Cloud Storage and Privacy: How much are you willing to pay to protect your data?

28 03 2013

Introduction

We have all been warned that our Internet purchasing habits and how much we share about our day-to-day lives could be placing ourselves at risk of being victimized.  However, in a recent study published by the European Network and Information Security Agency (ENISA), users, even those who have elevated concerns over privacy, do not heed some of these warnings.  In this study, a majority of consumers were willing to submit personal contact information for a mere 67¢ discount on a  $10.05 online-purchase (XRates) (N., Preibusch and Harasser).   So the question is no longer whether or not a user is willing to share information in exchange for discounts, it is how much information is he likely to share in exchange for discounted services?   This blog explores this question as applied to the adoption of cloud-based storage services.

How much privacy are we willing to sacrifice?

As evidenced by the ENISA study, it turns out that most people have a price, and if this study is any indication, the price is not very high.  In many cases, in the presence “free services”, many of us are willing to supply employment history, email addresses of our closest family and friends, phone numbers, birthdays, and political views – to name a few.   So, it turns out that nothing really is for free.  The price paid for free services is personal information that can be used to support targeted advertising revenue – based on your observed behaviors, spending patterns, your political, social, and financial associations, and more importantly, who you know.  By allowing service providers to observe you, they are able to develop a personal profile that can be sold to their ‘affiliates’ (Google).  To gain insight into what is being sold to affiliates, this author conducted a simple experiment using the third-party plug-in, PrivacyFix,  a tool that estimates the advertising value of Google account profiles.   For the experiment, the Google account was configured with a public profile (Google+) to include links to employment history, more than 100 friends, colleagues, and family, and an association with Carnegie Mellon University.  Even configured to blocks placed on most tracking mechanisms, this Google Plus account allows Google to track 55% of pages visited, and is valued at $25.30 per year in advertising (Anonymous).   This $25.30 subsidizes the free services Google provides, effectively offsetting the pricing of paid services.

Cloud based storage services are based on this same model – Google, DropBox, and Box.net (to name a few) offer free cloud-based storage services with options to increase your capacity.  Basic service starts at 5 GBytes, with increasing levels of storage capacity awarded through new customer referrals (e.g. family, friends, and colleagues) (DropBox).  For capacity needs beyond the default 5GByte level, subscription prices start for as little as $.17 per GByte per year (DollyDrive), and include “free” add-on services to support backup and recovery, revision control, but most importantly:  data sharing and collaboration.  Data sharing and collaboration promotes expansion of the customer base, but also promotes vendor lock by virtue of a shared infrastructure.

Strengths & Weaknesses of Commercial Cloud Storage Options

In spite of these somewhat troubling privacy concerns, new Cloud Storage service providers seem to be popping up each year, and while the cost of paid services still offered at a higher price point than local storage, there are some compelling reasons for migrating to the cloud in some cases.

Table 1 identifies some of the key strengths of weaknesses of todays cloud storage solutions as compared to local storage alone.  For most consumers, the key strengths that differentiate cloud storage from local storage (without software & hardware capital investment) is the infrastructure that supports collaboration and the ability to backup and restore data to an offsite location.

Strengths Weaknesses
Increased productivity – data can be seamlessly accessed across devices and operating systems (DropBox). Data Transfer Latency.  As compared to local data transfers, digital transfer technology can be 6800 times slower.[1]
Ease of setup and use.  Many cloud storage service providers include operating system plug-ins to provide accessible cloud storage as a locally mapped storage device. Confidential Information such as your name, likeness, age, email addresses and names of colleagues and friends, and unencrypted data may be shared with unknown third parties (Google).
Flexible Pricing.  Services range from free, to referral based, to pay as you go, to subscription based services (DropBox). Limited liability policies.  Many service providers require that the customer indemnify the service provider against claims for damage (Google).
Data Revision Recovery.  Many services provide the ability to track changes and recover previously saved versions of files (Dolly Drive). Dependency on external provider.  Service Provider may reserve the right to change the terms of agreements at any time (including the right to suspend or discontinue services) (Google).
Data Sharing & Collaboration.  Shared data can be configured to automatically replicate across subscribed devices and users, facilitating improved productivity for shared data (DropBox). Variable Security.  While security and redundancy can be built into any given platform, each provider balances differing sets of quality attributes, which may expose users at unintended vulnerabilities (Borgmann, T. and Herfert).
Elasticity.  Cloud storage capacity is resizable without the need for capital investment. Service switching Interoperability.   Switching service providers is possible; however, some providers deliver unique services, which are not easily transportable to a new service provider (e.g. Dolly Drive Backup versus Microsoft Azure).
Off-site storage.  In the event of catastrophic loss of local storage and processing hardware, Cloud based storage provides a low-hurdle alternative to backup and safe storage. Pricing for paid services.  In 2013, local hard-disk storage cost less expensive than cloud-based storage[2].

