Pricing a new product in Singapore, where there is no second region to test the number in

Set the floor from what the thing costs you to deliver, set the ceiling from what it is worth to the buyer in money they can count, and then choose a number between them based on how fast your advantage decays. That is the whole of pricing a new product, and the reason it still goes wrong is the step nobody writes down: deciding in advance what you will change when the number turns out to be wrong. In Singapore that step carries more weight than it does in a large market, because there is no quiet region to try a second price in. One city, one audience, one published figure, seen by most of the people who were ever going to buy.

The demand behind this question here is small and specific. Google Keyword Planner, queried against the Singapore geo target 2702 for this plan on 26 October 2026, reports 10 searches a month for pricing strategy for new product, and 10 is the figure for the whole normalised group rather than a total of its phrasings. Most remaining Singapore queries in this set sit at that same floor.

Why a new product is a different problem from a repriced one

Repricing an existing product is an exercise in evidence. You have a conversion rate, a unit history, a churn figure and a sense of which customers left when you last moved. A launch has none of that, and three things fill the gap whether you choose them or not.

• The buyer has no reference price for your product, so they borrow one. Whatever they already pay for the nearest substitute becomes the number yours is judged against, and that comparison happens whether or not the substitute is actually comparable.
• Your own first number becomes the anchor. Everyone who looks at the product in its first months learns a price, and that is the figure they carry into the next conversation, including the conversation where you have raised it.
• You have no conversion rate, so you cannot yet tell a price problem from a product problem. Low take up at launch is ambiguous by nature, and the way out of the ambiguity is to vary one thing at a time, which is exactly what a single market makes harder.

The four inputs you can get before you have a single customer

None of these needs sales data. All four can be on a page within a week.

• Your delivered cost per unit, honestly loaded. Not the marginal cost. The cost including the part of somebody's week that the delivery consumes. This is the floor, and a price below it is a decision to buy customers rather than to sell a product, which is a legitimate choice only if you have written down how long you will do it for.
• The buyer's measurable gain, in their money. Hours saved times their hourly cost, penalties avoided, revenue unlocked. This is the ceiling, and if you cannot state it in a sentence the buyer would agree with, the ceiling is unknown rather than high.
• The published prices of the nearest substitutes, with the date you read each one. Expect most of them to publish nothing, which is itself the finding rather than a dead end. Our page on competitor pricing, and the real work when most rivals do not publish a price sets out what to do with a market that quotes on request.
• The number your own conversations stop at. Ten structured conversations with people who have the problem, each one ending with a specific figure and their reaction to it, will tell you more than any benchmark. Ask what they pay now for the workaround, not what they would pay for yours.

With those four you have a floor, a ceiling, a competitive band and a reaction. The choice inside the band is the strategy, and it is one decision rather than a framework.

The one decision inside the band

High or low at launch comes down to whether your advantage decays. Ask how long it would take a competent competitor to offer something a buyer would accept as equivalent.

If the answer is months, the opening is a window and a high price is how you collect the margin while it is open. The discipline that makes it work is the plan for coming down, and our page on price skimming and the four decisions to take before you launch high works through the triggers, the stages and the floor.

If the answer is years, because the advantage is a relationship, a dataset or a network that gets better with use, share earlier is worth more than margin now and the lower opening is the rational one. The arithmetic for both, and for six other named approaches on one product, is laid out in our page on the same product priced eight ways and what each way does to the units you need, so this page will not restate it. The question worth bringing back here is the one that page asks in its fourth column: a lower price needs more units to stand still, and you are choosing the volume you then have to deliver.

What a single city market changes

In a larger country the standard move is to launch a price in one region, read the result, then roll the better number out. Singapore does not offer that. The whole domestic market is one city, the commercial search demand for most categories sits on a short list of queries, and your price page is seen by the people you were going to sell to and by the competitors who were going to watch. Three consequences follow, and all three are about sequence rather than about the number.

