Research · 7 August 2026
We Pulled the Google Rating of 220 Singapore Agencies. 62% Have a Perfect 5.0.
We recorded the Google star rating and review count for every marketing, social and video agency Google returns in Singapore. 62% score a perfect 5.0, and the median listing has 17 reviews. Here is why the rating is noise and what to read instead.
By Launch Media Studios
If you are shortlisting an agency in Singapore, the first filter most people reach for is the star rating on the Google listing. It is right there in the search results, it looks objective, and sorting by it feels like doing homework.
We wanted to know whether that number carries any information at all. So on 6 August 2026 we pulled the Google Business Profile for every marketing, social media and video production agency Google returns for the obvious Singapore search terms, and recorded the star rating and the review count for each one. That gave us 220 unique agencies.
A disclosure before the numbers, because it matters: we are an agency ourselves. Launch Media Studios competes in this exact market. We left ourselves out of the dataset entirely, and we do not name, rank or comment on any individual firm below. Everything here is aggregate. The point of the exercise was not to find out who is best; it was to find out whether the rating can tell you who is best. It cannot, and the reason is worth twelve minutes of your time.
How we built the list
We ran the queries a buyer would actually run (digital marketing agency, social media marketing, video production company), took every Singapore business Google returned across those searches, de-duplicated by listing ID, and recorded the star rating and review count exactly as Google displayed them on 6 August 2026.
The 220 businesses break down roughly as: 91 listed by Google under video production service, 76 under marketing agency, 24 under internet marketing service, and the remaining 29 spread across advertising agency, film production, public relations, branding, design and web categories. Google's own category labels are loose, so treat those buckets as indicative rather than precise.
Of the 220, 17 have no star rating at all, because they have zero reviews. That leaves 203 rated agencies, and those 203 are what every statistic below is calculated on. Between them they hold 7,674 reviews.
Finding one: almost everybody has the same rating
The median rating across 203 agencies is 5.0. The mean is 4.90. The standard deviation is 0.22.
Nine agencies in ten sit at 4.7 or better. Here is the full distribution:
| Google rating | Agencies | Share of rated agencies | Median review count in band |
|---|---|---|---|
| Exactly 5.0 | 126 | 62.1% | 13 |
| 4.9 | 35 | 17.2% | 29 |
| 4.8 | 17 | 8.4% | 36 |
| 4.5 to 4.7 | 17 | 8.4% | 19 |
| Below 4.5 | 8 | 3.9% | 7 |
Read the top row again. 62% of rated Singapore agencies have a flawless 5.0. A further 17% are at 4.9. Add 4.8 and you have covered 178 of 203 agencies, or 87.7% of the market, inside a band 0.2 stars wide.
Only 8 agencies out of 203 fall below 4.5. Exactly one falls below 4.0.
This is what a useless signal looks like. A metric where 88% of the sample sits in the top 4% of the available scale is not measuring quality; it is measuring whether a business has a Google listing and asks happy clients to review it. If you sort a shortlist by "highest rated" in this category, you are sorting 126 agencies that are all tied on first place, and the order you get back is whatever Google felt like showing you.
Finding two: the perfect scores are the thin ones
Here is where it gets practically useful. Among the 126 agencies with a perfect 5.0, the median review count is 13.
- 21 of the 126 perfect scorers (16.7%) have fewer than 5 reviews.
- 46 of them (36.5%) have fewer than 10.
- 77 of them (61.1%) have fewer than 20.
- Only 7 of them (5.6%) have 100 reviews or more.
A 5.0 built on 6 reviews and a 5.0 built on 300 reviews render identically in search results, and they are not remotely the same claim. The first says six people, who the agency very likely chose to ask, were happy. The second says the agency has survived hundreds of client relationships without a single serious public complaint, which is genuinely hard.
The arithmetic behind this is worth seeing, because it explains why small samples cling to 5.0 and why they stop being 5.0 the moment anything goes wrong:
| Current rating | Existing reviews | Rating after one 1-star review | Drop |
|---|---|---|---|
| 5.0 | 6 | 4.43 | 0.57 |
| 5.0 | 17 (the market median) | 4.78 | 0.22 |
| 5.0 | 50 | 4.92 | 0.08 |
| 5.0 | 150 | 4.97 | 0.03 |
| 5.0 | 300 | 4.99 | 0.01 |
An agency with 6 reviews is one unhappy client away from 4.4. An agency with 300 reviews barely moves. So when you compare a thin 5.0 against a thick 4.8, you are not comparing a better agency to a worse one. You are comparing an untested record to a tested one, and the tested one is carrying more information.
