The Artima Developer Community
Sponsored Link

Weblogs Forum
At the network's edge, is software a service business?

4 replies on 1 page. Most recent reply: Apr 29, 2004 9:49 PM by Brendan Macmillan

Welcome Guest
  Sign In

Go back to the topic listing  Back to Topic List Click to reply to this topic  Reply to this Topic Click to search messages in this forum  Search Forum Click for a threaded view of the topic  Threaded View   
Previous Topic   Next Topic
Flat View: This topic has 4 replies on 1 page
Frank Sommers

Posts: 2642
Nickname: fsommers
Registered: Jan, 2002

At the network's edge, is software a service business? (View in Weblogs)
Posted: Apr 27, 2004 1:26 PM
Reply to this message Reply
Summary
With powerful personal computers and local area networks easily affordable even to the smallest businesses, many software companies are making big bets on providing small business software. But any soul braving that market would do well to carefully think through their business model: they might find something completely unexpected.
Advertisement
As the large enterprise market for software matures, software firms must look elsewhere to sustain their growth. One of those areas is small-business software: Software for accountants, lawyers, doctors' offices, auto dealers, contractors, etc. That software includes both vertical-market software components, as well as the infrastructure part of supporting vertical market software (databases, operating systems, servers, etc).

All the major software firms have announced plans to significantly invest in that market segment. Microsoft, for instance, plans to spend on the order of $10 billion in the next few years on marketing and product development aimed at the small business market.

Indeed, most businesses are small businesses. A large part of the US economy consists of businesses employing a 100 people or less. These businesses provide many billions of dollars of output, and collectively employ far more people than the top 1000 largest firms in the US. It follows that making these business' lives easier with software products could prove a significant business opportunity.

But wait a minute before you rush out, quit your job, and start a company building small business software. Serving that market segment may turn out to be the ultimate trail of trials for a firm that embarks on that journey.

The reasons for that premonition are simple: First, small business software is mission-critical software. When you sit in a dentist's chair, and your dentist wants to pull up your records on his computer system before he pulls your tooth, that dentist simply can't afford for that system to fail at that moment. Or, if you're an auto dealer and a customer is ready to purchase a vehicle, you simply can't afford for that computer to fail at the crucial moment when you need to calculate the customer's loan payments.

At the same time, small business software must also be inexpensive, since small firms don't have the IT budgets of larger enterprises. (Many small firms don't have IT budgets at all.) If you charge above a few thousand dollars per customer, at least in the US, you simply won't reach a large part of your small business market.

Not only must small business software be inexpensive to buy, it must also be inexpensive to maintain. Since small firms don't have an IT department, they don't understand the continued investment required to maintain a typical enterprise software system. While maintanence of a typical enterprise software, such as an ERP system, costs far more than the initial purchase price, a small business is not prepared to accept a similar reality.

So what appears to be at first an attractive business opportunity, might just be the ultimate contortionist's excercises for software firms: Sell software suitable for the space shuttle in terms of reliablity, but at a millionth of the cost.

As if that were not a challenge enough, consider that the software environment on the space shuttle is controlled by the system provider: Astronauts are not, at least to my knowledge, allowed to install OS upgrades, download all the spam that fits in email, or browse pornographic Web sites with a dozen spywares and pop-up ads.

A small business' computing environment couldn't be more different: All users typically run with full administrative privileges (since most people just start using new PCs with the default settings), and are allowed unrestricted access to the computer and the Internet. That results in a dangerous elixir - the personal computer equivalent of Anthrax. Since there is no system administrator ever to manage those PCs, the computer environment at small firms tends to decay over time, until it reaches crisis proportions.

And when crisis hits, the software firm supplying the mission critical business software will likely take those blows. As the dentist is ready to start a procedure and needs to view an X-ray image, or the auto dealer is ready to close a sale and needs access to an inventory system, if the system fails at those crucial moments, they will then reach for the phone and utter the dreadful words that sends shrivels down the spines of anyone listening at the receiving end of the line: The system is not working.

