How to choose the best pricing intelligence software for e-commerce? How to avoid the bad choice?
In this article, we’ve listed 7 essential questions you should ask any service provider offering pricing intelligence solutions for online retailers and brands. After all, the results you get will depend directly on the partner you choose.
7 Key Questions About Data Quality to Ask Your Pricing Intelligence Service Provider
- What is the expected discovery/match rate?
- How often is the data collected?
- Is the technical specification agreed upon? Which products will be considered identical for price comparison?
- How does the service provider find products that your competitor did not have in their catalog yesterday but has today?
- How are products with incomplete or shortened names compared?
- How is the comparison of analogs carried out?
- If two products have already been matched, are they checked again?
1. What is the expected discovery/match rate?
What is the discovery/match rate
Discovery or match rate is the percentage of products found by the pricing intelligence software that exist both in your assortment and in your competitors’ assortment.
Say, you have 500 products and your competitor has 700, and the overlap between your assortments is 100 products. Then if the software identifies 99 out of these 100 products on the competitor’s site, the discovery rate is 99%.
Why it matters
If the discovery rate is low (below 90%), you won’t get a complete picture of the market. You might mistakenly believe your pricing is competitive, while missing better offers from competitors simply because the pricing intelligence software failed to find them. This directly leads to flawed pricing decisions.
Many pricing intelligence software providers claim, “Our discovery rate is 99%,” yet they only match products based on exact names, article numbers, or barcodes.
What happens in this case? If 30 out of your 100 products exactly match the competitor’s product names, the system will find those 30 and report a 100% discovery rate (based on exact matches). However, the remaining 70 products — which are also available both on your website and the competitor’s — won’t be found if their names differ by just one or two characters. As a result, the actual discovery rate will be 30%, not 100%.
Let’s take the “Laptops” category as an example. In many online stores, part numbers are often missing from product names. That’s why accurate matching requires taking into account technical specifications and, in some cases, even product configuration — which may vary by operating system version, presence or absence of a fingerprint scanner, screen type, and so on.
If the pricing intelligence software matches products based solely on names, this will lead to:
a) a low actual discovery rate;
b) a high percentage of false matches, where the system mistakenly treats different models as identical.
Another common case is when products are matched solely based on URL addresses. For example:
- You have a Samsung 55″ TV.
- The system finds a page on the competitor’s website with what seems to be the same TV: www.yourcompetitor.com/samsung-55-qled.
- It then stores this URL in a pair like: “your product = this product on the competitor’s site.”
The problem is that the product URL on the competitor’s website can change — for example, if the product page is recreated. As a result, the product still physically exists on the competitor’s site, but the system will no longer find it because it continues to search using the outdated link. This means that over time, the discovery rate will decline.
What answer are you looking for
A good discovery rate is 95% or higher, but it’s important to understand how exactly the system matches products. The number alone doesn’t tell the whole story — the search methodology matters just as much.
Red flags to watch out for
🚩 If a pricing intelligence software guarantees a 99% discovery rate but cannot clearly explain how it achieves that number, it’s more of a red flag than a reason to celebrate.
Conclusion
“Our discovery rate is 99%” sounds impressive, but to be sure, don’t just ask, “How many products do you find?” — ask, “How do you find them?” It’s important to understand how each service’s discovery methodology actually works and how the rate is calculated. It’s also worth testing this in practice.
How Pricer24 solves this challenge
We match products based on a combination of key parameters simultaneously, with the set of these parameters tailored according to your technical specification.
For example, when comparing laptops, we take into account article numbers/part numbers, product names, and, in cases of unclear matches, general technical specifications: RAM size, processor (CPU model, generation, frequency), GPU type, storage capacity, as well as specific features indicating the product’s configuration.
For perfumes, additional attributes may be included in the comparison, such as bottle volume, whether it is a tester or the main product, original or replica, and so on.
So, even if a product has a complicated name or a non-standard code, we can still find it.
Pricer24 delivers a 98% discovery rate. Achieving such a high figure is truly challenging. It requires extensive depth and experience working across different industries, as each product category has its own specifics.
2. How often is data collected?
What this means
The frequency of data collection indicates how often the system updates information about prices, product availability, promotions, and competitors’ assortments. This can be once a day, several times a day, or less often — depending on the platform’s technology and settings.
Why it matters
Prices and product availability in the market are constantly changing, especially in dynamic categories such as household appliances, cosmetics, children’s products, and electronics. If data is updated infrequently, you will be working with outdated information, which can lead to wrong decisions.
What answer are you looking for
It depends on the category, but data should be updated at least once a day for all your products.
Red flags to watch out for
🚩 Promises of “real-time” updates without technical explanation of how this is achieved often turn out to be a marketing gimmick.
🚩 The service provider avoids answering questions about update frequency or only gives vague statements (e.g., “high frequency”) without specifics. This is also suspicious.
🚩 Lack of an update history with timestamps showing when updates occurred in your dashboard.
How Pricer24 solves this challenge
We update data several times a day.
The exact number of updates is defined in the client’s technical specification and may vary across product categories according to business needs.