Table 1, Cloud Storage Strengths and Weakness

The big weaknesses are the limited liability and the potential exposure and spillage of confidential information.  Data Transfer latency, while not a show-stopper, is a significant hurdle to more wide-spread adoption, especially in light of the fact that the average data transfer rates in the United States are nearly 6800-times slower than local disk access (Streams) (Seagate).  Some mitigation strategies exist, such as pre-seeding data stores to mitigate latency, however, this remains to be a significant hurdle for some users.  If we assume that the ENISA study represents a predictive model for cloud storage adoption, then liability and confidentiality are not viewed as weaknesses, so the only weakness that really stands in the path of widespread adoption is price. Today, pricing of cloud-base storage for consumer level plans is about 4 times that of  than local storage (assuming that the average user capitalizes the cost of hard disk space every two years), generally starting at $.17 per GByte per year[1].

Moore’s Law and Storage

Now, if we take into account the pricing history of hard drives and capacity over the last thirty years (Figure 1 and Figure 2), we note that there is a close correlation to Moore’s Law.[2]  Note that in the years between 1992 and 2012 two years, the cost per Megabyte and drops by half every two years.  While it is too early to definitively predict, early evidence does suggest that Moore’s law may prove to predict the future of pricing for Cloud-based storage.  Just since 2011, the starting capacity for free services have doubled, and the pricing on paid services has dropped by half[3].

1
2

Deciding How Much to Adopt

While most users are likely continue using only the “free services” until such time that the price point for paid services drops below the cost of purchasing new hardware, the other strengths referenced in Table 1 may drive early adopters to migrate toward cloud-based storage solutions sooner.  For these early adopters, a cost-decision model may help to identify and quantify relevant economic facets.  Such a decision model would quantify up-front costs, annual investments costs, and operational costs to arrive at a total cost of ownership (Bibi, Katsaros and Bozanis):

TCO/Yr = Cu + Cad + Co

Where Cu are the total upfront costs (enrollment fees and setup, acquisition of hardware and software), Cad are annual investment (annual subscription fees and maintenance fees), and Co represents operational costs, such as annual Internet connection costs, utilities, and in some cases the cost of off-site storage and travel.

__________

Anonymous. PrivacyFix Plug-in Results on Google Plus Author. February 2013.

Bibi, S., D. Katsaros and P. Bozanis. “Business Application Acquisition.” IEEE Software (2012): 86-93.

Borgmann, M., et al. “The Security of Cloud Storage Services.” Technical. Fraunhofer Institute for Secure Information Technology, 2012.

Dolly Drive. “Cloud backup for Mac.” Dolly Drive. February 2013 <http://www.dollydrive.com&gt;.

DollyDrive. Pricing & Plans. February 2013. February 2013 <https://get.dollydrive.com/purchase&gt;.

DropBox. “Dropbox – Tour.” Dropbox. February 2013 <https://www.dropbox.com/tour/2&gt;.

—. “Plans – Simplify your life.” DropBox. February 2013 <https://www.dropbox.com/pricing&gt;.

Google. “Google Apps Terms of Service.” Google Apps. Google. Feburary 2013 <http://www.google.com/apps/intl/en/terms/standard_terms.html&gt;.

McCallum, J. Disk Drive Prices. February 2012. February 2013 <http://www.jcmit.com/diskprice.htm&gt;.

N., Jentzsch., S. Preibusch and A. Harasser. Study on monetising privacy, An economic modelf for pricing personal information. Technical. European Netowrk and Information Security Agency. Berlin: ENISA, 2012.

Seagate. “Hard Drive Data Sheet.” December 2012. Seagate.com. February 2013 <http://www.seagate.com/files/staticfiles/docs/pdf/datasheet/disc/barracuda-desktop-hdd-ds-1770-1-1212us.pdf&gt;.

Streams, K. Global Internet Speeds creep back to 2012. August 2012. February 2013 <http://www.theverge.com/2012/8/9/3230626/akamai-global-internet-speed&gt;.

XRates. Historical Lookup Euro Rates Table. 27 February 2012. 18 February 2013 <http://www.x-rates.com/historical/?from=EUR&amount=1.00&date=2012-02-27&gt;.

 


[1] According to the Internet archive waybackmachine.org:  DropBox Pricing 2011-2013.

[2] A profoundly accurate prediction by Intel co-founder Gordon Moore once stated that the number of transistors on a processor would double every two years.

[3] According to the Internet archive waybackmachine.org:  DropBox Pricing 2011-2013.


[1] Assuming a typical uplink data transfer rate of 7 Mb/s (Streams) as compared to SATA hard disk transfer rates is excess of 6 GB/s (Seagate).

[2] Based on 2012 prices of SATA II hard disk price:  $.07/GB as compared to Cloud-based Storage solution priced at  $.17/GB/Yr.

 

Advertisements

Actions

Information

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s




%d bloggers like this: