Paying a creator this year-end? Check the engagement is real first

Before you pay a creator for a year-end or Chinese New Year campaign, check that the engagement is real. Look at where the audience is, which languages they comment in, whether likes and comments match the follower count, and whether growth has jumps nobody can explain. For influencer marketing in Malaysia there is now a second reason to check: the draft Influencer Guidelines 2026 would expect brands, not only creators, to take reasonable steps against fake engagement.

With Chinese New Year on 6 February 2027 (and Ramadan close behind), many F&B brands are shortlisting creators and KOLs (key opinion leaders) now. A shortlist built on follower counts is quick. One built on evidence takes an afternoon longer and wastes far less budget.

What is fake engagement in influencer marketing?

Fake engagement is any follower, like, view or comment that does not come from a real person with a genuine interest in the content. It includes bought followers (pengikut palsu), purchased likes and comments, engagement traded between groups of accounts to game the algorithm, and testimonials written to look like real customers.

It matters to an F&B brand for a simple reason: fake followers don’t walk into outlets. A creator with an inflated audience can deliver an impressive reach figure and very few customers, and the campaign report will look better than the till receipts.

Not every odd number is fraud. Creators post viral hits, get reshared by bigger accounts and lose followers after a quiet month. The aim isn’t to accuse anyone, only to see whether the numbers hang together before you commit money.

What do the draft rules mean for influencer marketing in Malaysia?

The Communications and Multimedia Content Forum of Malaysia (CMCF), the industry body behind the Content Code, published its draft Influencer Guidelines 2026 for public consultation from 17 August to 30 September 2026. The guidelines are still a draft. CMCF said it would review the submissions before finalising them, and we could not find a final version at the time of writing.

As drafted, the guidelines would:

• Rule out fake engagement. Influencer marketing would not rely on fake followers, likes, views, purchased comments or fabricated testimonials that create a false impression of popularity.
• Share the responsibility with brands. Brands, agencies and influencers would be expected to take “reasonable steps to avoid them in campaign planning, partner selection and monitoring”.
• Ask for clear labels. Paid posts would carry a label such as “Ad”, “Advertisement”, “Paid Partnership” or “Sponsored”, early in the content. Short forms such as “sp” or “collab”, or a brand tag on its own, would not be enough.
• Think about language. Where content reaches audiences in Bahasa Malaysia, Mandarin, Tamil or English, a disclosure in only one language might not be sufficient.

The draft also covers AI-generated content, children and misleading claims, as The Star reported. Compliance would be voluntary, like the Content Code it would form part of, though the draft notes the Code already requires paid social media content to be clearly disclosed. CMCF chief executive Mediha Mahmood put the case plainly: influencer marketing “has become too important to be governed by guesswork”.

This is a summary, not legal advice. Check the final text when it is published, or ask your adviser.

How do you check a creator’s engagement before you pay?

Ask the creator or their agency for screenshots of their own platform analytics, covering audience location, age and recent reach, then look through their last 20 to 30 posts yourself. Work through these signals.

Signal	What to check	Red flag
Audience location	Share of followers in Malaysia, and in the cities where your outlets are	Most followers are overseas, or the creator won’t share analytics
Audience language	Whether comments are in the languages your customers use: Bahasa Malaysia, English, Mandarin or a mix	Comments in languages none of your customers speak
Engagement against followers	Likes, comments and views across the last 20–30 posts, against the follower count	Very low engagement for the size of the account, or wild swings with no reason
Follower growth	Month-by-month growth over the past six to twelve months	Thousands of followers gained in a day or two with no viral post to explain it
Comment quality	Whether comments mention the food, price, portion or outlet, or tag friends	Bursts of one-word comments, emoji strings or the same phrase from different accounts
Sponsored against organic	Engagement on past paid posts compared with ordinary posts	Paid posts get a fraction of the usual response, or a different set of generic commenters
Disclosure habits	Whether past paid posts carry a clear “Ad” or “Sponsored” label early	Paid posts tagged only “collab”, “sp” or the brand, or not labelled at all

No single signal settles it. A creator who fails one check may have a good explanation. One who fails three probably doesn’t.

Two creators, one budget. Take a hypothetical 15-outlet chain in the Klang Valley choosing a creator for a year-end menu launch. Creator A has 120,000 followers; posts swing from 300 likes to 3,000 with no clear pattern, and most comments are “Nice”, “Wow” or emoji strings, many posted within minutes of each other. Creator B has 18,000 followers; every post gets roughly 600 to 900 likes, and the comments read like customers: “Sedap gila, weekend ni nak try”, “好吃”, “Portion besar for the price”, and friends asking which branch it is. On reach, A wins easily. On the evidence, B is more likely to send people to an outlet. Brief B now, ask A for audience analytics, and decide once they arrive.

Put it in the brief. The draft guidelines note that clear written agreements and records help everyone understand their role. Three lines do most of the work.

