Niche analytics on Kaspi means answering three questions about a specific product card: how many people buy it, what share of those orders you could capture, and how much you'd earn at the price currently at the top. Kaspi doesn't publish sales figures for product cards, so demand is estimated indirectly and as a range rather than a single number. Below is a method you can apply by hand to any card in 20–30 minutes.
In short
- Kaspi doesn't show how much a product card sells. So any demand estimate is a model built on assumptions, and only a test batch can confirm it.
- A handy indirect signal is how fast new reviews appear. But the share of customers who leave a review is unknown, so calculate a range from the start.
- The demand of the whole card matters less than your share of it: how long you'll realistically stay at the top with your minimum price and rating.
- Calculate money from the price that's at the top today, not from the "market average".
- How to decide which products to check in the first place is covered in "What to Sell on Kaspi". This article is about checking a specific card.
What to collect for a product card
For each card, create a row in a spreadsheet with these fields:
| Field | Where to find it | Why |
|---|---|---|
| Number of sellers | The seller list in the card, including «Ещё предложения» ("More offers") | Intensity of competition |
| Prices of the top 5 sellers | Seller list | Spread and level of dumping |
| Rating of the top 3 | Seller page | Who you compete with at an equal price |
| Delivery methods and times of the top 3 | Seller list | Who takes a top-3 spot for speed |
| Total reviews of the product | Product card | Scale of the sales history |
| Reviews in the last 30 days | Reviews by date | Current pace of demand |
| Reviews by month over the year | Reviews by date | Seasonality |
| Is there an official seller or distributor | Seller list | Risk of a purchase cost lower than yours |
| Your purchase cost and minimum price | Your own calculation | Entry economics |
Don't stop at the first three rows. The order of sellers comes from bonuses, price, rating, reviews and delivery speed (Kaspi Guide), and Kaspi hides everyone rated 3 stars or lower under «Ещё предложения» ("More offers") (Kaspi Guide for customers). The cheapest seller is sometimes sitting right there.
Kaspi notes that app updates roll out gradually and the order may differ. Collect data at the same time of day and, if you sell in several cities, for each city.
Step 1. Demand: estimating from reviews
The formula
Not every buyer leaves a review. If you know what share does, you can estimate the number of orders like this:
Card orders per month ≈ New reviews in 30 days ÷ Share of buyers who leave a review
The problem is that this share is unknown and depends on category, price and sellers. Kaspi has no public data on it. So calculate a range using several assumptions.
An example with a range
A hypothetical example. Say 30 new reviews appeared in the card over the last 30 days. We don't know how many customers leave reviews at all, so we'll calculate three options: 5%, 10% and 20%.
| New reviews in 30 days | Share 20% | Share 10% | Share 5% |
|---|---|---|---|
| 30 | ≈ 150 orders | ≈ 300 orders | ≈ 600 orders |
| 80 | ≈ 400 orders | ≈ 800 orders | ≈ 1,600 orders |
Yes, that's a fourfold difference. This method can't be more precise, and pretending it can is harmful. Decide by the lower bound. If the niche interests you even at 150 orders a month, test it. If it only makes sense at 600, you're betting money on the most optimistic scenario.
How to narrow the range: for your own products, calculate what share of your buyers left a review over recent months. That's your own calibration, and it's better than anyone else's numbers, though it isn't perfect for a different category either.
Seasonality
Scroll through a year of reviews and sort them by month. If most fell within three months, the product is seasonal. That means your batch has to arrive before the peak, and whatever is left afterwards will sit until next year. Allow for the lag too: people write reviews some time after buying, so the real sales peak was a couple of weeks earlier.
Step 2. Your share
The card's demand isn't your demand. Orders mostly go to the sellers at the top of the list. Your share depends on how much of the time you can be there.
What your share depends on
- Your minimum price versus the leaders' prices. If the leader consistently sits below your minimum price, your price-based share is close to zero.
- Rating. At an equal price, the better rating ranks higher. With a new store at an equal price, you'll be lower.
- Delivery speed. The fastest delivery earns a top-3 spot even without the best price.
- Bonuses. Sellers offering bonuses (Kaspi's cashback points) can rank higher at a higher price.
Share scenarios
Let's continue the example. The card brings in about 300 orders a month (the middle of the range from step 1), and you keep 1,200 ₸ per order.
| Scenario | Your share | Orders per month | Profit per month |
|---|---|---|---|
| Rarely on top, minimum price above the leader | 10% | 30 | 36,000 ₸ |
| You share the top with one or two sellers | 25% | 75 | 90,000 ₸ |
| On top more often than anyone, purchase cost below market | 40% | 120 | 144,000 ₸ |
We set the shares in the table ourselves; they're scenarios, not statistics. Look at the bottom row: if the niche is worth the effort even there, you can move on.
