Product Matching Farm for Retailers

Product Match Farm for Retailers

Improve your data collecting and processing
When you’re building competitive pricing, one of the most difficult tasks is getting qualitative competitive data. Low match rates cause wrong conclusions, ineffective decisions and lost profits. Let’s take a look at how you can improve data collecting and processing with multiple quality checks for errors and anomalies.
Get Demo
Competitive data crawling

Matchings updated*:


Scan according
to schedule

*last 24h
Product match and data processing

First of all, we need to define what is a product match. Let’s suppose you’re selling headphones, and EVEN EarPrint H2 Wireless is one of your best sellers, so you want to set the optimal price for it. To do this, you have to find these exact headphones at the websites of your competitors and compare their prices.

And here the mess starts...

When matching products, you check many parameters for comparison to ensure the products are identical: model name, version, color, technical characteristics, ISBN, availability, discount, etc. When all parameters of a product are absolutely identical, it’s an exact match. If there’s a slight difference, for example, in color, we call it a similar match.

Popular challenges of product match retailer faces daily

  • titles of a product may be not identical, but semantically similar
  • some incoming data may be noisy or missing
  • primary images may be different
  • price may be highly different which indicates a mismatch
  • the data may be not timely or relevant at the time decisions are made

Data matching

Pain points

The economic effect of the business strongly depends on the quality of data, which is a basis for analytics and tactical decisions. Presume you have 5 competitors and 100 products in your shop. Even if product comparisons are 80% accurate, the other 20% may cause serious losses, meaning wrong prices for at least 20 items (percentage of comparisons is one of the main indicators of data quality).

But what if you sell thousands of thousands of products, not being sure the competitive data is full and accurate?

Competitor monitoring dashboard

Free White Paper Manual vs. Automated Matches Download Now

How to set an ideal matching process

To provide the best data quality, mixed type of product matching should be used: automatic comparisons to quickly collect large amounts of data and human-based comparisons to match products by more parameters.

This approach helps achieve the highest possible percentage of relevant comparisons with no mistakes.

Data comparison

Product Match at Competera

Here is how our Product Matching Farm works. All comparisons are initially automatic. We built an algorithm allowing quick collecting and comparison of all products with the help of an automatic script. But robots are not entirely perfect! That’s why we additionally use visual comparison.

How Competera handles complex task of product matches

While automatic matching allows full crawling of a website right after you add a new competitor to the dashboard, visual matching has its own advantages.

Comparison of any product categories in different industries (fashion, jewelry, etc.).
Visual comparison of products is possible even for extremely secured or complex websites of your competitors.
Ready-made matches 2 hours after the products were added to the Competera dashboard (automatic matching may take up to 5 days to set up).
Large capacity: 50K+ high-quality comparisons performed per day.

With the special algorithm of product search and multi-stage verifications, we developed our own constantly replenished data catalog of comparisons which improves capacity and verifies the results.

Our product match team has been growing over time. It is experienced in different industries and markets and is motivated by KPI-system and bonuses.

The main tasks of Competera Product Match Pool

Let you make correct pricing decisions on the basis of high-quality data is our main task. That’s what our Product Match Farm is performing to achieve this goal:

  1. We constantly improve the results of automatic product matching. Visual matching helps create learning schemes for algorithms of automatching. It improves automatching quality and increases the capacity of Competera.
  2. We compare products manually when it’s technically impossible to match them automatically. It concerns complex web pages with many buying options, e.g. perfumes or food for animals.

With the mixed approach, we deliver high percentage of comparisons by many parameters, low amount of zero (non-found) prices and very low amount of errors in the collected data.

Get the 95% accurate competitive data

Quick product comparisons, completeness and quality of data

Try Competera
Data crawling