At that point, both the customer and the software vendor are in a crisis mode, since the customer is stopped from performing a mission-critical action. As with a person who never bothered to go in for a checkup for decades, and then suffers from a massive heart attack, doctor and patient are in for a life-and-death struggle to save the patient. The medical analogy stops there, though: In the medical field, insurance companies (or the government) pick up the tab. In the software world, both the customer and the vendor end up as loosers.

I've have been building software for small businesses for the past 2 years, and, to be honest, have not yet been able to find a solution to that quandary. I hear and read the optimistic predictions about small business software, but then see the daily reality of managing crises and putting out fires that burn at the edge of the network, in the offices of small firms.

And here is the rub: Many small business owners are not able to distinguish between hardware and software, let alone types of software. They only know "the system." The system includes the total environment to perform the mission-critical business functions: The hardware, the OS, the vertical market app, the network, and everything in between. If any of these components fail, then the "system is down."

Actually, most of us software types think in a similar fashion when it comes to non-software: It's our car that doesn't work, not the ignition or the transmission. We just can't get to work because the system is down, and we don't really care at that point about the root of that problem. We just want the solution. So we take the car in the shop.

But there is no equivalent of "the shop" when it comes to a computer system. The local computer guy might help fix this and that, but they seldom perform the maintenance work required to prevent crises from breaking out. In the world of small business software, I believe that there is a need for the equivalent of a service station.

Lacking the notion of preventive maintenance, and lacking the sophistication in popular commercial operating system to peform that maintenance automatically, software vendors catering to small enterprises end up becoming service-based businesses, not product-based. In other words, they will end up, by default, delivering the value by performing services (crisis management, etc), and not by selling products. They end up as emergency room operators without the help of medical insurance or government assistance.

The problem with that is that a business must invest in the areas where it sees the most return. If we don't see return from developing and selling great software, we won't be justified to invest into that area of our businesses. Instead, we will probably have to invest in the services and crisis management areas.

Services might just be where the software business, in general, is going. To finish the auto business analogy, car manufacturers are realizing that they, too, become increasingly dependent on services to create revenue. Instead of investing in new technologies for making cars better and safer, they must equivally, if not more, invest in building service offerings to support the "systems" their customers care about.

Becoming a service businesses raises interesting questions for technology firms: How does R&D and intellectual property help that kind of a business? Should we look to successful restaurants (as an example of a service business) for business models instead of the Microsofts and Suns? What happens to the typically high profit margins of technology firms when they convert from being product-based companies to being service-based ones? If the auto industry is any guide, those margins will surely decline.

Admittedly, I wanted this blog to be as much an invitation for comments as a rant about the way I see things. You're input is most appreciated.


Brendan Macmillan

Posts: 8
Nickname: yow
Registered: Feb, 2003

Re: At the network's edge, is software a service business? Posted: Apr 27, 2004 5:36 PM
Reply to this message Reply
Sounds like a classic maturing market - I guess you know "Crossing the Chasm" (G Moore) well?

I think it's crucial to realize that the solution you provide does not need to be perfect, but just better than what they have now. That's progress by anyone's measure.

Perhaps a good way to approach it is to see what problems they experience with the present system - if you can fix those, customers will be much more understanding about any other problems (such as reliability).

Perhaps an ugly solution for reliability would work well enough: print out all the records as a backup? Or insisting on a backup PC or two - since hardware is now really cheap, esp last-year's hardware, which is more than powerful enough for this kind of app.

Or you could insist on making the application as simple as possible, so there is less to go wrong.

In the big picture, computers are different from every other kind of new technology, that had rapid adoption followed by maturity (eg TV, radio, telephone, railroads, cars etc). The difference is that computers work with mind-stuff - with structured knowledge, skills and expertise - so that as soon as we understand something new, or improve our understanding, we can automate it. If you believe that knowledge and skill can be improved forever and in unexpected ways (as I do), then you'll believe that computers will go on forever.

To support this, it does seem that computers have already gone through a few booms: mainframes, mini's, workstations and PC's. It's been argued that the PC is different, is the end of the line, because it is a mass market commodity - and so there is no fundamentally new market to expand into. But I believe that new uses will appear for computers, and each new use will create a new market - in the same way the other inventions I mentioned have. It's just that with computers, development and distribution and so on is so much faster.

