Interview with
Niel Robertson
Bradley: Tell us a little about your background.
Robertson: I have been building software companies for about ten
years now. For some reason I have always gravitated towards enterprise
application management – building software to help companies
manage the applications that run their business. In the 90s, I co-founded
my first company, Service Metrics. We were focused on helping enterprises
get more out of their web sites by better managing their web site
performance. At that time, every large company (and some small ones)
was scrambling to build a sales channel through the web. Hordes
of end users and potential customers were jumping onto the Internet
for the first time, using 14.4Kbps or 28.8Kbps modems. Surfing the
web on a 28.8 modem was a painful experience, and when you threw
a user into a buying process (for example, buying a book on Amazon.com),
you realized how very sensitive a user was to the speed of the web
site. Service Metrics measured the speed at which a user could perform
the most important transactions on a website from the end-user’s
perspective. With the data we collected, Service Metrics could identify
and dissect the largest bottlenecks to performance and help our
customers spend their money intelligently to speed things up. It
was exactly what the market needed and after only 18 months, we
sold the company to Exodus, which at the time was the world’s
largest hosting company.
Bradley: What did that experience teach you about market
timing?
Robertson: I’ll admit I got lucky with the timing of Service
Metrics. In hindsight, though I learned a lot about market timing
– and ensuring the market is ready for a good idea. I would
guess that as many good ideas have failed because they were too
early to market as too late. I have four guidelines I jokingly call
my “four rules of market timing”. I try to apply them
to any business idea I am considering.
My first rule is called “New dog, old tricks”. This
happens when companies invest in new technology platforms to essentially
do what they are already doing. This might be customer support,
sales channels, or monitoring their business. Inevitably, they forget
how much work and technology it took to get the old stuff working
right and always skimp on this piece of the puzzle out of the gates.
Essentially, Service Metrics helped people bring measurement of
the customer shopping experience back to companies now selling on
the web.
The second rule would be “Throwing good money after bad”.
This happens over and over in information technology shops. So many
times, a company commits to a certain technology choice and then
scrimps on the initial implementation or poorly scopes the project.
In either case they have become wed to it. At this point in time
their only choice is to throw more money (and for me – new
technology) to help get what they originally wanted.
The third rule is the “Earthquake” rule. This states
that the pain of a problem rarely comes at once – it usually
comes in waves. As with real earthquakes, you either get the big
one first and then some small aftershocks or you get a few tremors
first and then the big one. When the small tremors come first you
find companies building their own solutions or cobbling together
whatever they have lying around to solve the problem. If the earthquake
comes first you’re too late to the market. When we started
Service Metrics we found a number of large companies actually trying
to build a network of monitoring agents just to get some data. This
was a great sign as it was not a sustainable strategy given the
complexity of the problem.
The last rule I call the “Hokey Pokey”. You’ve
heard the song, it goes like this: “You put your left leg
in, you put your left leg out” etc. etc. The point being companies
inevitably decide to bring things in house or outsource them over
and over again. The theory is that this is going to save a lot of
money and solve all their problems. The problem is companies essentially
exchange the challenge of managing the technology for the challenge
of managing the service providers - or vice versa. When everyone
started outsourcing web site hosting, Service Metrics grew dramatically
by giving web site owners data about how well their hosting providers
were performing.
So these are the rules. I always reserve the right to break them,
but they have worked well for me.
Bradley: Have you ever ignored your rules?
Robertson: I was involved once (on the board) with an outsourced
network operations center provider – trying to get people
to hand over their management of network operations. The problem
was that we were trying to force the Hokey Pokey rule instead of
capitalizing on a natural shift in the market. Clearly, some people
have been successful at causing this shift, but in that situation
none of the other rules applied. Companies were not going through
a technology shift in terms of managing their network operations.
They had made their big investments in HP OpenView or CA and, generally
speaking, they were working fine. And there was no earthquake looming.
So, they were essentially 0 for 4. The company, despite the best
efforts of its management just could not find a market.
Bradley: How does your latest venture fit your four rules?
Robertson: When I was starting a new company, that eventually became
Newmerix (http://www.newmerix.com), I wanted to consciously build
a business case that fit these rules. After some tiny missteps,
I stumbled across the packaged application market. Packaged applications
are big software suites like PeopleSoft (now Oracle), Siebel, SAP,
Etc. Companies buy these suites and attempt to run their whole business
on them instead writing their own software. Two things really opened
my eyes when I looked at the packaged application market: the immense
amount of money that went into getting these packaged applications
up and running, and the rudimentary technologies available for most
companies to manage them .
So, I started to consider my four rules. “New dog, old tricks”
was clearly applicable. Essentially, when people bought these applications
they were replacing custom applications they had written themselves.
The promise of moving to a packaged application was no more development,
only configuration. But in the end, every business is so unique
that companies found they had replaced pure custom development with
a new form of software development focused on customizing these
packaged applications. However, all the advanced technology and
process and methodology developed over the course of 10 years for
custom development just wasn’t available in the packaged application
framework. There are many, many publicly available horror stories
of hundreds of millions of dollars wasted trying to get these applications
to essentially do what the company already had working with their
custom developed software.
Clearly “Good money after bad” was present as well
since companies had taken a major bet on these applications (many
were forced by Y2K to replace old Cobol programs) and now they were
held hostage by this new technology set.
As we started to look at what management software people were using
for these applications, we saw rule number three emerge –
tremors before the earthquake. Companies were cobbling together
whatever technology they had lying around in their custom development
environments and trying to get it to work with the packaged application
architecture--not a sustainable strategy. As a side note, the earthquake
part of that rule really hit with Sarbanes-Oxley last year when
the government basically said to IT departments “prove you
have your act together in IT”.
The last rule, the Hokey Pokey was actually the most important
for us. When companies first implemented these applications they
were usually installed, at great expense, by a specialized systems
integrator. When the bubble burst and $400/hr consultants were no
longer a viable IT staffing strategy, almost every major company
brought ownership of these applications “in-house” with
internal staff to manage them. This brought the problems of managing
these applications right to the company’s doorsteps, and provided
us a direct avenue to sell to the enterprise. So we set out to build
Newmerix, the first change management suite for packaged applications,
about two and a half years ago.
Bradley: Have the four rules worked?
Well fortunately we started a little over two years ago, building
the core suite of products based on what we saw happening in the
market. We anticipated a number of earthquakes, Sarbanes-Oxley was
one of them. Additionally, the consolidation of PeopleSoft/JD Edwards
and then PeopleSoft/Oracle have been additional market earthquakes
that highlight the pain of packaged application management. In the
longer term, we are encouraged because we are seeing the next wave
of the four rules. All of the major packaged application vendors
are shifting to an “Application Operating System” strategy.
In this model they want to provide all the layers to deliver all
of your applications. This shift is the next “new dog”
companies are grappling with – and we’re currently the
only company with a suite of products designed for the packaged
application architecture. We also see an emerging market for outsourcing
the development and maintenance of packaged applications so we expect
to take advantage of the need for good management tools if these
mission critical applications to go outside the enterprise. All
in all I would say, yes, the four rules have been an incredibly
successful guide for our business and continue to be so.
Ben Bradley is the managing director of The Bradley Group and GrowingCo,
Inc. – marketing and market research firms serving technology,
manufacturing and security clients. Do you have a question or topic
you would like Ben to address in an upcoming column? Please send
your comments to ben@benbradley.net.
|
