Parsing for business. How it works and what it’s for

Though there are negative reviews from some businesses, we believe that parsing is profitable and user-friendly, it allows collecting the necessary information in the network for taking important administrative decisions. How parsing works and what its use in business, so we decided to analyze this material.

We excluded boring definitions from the material, as they are not always clear, despite their diversity in the network. We faced the need of using this service for our clients a year ago and recommended them to try it out with specialized companies and were convinced of its effectiveness.

What is parsing?

It is the data collection on the Internet using a specialized software product. The most popular tools for business:

- price monitoring;

- trademark infringement monitoring;

- monitoring of the description of the goods;

- monitoring of goods turnover.

We often faced a negative attitude towards parsing. The online store owners are dissatisfied that their data is monitored by robots and stored.

You would agree that Yandex and Google are analyzing your site for page indexing. And you even want this indexing to be more frequent and search positions to be higher. This is almost parsing, but no one calls the work of search engines like that.

The same happens with commercial parsing, resulting in you get the collected data from the network.

Parsing in practice:

Last year we consulted an online shop of special tools, which held not the highest position on the market. The assortment of the shop was huge, 90% of which had weak liquidity. The stale goods limited working capital and led the business to a high risk of closure due to losses.

The key offer was an assortment reduction. But that wasn't enough. It was important to sell out the existing assortment and to define a strategy for the future management. The analysis of network requests via Yandex and Google gave no results, as people enter the general product specifications, such as "power jigsaw", but they buy specific models or specific manufacturers. Then we proposed to use Parsing.

We launched Parsing on the largest online stores in a similar area, all of them publish their information on balances. Using parsing daily, we started to measure the change of this indicator, as a result, we were able to determine the turnover of goods by specific models. Then, based on the frequency of the product search by category in Yandex and Google, we formed an advertisement for the most popular range, exactly in those regions where it is sought.

The result was stunning, though there were non-seller, we were able to start only 30 new items + 48 items from the existing range have increased in sales by 27. In three months, due to the top positions, the online store was able to start selling non-sellers with negative profitability, while maintaining a minimum business profitability.

After 9 months, the business got rid of all non-seller and continued to sell effectively the top positions, which we defined. By this period, sales increased 47 times, from the beginning of our work.

It's good to grow practically from the scratch, you'd say? And you'll be right to some extent, but the effectiveness of parsing is easily measured. And we are confident that in a bigger business, the correct use of their results will ensure a decent growth in sales.

What conclusion can be drawn from this experience?

In our opinion, everything is simple: You own the information, you own the world. And parsing is just a method of data collection.

And did you used parsing? What areas? 

Why Parcing is so unattractive?

Those who have ever registered for Avito and placed an ad received an amazing bonus in the form of nefarious calls from spammers. And who loves them? No one. Well, they were all caught by the same parsing.

Another popular way was to copy information. You are running an online store, spending a lot of resources to create a detailed and optimized for SEO product description, and you have it all copied and hosted by your competitors in three days.

However, we believe that any work can be used for both good and bad, you can't get away from it, you will have to accept it.

Parsing pitfalls

When using data collection in marketing, it is important to understand that all information is collected by the robot and although parsing companies adapt the program to your needs, you may still encounter errors.

For example, when we were monitoring online prices for auto parts under our order, we faced the problem of items and titles mismatching on different parts sites with the same applicability. Processing this data was not easy, and the most important thing was that the parsing did not solve these problems anymore. As a result, a lot of work was done to adapt the analytics to the processing, and even though we processed 20% of the material by hand, and 30% was thrown out without processing.

So, if you need data from the network, parsing is the best way to collect. But you should carefully evaluate the probability and the processing cost of the received data.