Sun Microsystems hopes itspurchase of the MySQL open-source databasewill be a key step towards positioning it as the pivotal
infrastructure player in a burgeoning online world.
But analysts believe the supplier still has many obstacles to
clear if it is to re-establish itself as the dot in dotcom - as Sun
was known during the first internet boom at the end of the last
millennium. Those obstacles include a number of big players, such
as IBM and Microsoft, who are also jockeying for supremacy in a
field that is currently wide open.
Sun chief executive Jonathan Schwartz laid out the company's
grand plan at its Emerging Markets Summit last August. In his view,
the more that consumers use the internet to access information,
goods and services, the more that they - rather than traditional IT
decision-makers - start to influence technology change.
His rationale is that the more time people spend online, the
more providers spend on technology as they try to meet customers'
changing requirements. So Sun's aim, says Schwartz, is to "drive
the adoption of technology into the hands of consumers and
developers, and to monetise this by selling infrastructure to
global businesses".
Therefore, everything Sun is doing in terms of technology,
branding and the like, is designed to "enable us to bring as many
people online as possible in the most cost-effective way possible
because, at some time in the future, they will open a bank account
and, as soon as they do, that creates an economic opportunity for
us", says Schwartz.
To help this process as well as boost brand awareness and
achieve its aim of becoming a thought influencer, Sun is also
forging relationships with academia, governments - particularly in
developing countries such as China and India - and non-governmental
organisations.
"We are trying to focus in on the next wave of developers, the
next wave of students [future technology purchasers], the next wave
of research and the next wave of economic development, so we can
best position Sun for growth in the next decade, not necessarily
the next quarter or the next few weeks," adds Schwartz.
He likens the company's business model to that of mobile
carriers who give handsets away free to make money on services, or
to Google, which provides search capabilities at no charge but
"monetises advertising". Just as the most popular handsets and
search services are free, Schwartz believes that "the most popular
software, over time, will also be free in terms of pricetag and the
licence used to convey the intellectual property".
Therefore, Sun's own promotional device to seed the
infrastructure market is open-source software. Schwartz explains,
"The move to free software, I would argue, is the single largest
transformation that the internet is delivering today and the
delivery of free services leveraging free software is defining how
consumers experience the internet."
If he is right, logic dictates that Sun needs a full suite of
open-source middleware, including development tools, an application
server and, of course, a database, that can be "leveraged" by
in-house developers and a third-party channel to create the
necessary applications to deliver free services.
All being well, these packages will act as a Trojan horse for
Sun to sell its infrastructure-based systems and related services
to end-users - an important consideration for a company that still
generates most of its revenue from hardware.
The fact that margins for even high-end RISC-based systems are
starting to dwindle as the market moves towards lower-cost
commodity machines suggests it is imperative for Sun to transition
its business model and create new engines for growth - and one of
the levers for this transition is software.
Jon Tutcher, head of developer marketing for Europe, the Middle
East and Africa, describes this as "hugely important" strategically
for Sun, not least as a "way of getting to talk to customers".
And this is where the
MySQL acquisition comes in. It is viewed as a crucial piece in
the overall jigsaw because, with 50,000 to 60,000 downloads a day,
it "provides us with a very ready access point to a new customer
set", says Tutcher.
And that, says Andy Butler, a vice-president at analyst Gartner,
is something the supplier needs if it is to start growing its
revenue again. Butler points out that three or four years ago,
there was real doubt over whether Sun would even survive, but
Schwartz has since turned the company around and brought it back to
profitability.
"The installed base is now much more willing to continue doing
business with Sun and at least consider it when moving to new and
next-generation products," says Butler. "But up until now, it has
not done much to appeal to new customers, who were probably more
likely to go to Hewlett-Packard or IBM."
Schwartz says he is "now confident that the ship will keep
floating" and can start looking at other strategies. As a result,
he says, Sun is "now moving into an era where it really wants to
start taking on new business" beyond its traditional financial
services and telco base.
To this end, it is taking a leaf out of IBM's book by exploring
alternative delivery models, such as subscription-based hosted
infrastructure services and software as a service.
Butler adds, "It is trying to get away from the 'we flog you a
box, you come back in five years and we sell you another one'
model, which is the old Sun." But to successfully move to an
annuity-based model, it also realises that hardware will not be
enough.
"At the very least, it needs strong services to sell with its
hardware and to basically lock people into it by using its software
tools and utilities in a utility model by which they rent," says
Butler.
Tutcher confirms Sun's interest in this area, indicating that
the aim is to build on the supplier's existing network.com
offerings. The grid-based system currently gives organisations
access to third-party high-performance computing (HPC)
applications, while ISVs and developers can also exploit its
resources to build, test and deploy their own at a cost of a 50p
per processor per hour.