It's true that this is a different kind of rhythm from what went before (business automation getting cheaper and more accessible), but I think there are many other tasks left to be automated. In fact, pretty much all of them!

But the big picture doesn't help you much right now. In the small, I'd say your customers need an appliance, not a general purpose computer. Therefore, restrict the functions to only what is actually needed for their use, and do that well. That is, an Enterprise Portal, just for their business - and all they have is a browser. You don't have to use their existing computer, because hardware is so cheap - you throw it into the deal.

This eliminates many of the problems you mentioned. And I think that your customers might really like a "specific purpose appliance".

Anyway, I hope you find some of these thoughts interesting/helpful. :-)

Frank Sommers

Posts: 2642
Nickname: fsommers
Registered: Jan, 2002

Re: At the network's edge, is software a service business? Posted: Apr 28, 2004 12:37 AM
Reply to this message Reply
> In
> the small, I'd say your customers need an appliance, not a
> general purpose computer. Therefore, restrict the
> functions to only what is actually needed for their use,
> and do that well. That is, an Enterprise Portal, just for
> their business - and all they have is a browser. You
> don't have to use their existing computer, because
> hardware is so cheap - you throw it into the deal.
>
> This eliminates many of the problems you mentioned. And I
> think that your customers might really like a "specific
> purpose appliance".
>
If you ask many small business owners, they will tell you that they explicitly don't want to use a portal, precisely because of their concern for data availability. Again, the key seems to be that the kinds of software I'm describing requires very high rates of availability at the crucial moment the user needs to use it. It's different, for instance, from an online banking application: If you can't check your account balance or transfer money in this very minute due to a temporary network outage, you can always go back 10-15 minutes later. That's just not acceptable for the small business owner who has a customer standing right there, ready to make a purchase. I guess the kind of software I'm describing is more like a cash register, which must work at the exact moment the customer wants to pay.

The cash register is also a good example of a specialized kind of computer used for only one purpose. But extending that concept might require a different PC for every major application.

Jeroen Wenting

Posts: 88
Nickname: jwenting
Registered: Mar, 2004

Re: At the network's edge, is software a service business? Posted: Apr 29, 2004 3:09 AM
Reply to this message Reply
> go back 10-15 minutes later. That's just not acceptable
> for the small business owner who has a customer standing
> right there, ready to make a purchase. I guess the kind of
> software I'm describing is more like a cash register,
> which must work at the exact moment the customer wants to
> pay.
>
I'm working at a company that's been targeting small and medium sized businesses in 2 very specific sectors for the last decade (and as a department of another company for over a decade before that).
The solution we've chosen is to have everything centralised on a server WE control (more or less).
Every customer on contracting us is provided with an AIX (or for the very small ones these days Linux) server with which we can make contact at any time through a DSL line (or callback/dialup in the past) through a router (previously modem) we also supply.
As part of the deal we also provide for the installation of the DSL line if needed, providing permanent internet connections at all workstations if the customer wants that.
All applications run on the server and are accessed from the clients.
Depending on the application that can be a terminal window, classic client/server application or for our latest offerings a browserbased solution.

We can also provide all the workstations if the customer wants that (many go for that option) or else we can reconfigure them (needed anyway to install our terminal emulator or c/s software of course).
Through the server, software updates can be pushed to the workstations giving us a single point of entrance and control.

For those customers without their own sysadmins (or if they want it) we can also provide either on-site or remote system administration services on contract basis, a service that's becoming ever more popular with customers (especially as more small customers are signing up).

> The cash register is also a good example of a specialized
> kind of computer used for only one purpose. But extending
> that concept might require a different PC for every major
> application.

For highly specialised tasks that doesn't matter.
There's people that only NEED a single application to be running (think again of the cash register, the cash register doesn't need to provide a web browser, a word processor and a graphics package. At most you may want some sort of spreadsheet integrated to view the sales proceeds of the day which is really part of the cash register functionality).