1. No bought engagement. The creator confirms that no purchased followers, likes, views or comments will be used on the campaign, and both sides agree what happens if any are found.
2. Labelling. The brief says how the post will be labelled, where the label appears and in which language, matched to how the content is delivered: caption, on-screen text or spoken.
3. Analytics after posting. The creator shares post-level analytics, such as reach, audience location, saves and shares, within an agreed number of days.

How do you know if the campaign worked?

Likes are the easiest number to report and the least useful. Three measures tell you more.

Sentiment. Read what the campaign started. Are people asking where the outlets are and what the set costs, or talking about something else entirely?

Share of voice. Did your brand take a bigger share of the category conversation during the campaign window than before, compared with competitors running their own year-end pushes? We covered how to measure it in knowing where you actually stand in your category.

Outlet sales. Compare the featured item’s sales at outlets near the creator’s audience with the same weeks at similar outlets, or track a creator-specific code or set menu. Conversation tends to reach the tills later, as we described in social signals are demand signals.

Keep a one-line scorecard per creator. Next year’s shortlist starts there.

Where Adqlo fits

Codemax Adqlo is social media intelligence for the industries and brands you choose, across TikTok, Instagram, X, Facebook and YouTube.

• Before the brief. Creator analytics show audience, engagement and authenticity signals for the creators in your category, so the shortlist starts from evidence. These are signals for your team to weigh. Adqlo doesn’t certify creators or decide who is genuine; people make that call.
• Reading the comments. Sentiment is tuned to category language, including Bahasa Malaysia, English, Mandarin and mixed-language posts like the ones above.
• During the campaign. Spike alerts flag when volume or sentiment moves outside the bands you set, routed to Slack, Teams or WhatsApp.
• Afterwards. Sentiment and share of voice show what changed against competitors. Because Adqlo feeds the AI-Kitchen Command Center, the campaign window can sit beside outlet sales rather than in a separate slide deck.

Questions worth asking before you sign

• Have we seen the creator’s own audience analytics, or only the follower count?
• Do the comments talk about food and outlets, in the languages our customers use?
• Is there a jump in followers nobody can explain?
• Do past paid posts carry a clear label early, in the audience’s language?
• Does the brief rule out bought engagement and say how the post will be labelled?
• Have we agreed to judge the campaign on sentiment, share of voice and outlet sales?

If a creator can’t help with the first question, that tells you most of what you need to know.

Frequently asked questions

How do you check if an influencer has fake followers?

To check whether an influencer has fake followers, ask for their own platform analytics and review their last 20 to 30 posts. Look at where the audience is, whether engagement matches the follower count, whether growth shows unexplained jumps and whether comments are specific or generic one-word bursts. One odd signal can have an innocent explanation; several together deserve a closer look.

What are Malaysia's Influencer Guidelines 2026?

The Influencer Guidelines 2026 are draft guidelines from the Communications and Multimedia Content Forum of Malaysia (CMCF), published for public consultation from 17 August to 30 September 2026. As drafted, they would cover disclosure of paid content, fake engagement, AI-generated content, children and misleading claims, and would form part of the voluntary Content Code. They were still a draft at the time of writing.

Do sponsored posts need to be labelled in Malaysia?

Sponsored posts in Malaysia are already expected to be clearly disclosed under the Content Code's advertising provisions, according to the Content Forum's draft Influencer Guidelines 2026. The draft would add detail: a clear label such as 'Ad' or 'Sponsored' early in the content rather than 'collab' or a brand tag, and a check on whether one language is enough for the audience. This is not legal advice.

What is a good engagement rate for an influencer?

A good engagement rate for an influencer depends on the platform, the content type and the size of the account, so no single number applies. Compare a creator with their own recent posts and with similar creators in the same category. Consistency is a better sign than a high headline figure, and a rate far out of line with similar accounts is worth questioning.

How should F&B brands measure an influencer campaign?

F&B brands should measure an influencer campaign on three things beyond likes: the sentiment of the conversation it started, the change in share of voice against competitors during the campaign window, and sales of the featured item at outlets near the creator's audience, ideally tracked with a creator-specific code or set menu. Keep a scorecard per creator for the next shortlist.

Can social listening tools spot fake engagement?

Social listening and creator analytics tools, such as Codemax Adqlo, show audience, engagement and authenticity signals for creators in a category, which makes unusual patterns easier to see. These are signals for people to judge, not proof. A marketing team still needs to review the evidence, ask the creator for their own analytics and make the decision.

Sources

1. Public Consultation Paper No. 1/2026: The Influencer Guidelines 2026 — Communications and Multimedia Content Forum of Malaysia, 17 Aug 2026
2. New guidelines for influencers to cover AI-generated content, child exploitation and fake claims, says Content Forum — The Star, 2 Sep 2026
3. Proposed rules set to rein in influencers — The Star, 3 Sep 2026