Updates cover:
- all products within the selected categories and brands, across all competitors specified by the client in the technical specification;
- new items that appear on competitors’ websites.
This allows you to have an up-to-date market picture every day, respond quickly, and make informed decisions.to have an up-to-date market picture daily, react quickly, and make well-informed decisions.
3. Is the technical specification agreed upon? Which products will be considered identical for price comparison?
What this means
It is in your best interest to agree with the service provider on clear rules that, among other things, define the criteria for product identity during matching — what should be considered the same and what should not.
Why it matters
The accuracy level of product matching depends on how the criteria for product identity are defined.
Here is a practical example: in the “Perfumes” category, one fragrance can be an official product, a tester, a promotional version, or even a replica — and these can differ significantly in price. One business may consider these variations equivalent, while another may not. Accordingly, the matching rules should be adapted to the client’s business logic.
Another example: white and gray products from the same manufacturer, such as smartphone models. Their prices may differ, even though they are essentially the same products. One business may treat these products as identical, while another may not.
Without a clearly agreed technical specification defining what the client considers identical products and what not, the risk of incorrect matching increases significantly.
What answer are you looking for
“Yes, we agree on the technical specification to ensure that the product matching rules align with your business logic and are as accurate as possible.”
Red flags to watch out for
🚩 Lack of a process for agreeing on the technical specification regarding this issue.
🚩 Use of standard or generic rules for product identity without adaptation to your category and/or business logic.
🚩 No possibility to adjust the criteria during the process.
How Pricer24 solves this challenge
We develop custom rules for each category and product type — tailored to every business we work with. In the process, we:
- analyze the specifics of the category;
- agree on the matching logic with the client;
- create an internal system of rules and exceptions that can be adjusted as needed.
4. How does the service provider find products that your competitor did not have in their catalog yesterday but has today?
What this means
The product assortment on websites (yours and your competitors’) is constantly changing: new items appear, old ones disappear. It is essential to track these changes promptly.
Why it matters
If the analytics service provider does not track the appearance of new products in a timely manner, you may miss critical changes in the category.
This is especially true for new arrivals in the assortment. Imagine you add new products to your catalog today. The pricing intelligence software searches for them in the competitor’s catalog but doesn’t find them. Those same products appear in the competitor’s catalog two days later, but you no longer receive information about them.
Typically, competitors’ store assortments overlap by about 70%, while 30% of their catalogs differ. Accordingly, if you have 1,000 products and 5 competitors, you should consider that there are approximately 1,500 products that can potentially appear at any moment and replicate your products. If your service provider lacks an approach to detect such products, you won’t have a complete market picture.
What answer are you looking for
The service provider should have a process that enables tracking changes in catalogs and responding to them promptly: creating new product links and highlighting problematic situations.
Red flags to watch out for
🚩 The pricing intelligence software cannot clearly explain the approach it uses to track the appearance of new products on competitors’ websites.
How Pricer24 solves this challenge
At Pricer24, we don’t limit ourselves to only the products in your assortment. The system automatically downloads all products daily within the selected categories and brands on competitors’ websites — including those that don’t yet have a match in your catalog. Thanks to this, we capture the appearance of new SKUs and compare the entire competitor category with your entire category. If a competitor launches a new product today, you will receive information about it within one business day.
5. How are products with incomplete or shortened names compared?
What this means
Different stores label products differently, often using abbreviations in the name or omitting parts of the name. If the system searches for matches by name, there is a high risk it won’t find everything, and you won’t get a complete picture.
Why it matters
You may miss some information, which is especially frustrating when it comes to new products on competitors’ websites that make their assortment more attractive. These are often items for which managers save time by not formatting the names according to standards.
What answer are you looking for
The service provider has a methodology for comparing products with incomplete or abbreviated names and can explain it clearly.
Red flags to watch out for
🚩 The service matches products only by name or by name and code (for cases when not all your products or competitors’ products have such a code).
🚩 The analytics provider lacks processes for studying the client’s catalog structure.
How Pricer24 solves this challenge
- Every day, we analyze all products in the relevant category on each of your competitors’ websites and find the most accurate matches between their entire assortment and yours. This allows us to identify products that were previously unmatched but can be linked.
- We study the structure of your catalog, the specifics of product card filling, abbreviations, and variations in names and attributes — everything that can speed up establishing connections. Our specialists understand which words in product names are significant, which are not, whether category prefixes are used, and how brands are identified. Over time, manual search becomes very fast. When your competitor adds a new product, we know exactly where in your catalog and under which filters to look for its match.
- We continuously enrich product information. If additional product data is available on the website, we add it to the product name — for example, tire size or laptop memory capacity.
This enables quick orientation in product characteristics even in cases where the system does not find exact matches or has doubts. Our specialists know where the relevant attributes for comparison are located, allowing us to avoid false matches.
For example, when comparing cat foods, the system may detect two very similar products that differ only by the cats’ age — “for kittens” and “for adults.” In such cases, the system highlights low confidence in the match. The matcher immediately reviews the product card, identifies the relevant age attribute, and determines that these products are different and therefore should not be compared.