However, both the pricing model and the fact that the system is
based on HPC applications means it is not "hosting in the true
sense of the word", he says, which means both approaches will have
to be tweaked.
For Sun to become "a bigger, more important player in the
Web 2.0 space", says Tutcher, it will need to create "a whole
portfolio of offerings" for developers that want to create and/or
host their own applications or access hosted infrastructure, if
they prefer.
"Part of this thing [Sun's reinvention] will be driven by
hosting - the part where Amazon is playing today with EC2 [Elastic
Compute Cloud]," says Tutcher. "We are partially there with our
network.com offerings, although not properly, but we do expect to
offer more."
But the downside for Sun is that it is far from alone in its
ambitions. IBM, Microsoft, Salesforce.com, Google and Amazon are
all keen to dominate a green-field area that is up for grabs and
has no clear leader - a fact that will prove both an advantage and
a disadvantage to Sun.
Another benefit of the MySQL acquisition, says Tutcher, is the
fact that open-source software gives the company "access to a much
bigger developer pool".
Interestingly for a company that has traditionally delighted in
antagonising Microsoft, it has not been too proud to co-opt part of
its business model. As Neil Ward-Dutton, a partner at analyst firm
Macehiter Ward-Dutton explains, Sun is trying to out-Microsoft
Microsoft by providing third-party developers with a platform for
building volume applications that is not just cheap, but free.
"I don't think it's about taking on Microsoft head to head or
becoming an alternative to Microsoft," says Ward-Dutton. "What Sun
is actually doing is copying Microsoft's client/server strategy
from the 1990s when it created an amazing developer community
around Windows NT by giving it great tools that were very cheap to
create a volume market - which is how it built its business in the
enterprise space."
In the "new platform" area of web applications, it is not
necessary for
Sun to take on Microsoft because this, too, is a "green-field"
area and the software giant is far from being in control so
far.
Again, Sun's dominance is far from assured in this sphere, not
least because all the big boys from IBM, Microsoft, Salesforce.com
and Amazon "are racing to build new, compelling platforms for
developers to capture the new generation of people who, 20 years
ago, were developing for NT", says Ward-Dutton.
This means that, once again, the field is "wide open" with no
obvious leader, which means "there is still all to play for", he
adds.
One of the key challenges Sun faces that its rivals do not,
however, is born of the very vehicle by which it hopes to make its
fortune. The open-source community is notoriously suspicious of
commercial companies trying to subvert its aims and will be keeping
a close watch for any sign that the supplier is either optimising
MySQL for its own software stack to the detriment of other
offerings, or going for a hard-sell approach with its other
products.
Sun will therefore have a fine line to tread, not least in how
it handles existing MySQL customers and rivals, such as open-source
distributors Red Hat and Novell, so as not to end up alienating
them and encouraging them to adopt alternatives - which would, of
course, undermine the premise of its acquisition.
As a result, says Butler, it is crucial that the database
division that now houses MySQL maintains its promised autonomy and
that there is "rigid control and guidance from the top to ensure
that everyone representing Sun, whether they are internal staff or
the channel, execute strategy consistently so as not to cause local
friction and real resentment". This is imperative, says Butler,
because "if it goes wrong, it will be because of local
execution".
Butler points out another challenge, however - that although
Sun's products are good, "they are not overwhelmingly superior to
those of other suppliers" and, apart from MySQL, are not, as a
rule, market leaders.
Clive Longbottom, a service director at Quocirca, questions why
open-source advocates would go to Sun to buy their software when
the LAMP (Linux, Apache, MySQL and Perl) stack "has more cachet
around it" than Sun's equivalent.
But Tutcher believes this is beside the point because the aim is
not so much to provide a rival SAMP stack - with the S standing for
its
OpenSolaris Unix operating system - as to offer customers
choice.
"As well as OpenSolaris, we also offer Windows and Linux and
there has also been a change at the processor level because we now
provide Sparc, Intel and AMD," says Tutcher. "So we offer choice at
the OS level, choice at the processor level and choice at the
middleware level. This means that when we go into a customer
account, we can make recommendations to help them solve their
problems."
Although it is still early days for Sun's strategy and too soon
to say whether it will succeed, in Butler's view it is probably the
right one for where the supplier is today - despite the inherent
risks.
He concludes, "No one is doubting Sun's ability to survive, but
the question now is, will it just keep churning its existing
customer base, which many other companies do successfully, or will
it succeed in achieving its aim of re-establishing itself as a real
innovator?"