The mentioned car salesman NEEDS a single application only.
Whether that application will then call other applications internally to yield data should be transparent to the salesman, he wants a single interface providing all the functionality he requires.
He doesn't want to have to start a catalogue program to look for the options, a word processor to type the contract, a credit checking application to determine the credit rating of the customer, and a link to the vehicle register to notify the government of the new registration.
He wants to push some buttons in a single application that combines all those functions into one seemless whole.

Brendan Macmillan

Posts: 8
Nickname: yow
Registered: Feb, 2003

Re: At the network's edge, is software a service business? Posted: Apr 29, 2004 9:49 PM
Reply to this message Reply
Thanks for the reply Frank.

I wasn't at all clear about a "portal": I forgot to mention that I meant having the server on the same PC. Otherwise, you would be quite right about the data availability (unless you had a great network connection). I assumed that a cash-register-like app wouldn't need a network connection for anything. If the network was needed for some functions, make it so that the core functions aren't affected by network failures. Hmmm... network access would help if you could repair problems remotely - much faster than waiting for someone to come out (which small businesses are used to - you would impress hell out of them).

I suggested using the browser as a portal, because it hides verything away from the user very simply and cleanly, and there is a lot of useful technology available for development (J2EE, frameworks etc etc), and easy to reuse previously developed functionality and so on.

I don't know your specific market at all - but I do think one app per computer makes a lot of sense based on my understanding of your description, and the need for reliabilty and ease of use. Although multi-function appliances seem popular at the moment (eg. copier+scanner+fax+printer+phone) they can be a pain to use, and if one part fails (eg PC, keyboard, monitor) - all the apps fail. Separate PC's give you redundancy, and that's a giant leap towards space-shuttle like reliability. :-)

With different appliances, you can also customize the outside (keyboard) to reflect that function, and so enhance ease of use - for example, those touch-screen cash registers you see everywhere these days (you know, some of them seem to just use flash)

Anyway, I think {browser+server; and 1 PC/app} gives great reliability and ease of use benefits.

It's counterintuitive to restrict a general purpose computer to one app, and certainly does feel technically wasteful to me - but benefits are in the eye of the buyer. ;-) I think it's a good whole solution to the problem you outline.


[I see Jeroen agrees with the 1 PC/app idea, but found a way to get network reliabilty - sounds a bit expensive to me, but I guess a dedicated phone line or DSL isn't that much. But I still submit that for many apps, a remote connection is not absolutely essential].


> > In
> > the small, I'd say your customers need an appliance, not
> a
> > general purpose computer. Therefore, restrict the
> > functions to only what is actually needed for their
> use,
> > and do that well. That is, an Enterprise Portal, just
> for
> > their business - and all they have is a browser. You
> > don't have to use their existing computer, because
> > hardware is so cheap - you throw it into the deal.
> >
> > This eliminates many of the problems you mentioned. And
> I
> > think that your customers might really like a "specific
> > purpose appliance".
> >
> If you ask many small business owners, they will tell you
> that they explicitly don't want to use a portal, precisely
> because of their concern for data availability. Again, the
> key seems to be that the kinds of software I'm describing
> requires very high rates of availability at the crucial
> moment the user needs to use it. It's different, for
> instance, from an online banking application: If you can't
> check your account balance or transfer money in this very
> minute due to a temporary network outage, you can always
> go back 10-15 minutes later. That's just not acceptable
> for the small business owner who has a customer standing
> right there, ready to make a purchase. I guess the kind of
> software I'm describing is more like a cash register,
> which must work at the exact moment the customer wants to
> pay.
>
> The cash register is also a good example of a specialized
> kind of computer used for only one purpose. But extending
> that concept might require a different PC for every major
> application.

Flat View: This topic has 4 replies on 1 page
Topic: Did I tell you I love Manning? Previous Topic   Next Topic Topic: Announcing the Artima Interest Groups


Sponsored Links



Google
  Web Artima.com   

Copyright © 1996-2014 Artima, Inc. All Rights Reserved. - Privacy Policy - Terms of Use - Advertise with Us