• Decide the final displayed figure before you publish anything. The number a Singapore buyer sees has a specific meaning here, and it is worth settling the display question before the architecture is built rather than re-pointing a ladder afterwards. Our page on pricing strategy for a Singapore business and the display rule that sets your price points covers that rule and the formats it accepts.
• Make the offer reversible, not the price. A launch discount with a stated end date, a founding cohort rate for the first twenty customers, a bundled first month: each of these lets you learn what a lower number does without publishing a lower price. When the offer ends, nothing has to be announced, because the price never moved.
• Treat the first cohort as the people who will describe your price to everybody else. In a market this size they overlap with the people you want next. Decide now what they keep if the price changes, and say it to them at the point of sale rather than when they find out.

The asymmetry nobody plans for

Coming down from a published price is easy to execute and expensive in signal: the customers who paid more learn that they did. Going up is the opposite, cheap in signal for new buyers and difficult with the people already on the old number. So the two directions are not mirror images, and the sensible default is to publish the price you want to be charging in twelve months and to put the discount, not the price, on a timer.

That one rule removes most of the awkward conversations. A customer who bought at a launch offer understands that an offer ended. A customer who bought at a price understands that a price was raised, and those are different feelings about the same amount of money.

Write these six lines down before the price is published

• The floor, with the cost assumptions that produced it.
• The ceiling, stated as the buyer's own gain in their own money.
• The number you chose, and the sentence explaining why it sits where it does inside the band.
• The trigger that would make you change it, written as an observable event rather than a date where you can manage it.
• What the first cohort keeps if it changes.
• The date you wrote all of this, because the four inputs move and a decision with no date on it cannot be reviewed.

Six lines, one page, and it is the document that turns a launch price into something a team can defend a quarter later. If the product is one item inside a range rather than a standalone thing, the step between your tiers matters as much as the absolute numbers, and the decisions behind that are in our page on product mix pricing and the step gap that decides which tier people buy.

Three launch prices that make trouble later

• A free tier with no path out of it. Free is a price, it has a server bill and a support cost, and it works when the free version creates the need for the paid one. When it simply solves the problem, the paid tier is competing with your own product.
• A number matched to a competitor whose cost base you cannot see. Matching a price means inheriting somebody else's margin assumptions along with their number. If their delivery is cheaper than yours, the match is a loss you have chosen.
• A discount dressed as a price. A launch rate with no end date becomes the price within two quarters, and the margin you planned on was in the difference. Put the date in the first announcement.

Questions about setting the price of a new product in Singapore

How do I price something with no competitors at all

Price against the workaround rather than against a rival, because the buyer is already spending something on the problem. That spend is usually a person's time, a spreadsheet, or a service bought for a different purpose. Cost that workaround in their money and you have a ceiling built from evidence instead of from a guess.

Should the launch price be high or low

It depends on how quickly somebody could offer an equivalent. A decaying advantage argues for collecting margin while the window is open, with a written plan for coming down. A compounding advantage argues for share first. Both are legitimate and only one of them fits your product.

How do I test a price when the whole market is one city

Vary the offer rather than the price. A founding cohort rate with a stated cap, a time limited launch discount, a bundled first period: each gives you a read on sensitivity without putting a second price in front of the same audience. The price stays where it is and the thing that ends is the offer.

Is it worse to launch too high or too low

Too low is harder to fix, because the figure sits in the market as the reference point and raising it is a conversation with every existing customer. Too high usually shows up as a short pipeline, which is uncomfortable and faster to correct with a time limited offer.

How long before I can judge whether the price is right

Long enough to collect a number of decisions you would act on, decided before launch. Write down how many conversations or orders would make the reading real and hold to it, because a price changed every fortnight produces a market with no idea what your product costs.

What should go in front of the board

The six lines above, with the sources and dates for the competitive figures. A price with a floor, a ceiling and a written trigger is a decision. A price with a benchmark beside it is an opinion with a citation.

The first thing somebody does with a new price

They check whether it is reasonable, and increasingly they do it by asking an assistant rather than by opening three competitor tabs. The reply comes back with a sense of what this kind of thing costs and sometimes with named suppliers and figures, assembled from whichever pages happened to state a number plainly. For a product that launched last month, that answer is being formed now, out of your own pricing page and whatever else is published about your category. Which is a practical argument for publishing the number rather than a contact form: a page that states a price, a unit and a date is a page that can be quoted correctly, and a page that routes everything to sales can only be described. AI Knows Us records what those answers say about a category and which sources each one drew on, so a founder can see how a new price is being repeated before a buyer repeats it back to them in a negotiation.