Linux Mindshare: Getting inside the buyer's head
a guest column by John H. Terpstra
In this guest column at DesktopLinux.com, former SCO evangelist and Samba.org
team leader John H. Terpstra weighs in on what is needed to win the consumer in
the bid for Linux on the backend and desktop. Terpstra argues that emphasis on open
standards and what Linux can offer an organization is a powerful message that is
often mired down in arguments about TCO and other contrasts that often don't add
up in a straight line comparison of Microsoft versus Linux. The essay explores the
issues and offers some food for thought . . .
Linux Mindshare: Getting inside the buyer's head
by John H. Terpstra
January 30, 2003
Comments about "single-source" as opposed to "proprietary" software abound today.
There are different definitions or perhaps interpretations of what constitutes these
approaches. One common thread in these opposing models is consumer expectation --
and how both sides approach the customer is key to shaping the desktop computing
What follows are key issues that must be addressed for Linux to win the buyer's
agenda. Over the past 5 years I have been asking a lot of questions. So far I have
polled well over 1000 VARs, about 300 senior business IT people, and around 50 government
employees from around the world. Here are some key themes from feedback received.
Users/buyers want freedom of CHOICE
This is a prevailing theme from all parties -- it is inherent in business, in life,
and in software choice. Consumers understand that a choice of a single solution
is not a choice at all. In a marketing climate that bombards buyers with messages,
explaining the open source proposition takes a little time and patience. This simple
process can be summed up for the buyer to recognize that:
Companies that have been helped over the proprietary data format hurdle using this
logical approach have adopted a stance that favorable of open standards. What follows,
in reality, is that the buyer has adopted open source.
- A single aquisition source point is risky
- A single solution from that point increases the risk
- Any attribute of that solution that is unique to that single solution expounds
- If that unique attribute is transient (ie. changes frequently) risk abounds
- Security factors that are associated with the unique attribute intensifies
a potential security emergency in the event of alien assault (eg: Worm and/or
- All factors that induce lock-in and remove choice force organizations into
a situation where choice becomes a non-issue, or "non-choice"
- Non-choice is not an option.
Complexity versus simplicity
This shift in paradigm, that is a logical progression for a consumer, translates
directly to productivity issues. Complex software requires additional user training,
which is costly in terms of a training budget, up front costs, and total cost of
ownership. However, often overlooked in the equation and considerably more costly
to a company or organization is lost productivity. Corporate budgets that are training-centric
can be shifted or replaced, but lost time in actual productivity in the company's
own business cannot be recouped.
For example, many companies are just as concerned at the cost of training as well
as the employee's productivity that is lost while being trained. In fact, one manager
stated the while the price paid for a staff member to attend IT training was used
as the reason for rejecting the employee's request for training, the real reason
for refusing the employee was the estimated TOTAL cost of training. This total of
course included an estimate of the value of lost time.
Highly complex products also incur higher infrastructure support cost. As an example,
the cost of gaining MS MCSE certification has surged from about $4000 in 1996 to
over $25,000 in 2002. Those with MS MCSE certifications seek radically higher salaries
for their merit. The best of MS MCSE's land well-paying roles because of the shortage
of skilled and competent MS MCSEs.
All of this just emphasises the fact that all complexity comes at a price. Technologists
are quick to provide supporting agruments for complexity, but upper management often
prefer a simpler and lower cost solution. We should keep this in the forefront of
our thinking when we offer competing solutions to an existing complex one.
Integrity is a big issue
Integrity covers a wide number of factors. Take any single issue -- performance,
reliability, security -- and on its own, there is not enough compelling evidence
to cause a change in buying behavior. Bundle these issues together and the buyer
feels greater need to remedy the larger problem.
One vendor's -- in this case Micosoft's -- answer to performance problems have
traditionally been to install faster, bigger, better hardware. This works in
favor of the hardware vendor and has actually caused much fear for those working
to promote Linux adoption.
Large customers/organizations that have a long-term budget cycles for IT infrastructure
can easily be compromised by unplanned hardware upgrades. The proliferation
of core infrastructure hardware, such as servers, exacerbates the cost of managing
the operational infrastructure base.
This is currently generating a growth explosion in the management software market.
There is significantly more business to be gained in the management software
market as a byproduct of recent server proliferation.
Linux is seen by many as suffering the same blight of increased complexity and
disk-image size as other non-open source solution offerings. Linux can run in
larger spaces and on resource rich machines. But it does not have to do this.
Scalability to fit on more constrained resources is something Linux can do rather
well. Additionally, there is plenty of evidence that Linux-based alternatives
offer very attractive server consolidation opportunities by reducing system
administration and overall task of system management.
- Security record
There is growing and wide-spread concern over malicious threats posed by exposure
to viruses, worms, and trojan horse invasions of information systems. Pro-Microsoft,
or more anti-open source, press have recently hyped open source as being "more
vulnerable" to attacks than Microsoft-based products. There is evidence that
the message is appearing ever more frequently. The message has been based on
Every incident of a security advisory is counted. For each Microsoft product
the threat is counted just once. Linux incidents are often counted for every
communication by each and every player in the field despite the relationship
of a security flaw to a single piece of code. I believe that this is a deliberate
activity because security is a prominent and effective issue that lends itself
to being used as a powerful communications weapon.
- Enforced software registration
Many parties are concerned that Microsoft may use their internal software licensing
database as a tool to prosecute suspected copyright infringers. There have been
estimates that legitimate corporate software users have bought up to 20% more
additional licences than required simply out of fear of prosecution for violation
of Microsoft software licensing.
