SaaS and Stand-alone Web Data Extraction Tool Comparison

Pros

Cons

Infrastructure Independence

Theoretically, the “cloud-based” solution can run from any OS and any browser. But today cloud-based solution requires a download file to your personal computer (or local storage) or install a browser extension. To sum up they are not platform or machine independent.

You don’t have to host anything yourself

It means accessing your data is fast and easy, for your computer, or others on your network. All the website page scrape, data extraction, and parse get handled on someone else’s server.
But you need to host your own business site anyway, isn't it?

You don't manage web proxy and captcha solvers.

If you don't need to extract more than 2k pages per time, the browser extension version is good enough.

Restrictions on available data

There may be restrictions applied on what websites you can download data from sites that use AJAX or complex JavaScript sceneries cannot be handled by most cloud solutions. For example, it can be a forum with multiple pages.

Performance limitation

"Data extraction takes over 10 hours without ending". Performance issue is a kind of frequently asked question by customers of SaaS. It could be a serious limitation.

No options for fine tune out results

There are no options to sort out your search results, no option for fine tune out results in most cloud solutions.

Usually, there is no an easy way to directly fed your data through to a database. Finally, you can have difficulties opening/manipulating the delivered files (+2M rows) in excel.

UI annoyances

Usually, the robot editor has various UI annoyances. If you need a "business volume" of data extraction, setting up your accounts will not easy. It will be done manually by SaaS support team and it can take a couple of weeks to get all done.

If SaaS team constantly push out updates and UI re-designs. It will make using the platform very confusing as its always changing and you constantly have to learn how to use it again. You have to keep deleting and reinstalling their plugins when they do updates as the old ones no longer work.

SaaS may intrude your data privacy

You might remember this particular privacy scandal when data-mining company exposed voter info. The data was leaked accidentally by a data-mining company that was responsible for compiling it. Moreover, who would have thought that a misconfigured spambot would be able to leak over 700 million email addresses and phone numbers in one of the biggest data breaches ever.

Overall, those privacy scandals brought to light how pretty much data in third-party services have privacy endangered.

If the SaaS provider shuts down your business operations can be broken down too and your data lost. For instance, in 2016 the Kimono web service was shut down and and the cloud service has been discontinued.

Pros

Cons

Control over data extraction software

If you build your own infrastructure you will get security and confidentiality of your data, the high availability, complete transparency, and guaranteed confidentiality. Dedicated servers offer ultimate flexibility. It’s your device: you can install whatever you like whenever you need it. No problem.

High performance

The custom designed scraper software offer the high performance of mass data extraction.

Proper workflow

Development of your independent software would allow you to build a proper workflow so you can manage the entire data retrieval process. You will get your data in a format most useful to you. Completely customizable.

No limits on available data

Custom scraper allows to intelligently extract and timely update structured data from any interactive website. Extract data from sites that use anti-scraping technology. JavaScript, AJAX, complex JavaScript scenaries, or any dynamic website - your custom software can cover all.

Infrastructure management

You get full responsibility for your own infrastructure management and monitoring. It will include server, proxy and captcha solver management.

Technical expertise

Harder to set up for a beginner if you do it by yourself. Technical expertise is necessary to connect, install and configure the software required for your system. Hosts often provide some level of assistance but you’re largely on your own. You might require additional staff costs to install, monitor and maintain your server.

Cost of SaaS Scraping Solution

Cost of Custom Web Data Extraction Tool

First, you will need to pay the total amount upfront before getting your data.

SaaS fees depend on the amount of data

Ongoing monthly SaaS fees depend on the amount of data you extract. Often SaaS publics pricing for websites without any anti-scraping mechanisms and less than 25K pages. You will need to keep in mind the fact that each sub-page costs a credit. It can quickly get expensive if you are extracting data from a number of sub-pages.

As your web-scraping data amount increase, subscriptions payments can be very high. These can top the cost of a custom development and set up of your own software.

You will need a very accurate check their terms. Have they an account query limit that will prevent you from exceeding your plan?

Practice to charge an extra month

There is shady practice to charge the customer an extra month. SaaS usually have put in their terms and conditions that cancellation takes 31 days, meaning you have to pay for the next 2 months in reality. Often the terms and conditions are filled with hidden tricks - if you don't cancel 7 days earlier, they charge you again with not possible to request a cancellation.

You cannot put your account on pause

If you do not need your data for a specific month, you cannot put your account on pause. As soon as you stop paying for your SaaS account all of your structured queries permanently disappear.

One-time job

If you need a one-time scraping job there are many freelancers at upwork.com.

Regular scrape a huge number of products

If you need to scrape a huge number of products. Stand-alone automated web data capture solution has a high initial price on account of the purchase of technology. But as the project proceeds, it is found to lower the operating costs significantly on account of low manpower requirement.

Further, with the majority of web data today existing in electronic forms, the cost of using such automated technology has also reduced.