Quick answer: Treat East Malaysia (Sabah and Sarawak) as its own base rather than a slice of one national sample. Decide first whether you need results you can read for East Malaysia on its own or only a representative national picture, then set a separate target and quota for each region, check language needs, and confirm feasibility early because online availability is usually thinner than in Peninsular Malaysia.
Why is East Malaysia treated differently from the Peninsula?
Sabah and Sarawak differ from Peninsular Malaysia in population mix, language, urban and rural spread, retail channels and often brand availability. A product that is everywhere in Klang Valley may have patchy distribution in parts of the East, and the media people use can differ too.
This matters for sampling. If you simply let a national sample fill naturally, East Malaysia will usually be a small share of completes, and you will not be able to say anything reliable about it. If you force it up without planning, you may pay for a sample the rest of the study does not need.
Do you need a national picture or a readable East Malaysia result?
This is the first decision, and it drives everything else. The two goals need different designs.
| Goal | What it means for the sample | Typical approach |
|---|---|---|
| A national picture that includes East Malaysia | East Malaysia appears in proportion to the population, so it cannot be read on its own | Proportional quotas, weighting if needed |
| A readable East Malaysia result | East Malaysia needs enough completes to analyse by itself | Separate target for East Malaysia, often boosted |
| A comparison of East and West | Both regions need comparable bases | Two bases with matched quotas |
A common mistake is to want the third result while designing for the first. If someone in the room will ask what East Malaysian consumers think, build the sample so that question can be answered.
How should you set quotas for East Malaysia?
Calculate feasibility and quotas separately for each region, because incidence and panel depth can differ. A behaviour that is common in the Peninsula may be less common in the East, or the reverse, so a single national incidence figure can mislead you.
- Set a separate target for each base, for example Peninsular Malaysia and East Malaysia, rather than one blended number.
- Within East Malaysia, decide whether Sabah and Sarawak need to be read separately. They are distinct markets, and splitting them doubles the number of cells.
- Check how ethnic and religious composition differs from the Peninsula, and set quotas that reflect the local mix instead of copying a national split.
- If a screening criterion depends on ethnicity or religion, build the audience from the local data, not from a national average.
- Keep the number of layered quotas within East Malaysia small, because the pool is thinner.
For the arithmetic behind this, see our guide to checking a niche audience is feasible in Malaysia.
What about language and questionnaire design?
Most Malaysian online surveys run in English and Bahasa Malaysia, and many add Chinese. Respondents in the East may be more comfortable in Bahasa Malaysia, and some prefer a local language in conversation, although written surveys are usually in the standard languages. Offer a language switch, and have a native speaker review the translation rather than relying on automatic translation alone.
Watch the examples in your questions as well. Retailers, media and even food references that feel natural in the Peninsula may not exist in the East. Keep brand and channel lists local, and use an open option so respondents can tell you what you missed.
When is an online survey not enough?
Online research reaches people with a smartphone and a connection, and coverage is thinner in rural and interior areas. If your audience is mostly rural, older, or depends on face-to-face contact, a local field agency running interviews on the ground may serve you better, or you may combine both methods. Be honest about this in the design, and say in the report who the sample does and does not reach.
A short pre-fieldwork checklist
- Have you decided between a national picture, a readable East Malaysia result, or an East versus West comparison?
- Is there a separate target and quota for each base?
- Have you checked incidence for each region, not just nationally?
- Is the questionnaire available in the languages your respondents use, with a native review?
- Are brand, retailer and media lists relevant locally?
- Do you have a plan for what to relax if East Malaysia fills slowly?
Allow extra time for the thinner region. Our guide to how long market research takes shows where fieldwork sits in the schedule, and the survey sample size guide helps you decide how many completes each base needs.
Where does Vase.ai fit?
If East Malaysia matters to your decision, send us the research question, the audience, the regions you need to read separately and your date. We will suggest a study design and sample, and flag feasibility risks up front.
We build Vase.ai, a research platform with a verified panel of SEA consumers and separate quota controls for each region. Telling us early whether you need a national picture or a readable East Malaysia result is the single most useful detail, because it shapes the whole design.
Frequently asked questions
Can I just add Sabah and Sarawak to a national sample?
You can, but they will usually be a small share and you cannot read them on their own. If you only need a national picture that is fine. If anyone will ask about East Malaysia specifically, set a separate target.
Do I need to treat Sabah and Sarawak as two separate markets?
Only if you need to read them separately. They have different populations and behaviours, so combining them is a simplification. If the decision depends on either state alone, give each its own base.
Is an online survey reliable in East Malaysia?
It is reliable for people who are online, which is a growing share but not everyone. For audiences who are largely offline, rural or older, consider a local field agency or a mixed approach, and state the coverage limits in your report.
