will find in this section
Our insights and guides are here to help clients, communications teams, policy teams and researchers use evidence carefully. Some pages explain how to read a poll without overreaching. Others focus on the steps involved in commissioning a survey, from defining the audience to deciding how results will be reported.
If you are comparing proposals, this section can help you ask better questions before fieldwork starts. If you already have findings in front of you, it can help you check what the data does and does not support. The aim is simple: clearer decisions, fewer assumptions and better use of research in real working conditions across South Africa.
Choose the right guide for your task
Different guides answer different needs. Start with the problem you are trying to solve, not with a method name. That usually keeps the project focused and avoids adding questions that look useful but do not help the final decision.
- Use “How to read a poll” when you need to interpret headlines, charts or topline findings with care.
- Use “Commissioning a survey” when you are planning a new project and need to brief it properly from the start.
- Look at methodology pages as well if you want more detail on sampling, weighting or fair question wording.
- Read before approvals are signed off if several internal teams need to agree on scope, audiences and outputs.
A short guide at the right moment can save a lot of back and forth later, especially when timelines are tight or several provinces are involved.
Reading a poll versus commissioning a survey
These are related tasks, but they are not the same. One is about interpretation after data exists. The other is about making sound choices before any interviewing begins. Mixing the two can lead to confusion or weak briefs.
- Reading a poll: focus on wording, field dates, geography, interviewing mode, likely audience reached and the method notes supplied with the findings.
- Commissioning a survey: focus on the business or policy decision, target groups, language needs, where interviewing must happen and how results should be delivered.
- Reading a poll: ask whether the conclusion goes further than the evidence allows.
- Commissioning a survey: ask whether the questionnaire will produce information that the team can actually act on.
In practice, strong projects do both well. They begin with a careful brief and end with careful interpretation, not a rush to claim more certainty than the evidence can carry.
Questions worth asking before fieldwork starts
A good brief is usually plain and specific. It should explain who needs the findings, what decision is being supported, and what must be understood by province, language group or customer segment. In South Africa, practical details matter early: travel time, municipal access rules at some venues, call windows that suit working households and language coverage for face-to-face interviewing.
It also helps to decide what a useful final output looks like. Some teams need a concise management summary. Others need fuller reports with method notes and cross-breaks. Get that agreed before the questionnaire is finalised, otherwise the project can drift a bit and the reporting stage becomes harder than it should be.
Using findings carefully inside your organisation
Research is most useful when it is shared with the right level of caution. A single chart should not carry the whole story. Findings need context: what was asked, who was reached, how interviewing happened and what limits apply. This matters whether the topic is consumer behaviour, service experience, membership views or public opinion.
- Keep the original question wording close to any chart or summary point.
- Separate measured findings from interpretation or recommendation.
- Check whether provincial differences are meaningful before changing plans around them.
- Do not treat one wave of work as permanent truth if conditions have shifted.
- Store method notes with the report so colleagues can refer back later.
Careful use protects good research. It also makes internal conversations more useful, because teams can debate the implications without arguing about what was actually measured.
How these guides connect with our work
Populism Polls runs public opinion polling and survey research nationally from our base in Braamfontein, with work across all nine provinces. That includes national and provincial polls, consumer and brand surveys, phone and online panels, face-to-face field interviews in all official languages, questionnaire design, sampling and weighting, and reporting with clear method notes.
We are non-partisan. We do not campaign, canvass or predict election results, and we do not publish client data. These guides reflect that approach: fair questions, transparent methods and findings that can be used sensibly. If you are planning a project and want to discuss scope, use the enquiry form on our contact page and tell us what decision your team needs to support.
What our Opinion Polling customers say
Verified Google reviews
Questions we get asked
Who are these guides written for?
They are written for brands, agencies, NGOs, associations and policy teams that need to read polls properly or plan survey research with a clear brief.
Do the guides include technical methodology?
Yes. This section works alongside our methodology pages, so you can move from plain-language guidance into topics like sampling, weighting and fair question wording when needed.
Can we ask about a project after reading the guides?
Yes. If you want to discuss a poll or survey project, send your requirements through the enquiry form on our contact page and we will come back to you.
Tell us about the job
Send sizes, materials and where you are through our contact form, and we will come back with the practical options.
Open the contact form


