How to Choose an SEO Company in Penang: 12 Questions to Ask First

Published 2026-03-31 by Robin Ooi. Last reviewed " + TODAY + ".

SEO is unusually hard to buy. The work is invisible, the results are delayed, and the vocabulary is opaque enough that a confident presentation is difficult to distinguish from a competent one. Most bad SEO purchases in Malaysia are not made by careless buyers. They are made by careful buyers who did not know which questions carried weight.

These twelve do. Ask them in a first call. You are not testing knowledge, you are testing whether the answers are specific.

The twelve questions, and what a good answer sounds like

Working through the ones that matter most

Question 1: what are we measuring?

Rankings are a diagnostic, not an outcome. A provider whose entire reporting story is ranking positions has chosen the metric that is easiest to move and hardest to connect to money. Ask what happens if rankings improve and enquiries do not. If there is no answer to that, there is no plan for it, and it happens often. It is common enough that traffic without enquiries deserves its own diagnosis.

Question 3: where do the links come from?

This is the single most revealing question, because the risky methods are the ones people will not describe. Reluctance to name sources usually means private networks or bought placements on sites that exist only to sell links. Those work until they do not, and the failure lands on your domain rather than the provider's.

A good answer sounds concrete. Digital PR and journalist requests. Industry directories that have genuine editorial standards. Guest contributions to publications with real readers. Partner and association pages. Research or data assets that earn citations. Ask to see one live example. Not a spreadsheet, a URL.

Question 7: what would make you say no?

This question is uncomfortable to ask and very informative to hear answered. Any provider with real judgement has turned work down. Budgets that cannot fund the outcome being requested. Categories where the promised timeline is not achievable. Businesses where the constraint is the offer rather than the visibility.

Somebody who has never declined a project is selling capacity. That is a legitimate business model. It is just not the same product as advice.

Question 8: what is the AI search answer?

Almost every Malaysian provider now says they do AI SEO. Very few can describe what it changes in the actual work. A real answer covers structured data that defines your business as an entity, content shaped so that a direct answer can be lifted from it, and independent third party sources that corroborate what your own site claims.

The corroboration part is the one people skip, and it is the part that matters most. An assistant will not confidently name a business it has only ever read about on that business's own website. If you want the longer version of this, how AI assistants decide who to name covers the six signals involved.

Four answers that should end the conversation

1. A ranking guarantee. Nobody controls the ranking system. A guarantee is either attached to keywords with no search demand, or it is a signal that risk will be taken with your domain to hit a contractual number.
2. Refusal to name link sources. Covered above. The methods people hide are hidden for a reason.
3. No access to your own Analytics or Search Console. These are your properties. A provider who creates them under their own account and will not grant you admin has built an exit cost into the relationship deliberately.
4. A twelve month lock-in with no delivery milestones. An initial three to six month term is reasonable, because the work genuinely takes that long to evaluate. Twelve months with no break clause and no defined deliverables is not a term, it is a trap.

Three questions specific to Penang

Beyond the general twelve, three local checks are worth adding.

Do they understand which Penang you are selling to? Penang runs several separate search economies at once. Export buyers researching from overseas. Medical travellers comparing countries. Locals looking for someone nearby right now. A provider who treats all three the same way will build the wrong content for at least two of them.

Can they handle the language mix? If your customers search in Bahasa Malaysia or Mandarin as well as English, a monolingual strategy leaves a segment on the table. Language choice on Penang websites goes into what that costs and when a bilingual build is worth it.

Have they worked with your constraint type? MNC subsidiaries where head office controls the website. Manufacturers who cannot publish customer names. Clinics bound by healthcare advertising rules. These are normal in Penang and each one reshapes the programme. A provider who has not met them will propose something you cannot execute.

Before the first call

Run the free scan on your own site first. It takes ten seconds and it changes the conversation, because you arrive with a document rather than a question. If a provider's diagnosis contradicts what the scan found, ask them to explain the difference. The answer to that is worth more than the rest of the pitch.

Then decide what shape of provider you actually need. Agency, freelance consultant, in-house hire and doing nothing are four different cost structures, and the comparison between them is a genuinely open question rather than a sales argument.

If you want the twelve questions answered directly, that is essentially what the free consultation is. Thirty minutes, no pitch deck, and you keep the answers whether or not anything follows.