Finding three: the rating tells you nothing about the volume, and vice versa
We checked whether higher-rated agencies tend to have more reviews. They do not. The correlation between star rating and review count is 0.014, which is statistically indistinguishable from zero. On a rank correlation the relationship is very slightly negative (-0.18), meaning the busiest listings are, if anything, a fraction lower rated than the quiet ones. That is exactly what you would expect: more clients means more chances for one of them to be unhappy in public.
Grouping agencies by review volume makes the flatness obvious:
| Review count | Agencies | Share | Average rating | Share of all 7,674 reviews |
|---|---|---|---|---|
| 1 to 4 | 25 | 12.3% | 4.85 | 0.8% |
| 5 to 9 | 36 | 17.7% | 4.90 | 3.2% |
| 10 to 24 | 63 | 31.0% | 4.90 | 12.6% |
| 25 to 49 | 44 | 21.7% | 4.92 | 19.5% |
| 50 to 99 | 19 | 9.4% | 4.89 | 18.0% |
| 100 or more | 16 | 7.9% | 4.92 | 45.8% |
The average rating moves by 0.07 stars across a group that ranges from a handful of reviews to several hundred. The rating column is flat. The volume column is not flat at all, and that is the column with the information in it.
Finding four: review volume is extremely top-heavy
The median Singapore agency in our dataset has 17 Google reviews. The largest has 471, which is 27 times the median.
- 61 agencies (30.0%) have fewer than 10 reviews.
- 112 agencies (55.2%) have fewer than 20.
- Only 16 agencies (7.9%) have 100 or more.
- The 10 largest listings hold 36.6% of all reviews in the market. The top 20 hold 50.5%.
- The bottom half of the market, by review count, holds 10.9% between them.
So a small group of agencies has accumulated most of the public evidence that exists in this category, and everyone else is operating on a handful of data points. Note that this is a measure of listing maturity and review-collection discipline, not of talent. Plenty of excellent small studios have 8 reviews. The point is that with 8 reviews, the star rating cannot tell you which ones they are.
A method that actually works
Since the rating is noise, here is what to do instead. It takes about twenty minutes per shortlisted agency, which is cheap against a retainer of S$3,000 to S$10,000 a month.
1. Filter on volume first, rating second
Set a floor on reviews before you look at stars at all. In our dataset, requiring 4.8 or better and at least 50 reviews cuts 220 agencies down to 32. That is a real shortlist. Requiring 4.8 or better on its own leaves you with 178, which is not a shortlist, it is the phone book.
2. Read the text, sorted by newest
Default Google sorting hides the recent ones. Sort by newest and read the last twelve months. Agencies change: staff leave, founders get busy, the team that earned the 2023 reviews may not be the team quoting you. A 5.0 whose last review was two years ago is a historical artefact.
3. Look for named deliverables, not adjectives
"Professional and responsive" is written by someone who was asked for a favour. "They rebuilt our Shopify product pages and our checkout completion went up" is written by someone who remembers what happened. Count the reviews that name a specific service, a specific person, or a specific outcome. In most listings that count is far lower than the total.
4. Check whether reviews arrive steadily or in clumps
Fourteen five-star reviews inside one week, then nothing for a year, is a review drive, not a track record. Steady arrival over years is the harder thing to manufacture and the better signal.
5. Read the negatives properly, and read the replies
One or two critical reviews on a large listing is normal and healthy; it also proves the listing is not curated. What matters is the reply. A calm, specific, non-defensive reply from the agency tells you more about how they will handle your problem than any five-star review will.
6. Then ignore Google and ask for two references
Ask for two clients in your industry, at your budget, from the last twelve months, and ask to speak to them. An agency that can produce those in a day is telling you something no star rating can. If you want the longer version of this process, we wrote it up in our guide to choosing a social media marketing agency in Singapore.
What this means for us, honestly
We are new. Our own Google listing is thin, which by the logic above means our star rating should not persuade you of anything either, and we would rather say that plainly than pretend otherwise. It is one of the reasons we published this dataset instead of a badge.
The wider point is that this category has quietly agreed to compete on a metric that no longer separates anyone. When 62% of a market is perfect, perfect is the entry fee, not the differentiator. The things that still separate agencies are the boring ones: how many client relationships they have actually completed, whether the work is visible, and whether real clients will pick up the phone about it.
If you want to see how we work before there is a review pile to read, that is what our social media agency page is for.
Method note: 220 Singapore agency listings collected from Google search results on 6 August 2026, de-duplicated by Google listing ID. 203 carried a star rating; 17 had zero reviews and no rating. All percentages are calculated on the 203 rated listings unless stated. Ratings and review counts are point-in-time and will drift. Launch Media Studios was excluded from the dataset.