We should use opportunity to highlight the safety from license infringement
afforded by open source software. Users are more completely within their rights
to total confidentiality in use of open source software. At the same time we
should stress the urgency to comply with all software licensing terms and conditions.
If the cost is too high, come to open source solutions and avoid the tax!
- Uptime reliability
Users respond to system outages by demanding assured service availability. IT
staff respond by installing duplicated servers and clustered solutions. In large
companies the growth in numbers of clustered servers is astounding. A major
increase in network infrastructure management costs is unavoidable as the number
of servers escalates.
Clustering of file and print services is a great way to bury (hide) the inadequacies
and unreliabilities of the systems a business depends on. There is significant
emotional appeal in the mental concept that having a "fall-back" server is a
"good thing". But should we really buy three new cars in case two should break
Whatever happened to efficiency and cost control? Of course, there are situations
that are best addressed by a clustered server environment.
Examples include, clustered compute farms for scientific purposes, load distribution
for web servers, etc. We should recognise the difference between a legitimate
situation and one where the cluster servers purely to render invisible grossly
inadequate systems. In the latter case, the increase in cost and complexity
alone should drive the buyer to a better solution.
In any case, when back-end servers are reliable, users tend to be happier. Happy
users equate with a more peaceful life in the network administration department.
There is also a direct relationship between such peace and the costs of network
maintenance and management.
- Update and Migration Control
IT management remains concerned over their ability to control the introduction
of software features and updates system-wide. This is a point of concern in
respect of all automatic update processes and service offerings.
Samba users that rely on Linux-based systems for file and print capabilities
have been caught off-guard by incompatibilities introduced by MS Windows' automatic
update features. Interestingly wnough, incompatibilities and inconsistencies
have also affected pure MS-only environments as well.
- Too many platforms
Microsoft would like it's customers to propagate "Windows" as a single platform.
This message has only partly succeeded. Many users regard each MS Windows product
as a separate platform. Some even regard each service pack update as a new platform.As
an example, one website recently reported to the Samba mailing list that the
"platforms" they run include: "MS Windows 95, MS Windows 98, MS Windows ME,
MS Windows NT4 SP3 and MS Windows NT4 SP6a, MS Windows 2000 SP2 and MS Windows
Unfortunately, the Linux camp has not succeeded with our message that Linux is Linux
is Linux either. This can be addressed, and is currently the focus of organizations
such as The Free Standards Group and other initiatives that promote unified deployment
interoperability. The single best feature of Linux is one single file format standard
that complies with open industry standards.
There is a significant perception in MS Windows BackOffice 4.5 and MS Windows 2000
BackOffice are not meeting user expectations. Oftentimes, Microsoft users are not
getting the functionality they thought had purchased.
Factors that create the gap between expectation and realization are many and varied.
Most stem from buying based on a slick sales brochure, rather than from research
and careful trial-based evaluation. There is a rabid belief that Microsoft solutions
are so simple that anyone can deploy them. The truth is, they are complex. And,
just like Linux implementations, require expertise.
Buying prior to testing, and without evaluation usually leads to implementation,
since oftentimes software has already been paid for upon acquisition. Deficiencies
are later excused since the budget has already been spent. Open source effectively
creates a different evaluation and purchasing cycle -- one that focuses on the merits
of the software itself.
- Fear of problem situations
It can take 4 hours or more just to get a Microsoft support person on the telephone.
The cost per incident is high (approx. USD$150), but the success rate for issue
calls is not reported to be very high. Many inquiries result in confirmation
of a problem that cannot readily be solved. Although the resolution rate is
far less than ideal, the inference that creates customer concern about open
source is the continued perception that there is no one to call and fix Linux.
The long and the short of it is that Microsoft customers and VARs feel exposed
and helpless in the face of problems with core infrastructure IT solutions.
Linux is seen as posing an even greater risk since there is no one organization
Reality, and a positive track record, have proved that open source Linux has
responded swiftly and completely to security alerts and other issues.
- NO CHANGE
One of the strongest selling points for consumers evaluating a Linux-based solution
is the fact that a change of back-end or desktop solution does NOT need to cause
ANY change in user experience at using the changed environment.
- SOLUTIONS focus
Microsoft sell MS Windows 2000 as a total solutions package. They do not sell
technology. They do sell the sizzle - but not the sausage. Linux propositions
have generally been heavily focused on technology.
In any approach to selling Desktop Linux to the enterprise, it is important that
the community and commercial interests touting open source related solutions focus
on the end goal. That is satisfaction and open standards - and NOT necesarily Linux.
Sure, Linux can cost less, but the end goal affordable results is a far better reason
to select Linux over other OSs. Linux is open source, and that is powerful. However,
customers need to focus on Linux as a platform that is based on open standards.
The real value to the comsumer is the proposition of not being tied to any company
or specific technology that could lock them into a costly or dead end proposition.
Embracing Linux is NOT anti-Microsoft, it is pro-user and pro-results for any company
seeking to unencumber its software infrastructure.
At the end of the day, Customers/Buyers want freedom of choice, seamless interoperability,
managed and controlled budgets, and happy end users -- all things Linux can deliver
better than any other operating system available today.
By focusing on the hot spots of the customers'/users'/buyers' agenda we standa greater
probability of gaining his/her attention and commitment to our cause.
author: John H Terpstra is a respected business computing authority, John Tersptra
co-founded Samba.org and specializes in transitioning businesses to Open back office
alternative technologies. firstname.lastname@example.org.
Copyright © 2003 by John H. Terpstra.