6. How is the comparison of analogs carried out?
What this means
Sometimes two products can look identical — the same size, color, and texture. But one is from a well-known, quality brand, while the other comes from a no-name brand. Visually, they are the same product, but brand strength changes how consumers perceive its value.
Why it matters
For example, facing tiles from an unknown brand can be up to 40% cheaper than similar branded tiles. At the same time, buyers may consider them as alternatives. If you don’t account for acceptable price deviations due to brand value when comparing such products, you may end up with a distorted competitive picture.
Another situation: a manufacturer creates two part numbers for the same laptop model to supply a batch to a specific retail partner under exclusive terms. The laptops are identical, although the part numbers and prices differ. You won’t notice this if analog comparisons are done using a fixed template.
Each retailer defines competitive products differently — depending on their own strategy. You may consider a product from another brand as an analog even if it is 10% more expensive than yours due to brand strength, or you may not, comparing products only within your price segment. Direct “price-to-price” comparison in the latter case is not representative, as it does not account for the brand value factor, which is a very real criterion for customers when choosing. A buyer might think: “I’m considering two phones — a more expensive one from the most popular brand and a cheaper one from a less popular brand — and I’m unsure which to choose.” You need to see which analogs your products compete with not only by price but also by value.
What answer are you looking for
The analytics provider should be able to explain how they take brand and product value into account when comparing analogs and provide examples. You can describe your specific case to a potential service provider and listen to how they handle it to assess the depth and thoroughness of their approach.
Red flags to watch out for
🚩 The pricing intelligence software compares identical products but lacks a methodology for analyzing similar ones.
How Pricer24 solves this challenge
Precisely for such cases (and more), we offer custom product comparison when the standard approach doesn’t work.
We are developing functionality that will allow analogs to be taken into account. For example, if a product from a strong brand is more expensive by a certain percentage but remains a relevant competitor, the system will be able to “highlight” this.
7. If two products have already been matched, are they checked again?
What this means
The links between identical products in your catalog and the competitor’s catalog are not fixed. If they are not updated from time to time, data accuracy and completeness suffer because changes are constantly happening on both your site and the competitor’s site.
Why it matters
Product matching initially follows this logic: the system searches the competitor’s catalog for the best match for each product. But tomorrow, an even better match with a different price may appear on the same site.
For example, a store may create two product pages for the same item but with different prices to violate the MSRP/MAP without consequences, knowing that the manufacturer and competitors will analyze the price on its old page for some time, where it remains within the MSRP limits.
If matches are not systematically rechecked, incorrect links will accumulate in the system — and this will affect all subsequent analytics.
What answer are you looking for
“Yes, we regularly recheck product matches” — followed by an explanation of how this is done.
Red flags to watch out for
🚩 Lack of repeated checks on matched products. If the service provider does not regularly review already matched products, errors will accumulate and affect analytics.
🚩 Overreliance on “guaranteed” matching accuracy solely through AI. No algorithm is perfect. If a pricing intelligence software promises 100% matching accuracy without human involvement, this is also a warning sign.
🚩 Absence of different algorithms and approaches for various categories. For example, matching by article number works well when comparing laptops, but for clothing, article numbers are often entered incorrectly, leading to poor matching.
How Pricer24 solves this challenge
- When initially setting up a client’s account, we don’t just perform automatic configuration but also conduct a manual review that detects errors in brands, confusion between similar products, and so on. By understanding why these errors occur, we can anticipate similar situations and adjust the algorithm’s logic to proactively prevent potential future mistakes.
- Every night, AI-driven algorithms run to recheck the matching. If a more accurate match appears, the system flags it as “review,” suggesting a new option. Our specialists see these flags daily and promptly correct incorrect matches, keeping everything up to date.
- We use checks for multiple matches — when one client product is matched with several competitor products simultaneously. This either signals a duplicate or provides an opportunity to find and fix an error.
- A separate tool for our specialists is reports that highlight significant price differences. Such cases also trigger additional review of matches (whether the price difference indicates a false match).
Thus, we have several levels of post-matching verification, ensuring high-quality data control throughout the entire cooperation with the client.
Conclusion: How You Should Choose a Pricing Intelligence Software
Accuracy, completeness, and data relevance are critically important factors for price analytics.
If the system makes mistakes or misses some items, it will directly affect your pricing decisions.
Why do we do this better than many others on the market? Because we take into account dozens of nuances depending on the client’s specifics, have an experienced team, and constantly improve the technical side of the process.
But how can you be sure of this? We suggest testing it in practice.
The best way is to give all service providers the same test task:
- 100 of your products;
- 5 competitors;
- task — find the maximum number of exact matches.
Then simply compare how many matches each provider actually found. Even if the difference is only 3–5 products, it already speaks to the quality and approach. On a real scale, such a difference can amount to hundreds of SKUs. It’s also important to understand that over time this indicator only decreases.
We invite you to a demo — we will show the results on your products and give you the opportunity to see the quality firsthand.