Step 3. Competition: price spread and structure
Price spread
Spread = (Price of the 5th seller − Price of the 1st seller) ÷ Price of the 1st seller × 100%
If prices are squeezed within 1–2%, the card almost certainly has bots running and 1 ₸ dumping going on. There, purchase cost and reaction speed win; everything else is secondary.
A 3–10% spread leaves breathing room. It looks like some sellers change prices by hand or deliberately protect their margin.
More than 10% means either someone at the top is selling at a loss to dump stock, or the leader has a completely different purchase cost. Look at the card again in a week. If the picture hasn't changed, that's how the market works, not a random spike.
We chose these thresholds ourselves as a guideline; Kaspi has no such figures.
Who's on top
If an official seller or distributor is at the top, your purchase cost is almost certainly higher. You can't win a price war against them; what's left is delivery speed, bonuses and advertising. If the top spot belongs to a store rated below 4.5 with a dozen reviews, that's soft competition.
How many sellers
The number on its own means little. Ten sellers with prices spread widely and low ratings are easier than three strong stores at the same price with a 4.8 rating.
Step 4. Money at the top price
Calculate profit per order at the price the first seller has right now, minus 1 ₸. Include every item: purchase cost to warehouse, the Kaspi fee including VAT for your category, Kaspi Delivery by weight and order value, tax, packaging, advertising and a returns reserve. The full calculation is in "How to Stop Selling at a Loss on Kaspi", and the minimum-price formula is in "Price Dumping on Kaspi".
Cushion = (Top price − 1 ₸ − Your minimum price) ÷ Top price × 100%
- Negative cushion. In the current market you'd be selling at a loss. The niche is closed until you find a cheaper purchase cost.
- Cushion of 0–3%. You can get to the top, but any downward step by a competitor will knock you off. Only suitable with a strong rating or fast delivery.
- Cushion above 3%. There's room for dumping and advertising.
The thresholds are rough. Adjust them to how volatile prices are in your niche.
Step 5. The decision and a trial batch
Put everything into one row and decide.
| Signal | Green | Yellow | Red |
|---|---|---|---|
| Demand, lower bound | The niche is attractive even there | Attractive in the middle of the range | Attractive only at the upper bound |
| Your share in the worst case | Covers launch costs | Pays back in 2–3 months | Doesn't pay back |
| Price spread | 3–10% | 1–3% | Over 10% with no explanation |
| Who's on top | Stores with no purchasing advantage | Strong stores | A distributor or official seller |
| Cushion to minimum price | Over 3% | 0–3% | Negative |
| Seasonality | Steady demand, or more than two months to the peak | Peak is soon | Peak has passed |
If you see even one red, either walk away or solve that specific problem first. Two yellows mean you can try, but with a smaller batch.
Trial batch and stop criteria
Trial batch ≈ Your store's expected orders over 2–4 weeks in the worst-case scenario
Before launching, write down the criteria that will make you stop:
- in 2 weeks you sold less than half of what the worst case predicted;
- the average selling price is below the minimum price, which means the calculation was wrong;
- quality-related returns are approaching 2%, the threshold above which Kaspi takes action (Kaspi Guide).
After launch: Kaspi's conversion rating
30 days after a product is added, and once it has 2,000 views, a conversion rating appears in the «Маркетинг» (Marketing) service: from 1 to 6.9 is low, from 7 to 10 is high (Kaspi Guide). It's the first official signal of how the product sells within its category. A low rating at a normal price is a reason to work on the card before buying more stock.
Example: analyzing a single card
Let's work through one card from start to finish: a humidifier. The numbers are hypothetical; this isn't a client case.
- What the card shows. Seven sellers. The top five prices are 18,990, 19,050, 19,200, 19,400 and 19,900 ₸; the top three ratings are 4.7, 4.6 and 4.8. Ordinary stores are at the top, with no distributor among them. 24 new reviews in 30 days, and last year most of them came in October–January.
- Demand. From 24 reviews at shares of 20%, 10% and 5%, you get 120 to 480 orders a month. We take the low figure, 120.
- Your share. A new store's rating is lower than these three, and at the same price you'll sit below them. You can realistically count on 10–15% of the card, which at the lower bound of demand is 12–18 orders a month.
- Spread. (19,900 − 18,990) ÷ 18,990 ≈ 4.8%, the middle of the scale. One of the sellers seems to be managing their price by hand.
- Money. Your minimum price came out at 17,600 ₸. The cushion to the top price is (18,989 − 17,600) ÷ 18,990 ≈ 7.3%, so there's room to maneuver.
- Season. It's mid-October, and the peak is just starting. If your batch takes longer than two or three weeks to arrive, you'll land in the middle of the season.
- Verdict. Demand at the lower bound is modest, but the price cushion is good and the season is ahead: share is yellow, money is green. We test with a batch of about 20 units. The stop signal is fewer than 6 sales in the first two weeks.
