AnalyticBridge
Data Intelligence, Business Analytics
Analytics 3.0. What is it? What are the risks and opportunities?
What is it?
Software accessible online, via your web browser, instead of available as a package installed on your desktop or on your company's network. Also, software accessible online that can run automatically in the background via API calls. This is SaaS for analytics.
For instance, one could install a copy of SAS on her web server, and everybody with an Internet connection could access it in batch mode (API) or via a user interface if a GUI is created. SAS code could be executed from your browser, you would not need to have a copy of SAS on your machine. This scenario is actually more likely to happen with open source software, and this is already true for R. Unfortunately, these open source implementations (unlike SAS) can not easily process large data sets. AnalyticBridge is currently developing an open source solution to score large data sets: it will be written entirely in Perl, and you will be able to run the software in batch mode (API calls) or online, from your browser.
What are the risks?
Software piracy. Everybody could install a web server / web interface to allow the whole world to do statistical analysis on her server, via a browser. So if you set up a SAS web server, pays a $10,000 yearly license fee to SAS, charge $1,000 yearly fee for people to use your service, and have 20 clients, then:
What are the rewards?
This technology offers access worldwide to everybody, and in particular to people who can not afford to purchase, install and maintain expensive packages that handle large data sets. Even for a company like SAS, benefits are considerable: they could have a new business model, where individuals running their own SAS server would be a SAS affiliates and pay a fee to SAS. In the above example, maybe SAS would charge you $500 per online user, in top of your license fee. In turn, you would have to charge money to offer your service: you would have to charge more than $500 per user. It is a win-win model, where SAS, you and end-users all benefit. SAS would benefit by reaching out to people who would never purchase a full (desktop or Intranet) version of SAS, because of the high cost and other drawbacks (maintenance).
Software accessible online, via your web browser, instead of available as a package installed on your desktop or on your company's network. Also, software accessible online that can run automatically in the background via API calls. This is SaaS for analytics.
For instance, one could install a copy of SAS on her web server, and everybody with an Internet connection could access it in batch mode (API) or via a user interface if a GUI is created. SAS code could be executed from your browser, you would not need to have a copy of SAS on your machine. This scenario is actually more likely to happen with open source software, and this is already true for R. Unfortunately, these open source implementations (unlike SAS) can not easily process large data sets. AnalyticBridge is currently developing an open source solution to score large data sets: it will be written entirely in Perl, and you will be able to run the software in batch mode (API calls) or online, from your browser.
What are the risks?
Software piracy. Everybody could install a web server / web interface to allow the whole world to do statistical analysis on her server, via a browser. So if you set up a SAS web server, pays a $10,000 yearly license fee to SAS, charge $1,000 yearly fee for people to use your service, and have 20 clients, then:
- Your net gain is $10,000 per year (20 x $1,000 - $10,000)
- SAS is missing revenue due to your operations
- SAS will sue you, but if your server is located in China (say), this might be difficult
- The issue is that even if 20 people are using your service, they all use YOUR copy of SAS and thus, from a license viewpoint, it is as if just one user (you!) is using your SAS copy
What are the rewards?
This technology offers access worldwide to everybody, and in particular to people who can not afford to purchase, install and maintain expensive packages that handle large data sets. Even for a company like SAS, benefits are considerable: they could have a new business model, where individuals running their own SAS server would be a SAS affiliates and pay a fee to SAS. In the above example, maybe SAS would charge you $500 per online user, in top of your license fee. In turn, you would have to charge money to offer your service: you would have to charge more than $500 per user. It is a win-win model, where SAS, you and end-users all benefit. SAS would benefit by reaching out to people who would never purchase a full (desktop or Intranet) version of SAS, because of the high cost and other drawbacks (maintenance).
Views: 126
Comment
-
Comment by Vincent Granville on January 2, 2010 at 9:58am -
But isn't this already the case if you perform a traditional SAS analysis for a client? If your SAS server is accessible only to paid clients (e.g. it requires a login / password and approved membership), your online service is very similar (from a legal point of view) to an off-line service offered by a consulting company. And you can purchase an insurance to protect yourself against a range of potential legal issues, such as delivering erroneous results that will harm your client.
-
Comment by jyotiska bhattacharjee on December 29, 2009 at 9:46pm -
suppose one have a copy of sas and giving poeple to use it online via a web browser in exchange of a fees, then the owner must understand that all work they do on sas , all the analysis will be on the owner's copy and in case some one using a restricted data of national security or stolen data of a company then it will be a problem for the owner he might even end up in jail...if one is giving permission to the users the owner should also ask them for source authentication...but this will slow down the process....also wont slow internet connection will have any impact on the computing speed??
Top Content
Follow us
Big Data Jobs
© 2013 AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC
You need to be a member of AnalyticBridge to add comments!
Join AnalyticBridge