The value of web data is increasing in every industry from retail competitive price monitoring to alternative data for investment research. As the trusted research firm, Gartner, stated in their blog:
“Your company’s biggest database isn’t your transaction, CRM, ERP or other internal database. Rather it’s the Web itself…Treat the Internet itself as your organization’s largest data source.”
In fact, the internet is the largest source of business data on earth and it’s growing by the minute. The infograph below from Domo shows how much web data is created every minute from just a few websites out of a billion.
This article will walk you through a simple process of getting web data using Import.io data extraction software. First, let’s look at other uses of web data in business.
How do businesses use data from a website?
Competitive price comparison and alternative data for equity research are two popular uses of website data, but there are others less obvious.
Here are a few examples:
Teaching Movie Studios how to spot a hit manuscript
For StoryFit, data is the fuel that powers its predictive analytic engines. StoryFit’s artificial intelligence and machine learning algorithms are trained using vast amounts of data culled from a variety of sources, including Import.io extractors. This data contributes to StoryFit’s core NLP-focused AI to train machine learning models to determine what makes a hit movie.
Predicative Shipping Logistics
ClearMetal is a Predictive Logistics company using data science to unlock unprecedented efficiencies for global trade. They are using web data to mine all container and shipping information in the world then feed predictions back to companies that run terminals.
XiKO provides market intelligence around what consumers say online about brands and products. This information allows marketers to increase the efficacy of their programs and advertising. The key to XiKO’s success lies in its ability to apply linguistic modeling to vast amounts of data collected from websites.
Virtuance uses web data to review listing information from real estate sites to determine which listings need professional marketing and photography. From this data, Virtuance determines who needs their marketing services and develops success metrics based on the aggregated data.
Now that you have some examples of what companies are doing with web data, below are the steps to getting data from a website.
Steps to get data from a website
Websites are built for human consumption, not machine. So it’s not always easy to get web data into a spreadsheet for analysis or machine learning. Copying and pasting information from websites is not feasible.
Web scraping is a way to get data from a website by sending a query to the requested pages, then combing through the HTML for specific items and organizing the data. If you don’t have an engineer on hand, Import.io provides a no-coding, point and click web data extraction platform that makes it easy to get web data.
Here’s a quick tutorial on how it works:
Step 1. First, find the page where your data is located. For instance, a product page on Amazon.com.
Step 2. Copy and paste the URL from that page into Import.io, to create an extractor that will attempt to get the right data.
Step 3. Click on Go and Import.io will query the page and use machine learning to try to determine what data you want.
Step 4. Once it’s done, you can decide if the extracted data is what you need. In this case, we want to extract the images as well as the product names and prices into columns. We trained the extractor by clicking on the top three items in each column, which then outlines all items belonging to that column in green.
Step 5. Import.io then populates the rest of the column for the product names and prices.
Step 6. Next, click on Extract data from website.
Step 7. Import.io has detected that the product listing data spans more than just this one page, so you can add as many pages as needed to ensure you get every product in this category into your spreadsheet.
Step 8. Now, you can download both the Images as files and the Product Name and Price as a spreadsheet.
Step 9. First, download the Product Name and Price into an Excel spreadsheet.
Step 10. Next, download the images as files to use to populate your own website or marketplace.
What else can you do with web scraping?
This is a very simple look at getting a basic list page of data into a spreadsheet and the images into a Zip folder of image files.
There’s much more you can do, such as:
- Link this listing page to data contained on the detail pages for each product.
- Schedule a change report to run daily to track when prices change or items are removed or added to the category.
- Compare product prices on Amazon to other online retailers, such as Walmart, Target, etc.
- Visualize the data in charts and graphs using Import.io Insights.
- Feed this data into your internal processes or analysis tools via the Import.io APIs.
Web scraping is a powerful, automated way to get data from a website. If your data needs are massive or your websites trickier, Import.io offers data as a service and we will get your web data for you.
No matter what or how much web data you need, Import.io can help. Talk to a data expert to determine the best solution for your data needs.