Note that no figure in this analysis is "knowledge"; each is an estimate with assumptions. The value of the method is that the assumptions are written down, and in two weeks you can check them against reality.
Common mistakes
- Treating a demand estimate as fact. Kaspi doesn't publish sales; any number is a model.
- Calculating on the upper bound of demand. The decision has to survive the lower bound.
- Confusing the card's demand with your own. Orders go to whoever is on top.
- Calculating profit at the average price. People buy from the first seller; calculate at their price minus 1 ₸.
- Looking only at the first rows of the list. Low-rated sellers are hidden under «Ещё предложения» ("More offers").
- Collecting data only once. Prices and sellers change; repeat a week later.
- Not writing stop criteria down in advance. After buying stock, there's always the urge to "wait a little longer".
Checklist
- All fields from the table have been collected for the card.
- Demand is estimated as a range using several assumptions.
- The decision has been tested against the lower bound of demand.
- Your share scenarios are calculated.
- The spread of the top-5 prices is calculated.
- It's clear who's on top and what their purchase cost is likely to be.
- The cushion from the top price to your minimum price is calculated.
- Seasonality from the year's reviews is taken into account.
- The trial batch size is based on the worst-case scenario.
- Stop criteria are written down before buying stock.
How to automate it
Analyzing one card by hand takes 20–30 minutes. A hundred takes a week if you don't get distracted. SaleScout's niche analytics module collects card data on its own, without scrolling through each one, but the decision by the criteria in this article is still yours. Once a product is on sale, SaleScout's Kaspi bot can manage its price: it checks prices every 20 seconds and never goes below your minimum price. You can try the bot free for 7 days, no card required.
FAQ
What is niche analytics on Kaspi?
It's assessing a specific product card before buying stock: how many people buy it, what share of orders you could capture, how strong the competition is and how much you'd earn at the price currently at the top. Kaspi doesn't publish sales, so demand is estimated indirectly and as a range.
How can I estimate demand for a product on Kaspi without tools?
Count the new reviews from the last 30 days and divide by the share of buyers who leave a review. Nobody publishes that share, so calculate several options, for example 5%, 10% and 20%, and decide on the lower bound. Calibrate the share using your own products.
How many competitors in a Kaspi product card is too many?
The number on its own says little. Look at the spread of the top-5 prices, the leaders' ratings, and whether a distributor or official seller is at the top. Three strong stores at the same price can be harder than ten weak ones.
How do I know whether I'll make money in a Kaspi niche?
Calculate profit at the first seller's price minus 1 ₸ with all costs: the fee including VAT, delivery, tax, advertising and returns. If the top price is below your minimum price, you'd be selling at a loss in the current market.
Can I trust "best niches on Kaspi" lists?
As a source of ideas, yes; as grounds for buying stock, no. A list doesn't know your purchase cost, rating, delivery speed or budget. Check every idea from a list against the card: demand as a range, share, competition and money at the top price.
What size trial batch should I buy when entering a niche?
As much as your store would sell in 2–4 weeks in the worst-case scenario. And before you pay, write down when you'll stop: less than half the expected volume sold in two weeks, the average selling price dropping below your minimum, or quality returns creeping up toward 2%.
What is the product conversion rating in Kaspi Marketing?
It's how Kaspi shows how often your product is bought compared with its category. You'll wait a month for it: it takes 30 days from when the product was added and 2,000 views. The scale is simple: 1 to 6.9 is low, 7 to 10 is high, and it's the first official signal about a product after launch.
Sources
All links checked Oct 15, 2026. The sources are in Russian; original titles are given with English translations.
- Kaspi Guide for partners. «Как располагаются продавцы в карточке товара?» ("How are sellers ranked in a product card?") — https://guide.kaspi.kz/partner/ru/shop/goods/general/q3190 (updated Jun 15, 2026)
- Kaspi Guide for customers. «Почему в карточке самая низкая цена, но я не могу найти её у продавцов?» ("Why does the card show the lowest price, but I can't find it among the sellers?") — https://guide.kaspi.kz/client/ru/shop/choice/q18442
- Kaspi Guide for partners. «Почему меня отключили от Магазина на Kaspi.kz?» ("Why was I disconnected from the Kaspi.kz Store?") — https://guide.kaspi.kz/partner/ru/shop/conditions/q2920
- Kaspi Guide for partners. «Что такое оценка конверсии товара?» ("What is a product conversion rating?") — https://guide.kaspi.kz/partner/ru/marketing/advert_goods/q3350 (updated Jun 30, 2026)
Bota, SaleScout analyst
SaleScout is our company's product. Kaspi and Kaspi.kz are trademarks of Kaspi.kz JSC; SaleScout is not affiliated with Kaspi.
