Fog Creek Software
Discussion Board




Rent/Subscription Based Software?

Any thoughts on whether this is going to fly in the business world?  Seems that most businesses aren't willing to house their data outside their own data centers.  Then again, there's always salesforce.com.  Will this be a successful way to go or is it another fad?

GiorgioG
Monday, April 21, 2003

A *lot* of companies got burned by the ASP model a few years ago. The company I was at outsourced their Exchange Server to a big ASP that oversold their capacity. Considering how painfully easy Exchange is to admin, that was purely wasted money.

I think subscription based software is equally a poor move. If software had always been leased, it would be different; but moving *to* a subscription model from a sales model won't work - if you price high enough to weather the drop in revenue, then the competition will eat your lunch by selling cheaper than you rent. If you price low enough to compete with that, you had better be well prepared to face the significant drop in revenues.

And with respect to the only company that I think has the cash reserves to pull it off - Microsoft started two years too late. Activation and/or leasing should've been implemented with Win2k, a must-have OS. Starting Activation with XP just makes Win2k that much more popular, and any leased OS will die stillborn.

I hope. :)

Philo

Philo
Monday, April 21, 2003

I would tend to agree with you except at some point (somewhere in the next few years), there will be no reason to upgrade from WinXP (hell, when I get my new laptop next month, Win2K is replacing XP.)

What this means to me is that all software solutions (minus custom s/w) are reaching the point of diminishing returns for upgrading.  Who here doesn't know 1 client that's still using Office 97?

How does Microsoft get you to upgrade to Office 2006?  It doesn't have to, you're paying $30/yr for a 'subscription to Office'

So I tend to think this is going to happen to most boxed software packages.

It's still up in the air in my mind, though I could certainly see something like this working.

GiorgioG
Monday, April 21, 2003

Simple: Just explain to customers that the current model where the only way to remain in business is to constantly come up with new versions with features that make less and less sense... doesn't, er... make sense :-) ... while having customers rent the software has the advantage of keeping the company in business (someone still has to answer the phone, fix bugs, maintain the code when the next version of Windows or technology comes up) _and_ keep some developers around so they add _useful_ features per customer requests. It's a win-win situation.

After talking with some customers, we came up with the idea of having them amortize the software each year (ie. "MySoftware 2003", "MySoftware 2004"). Financially, the price is the same as what they paid for maintenance, but they seem to like the idea of inputing this as investment better.

But then, this only works for software that requires occasional changes, and is important enough to the company that they can't afford the editor just going out of business in the foreseeable future. In other words, it doesn't work for eg. utilities like zippers, etc.

Frederic Faure
Monday, April 21, 2003

Our (admittedly limited) research has shown a higher tendency to want to use ASP-model software, not lower, because of smaller or non-existant IT budgets. Zero deployment and maintenance is still a pretty significant win.

Brad Wilson (dotnetguy.techieswithcats.com)
Monday, April 21, 2003

Brad -
Because when your network or internet connection is down it's better if people can do *no* work whatsoever?

Philo

Philo
Monday, April 21, 2003

Philo,

I'm not sure what you're saying here. You're saying that we need to be concerned about when our network goes down, or when their network goes down?

If the former, then that's what redundant links are for.

If the latter, I'd have to say that the network being down is a rather crippling thing for just about any business today. As much communication is probably done by e-mail as is done by the phones, so it's probably roughly equivalent to having no phones. You don't think they'd be fixing it right away?

Brad Wilson (dotnetguy.techieswithcats.com)
Monday, April 21, 2003

"Considering that Exchange is painfully easy to admin":

Anyone else wondered whether Philo has ever actually been admin for one or more Exchange server, and if so what his or her secret was?

We switched away from Exchange because of the amount of time it took simply to keep the machine properly patched.  Painful, alright.

Not to mention that of the various clients and vendors that we know that still use Exchange - all have dedicated staff capable of dealing with Exchange or outsource the function - and all have been unlucky enough to experience data corruption at one time or another, resulting in at least a day's down time and requiring several high priced consultants to come in and stay up all night to repair it.

Maybe their techniques *all* suck. And likely their disaster planning is completely inadequate (most companies don't have any).  But characterizing administration of Exchange as "painfully easy" is contrary to most corporate experiences in this area anyway.

Bah Humbug
Monday, April 21, 2003

+1 for never dealing with Exchange Server again. It's a monster pain in the ass. We experienced what you and others have: stability problems, patching issues, data corruption. If the server didn't shut down cleanly (like, say, a power outage and we aren't at work to shut the machine down right), the Exchange store was almost surely corrupted and had to be restored from backup. No way would I wish Exchange admin on my worst enemy. :-p

Brad Wilson (dotnetguy.techieswithcats.com)
Monday, April 21, 2003

Exchange is painful to admininster. That part Philo got right.

With regard to stability it is worth remembering that it uses the same Jet database engine that Access does.

Renting software is what every company would like with two conditions; one that the rent is exorbitant and two that all data or documents you create with the rented software is inexorably tied in with the program you are renting so you never can get out of it.

Unfortunately most customers don't have a suicide wish.

Stephen Jones
Monday, April 21, 2003

Brad -
I'm saying that when our internet connection goes down then sure I can't send email or surf the web, but I can write code, the tech writers can work on documents, our webmasters can craft pages, the graphic artists can work on artwork, managers can write memos or reports, and, in fact, you can even *draft* email, you just can't send it.

If all our software was subscription, based on some proposed models (including some of MS's favored proposals) then either you cannot access the applications you need, or if they can't phone home, they won't run.

So then I guess you could hold meetings...

Now if the model evolves such that software is installed locally and will run up to 48 hours without calling home, then that might be viable.

Philo

Philo
Monday, April 21, 2003

Obviously enterprise software falls out of the bounds of what Philo is talking about above. When your network goes down, it's fairly reasonable to expect that the connection to your large, multi-user DB will fail too. Boy, do customers love having DBs on site during the rare cases when the network connection goes out (they can continue to use their apps). Boy do their field workers hate it when the power in the office goes out, because everyone goes dead.

On balance, many enterprises opt to host their apps offsite at in a large colo facility. That's a given. Once you go that route, it isn't much of a leap to think about outsourcing....

J
Tuesday, April 22, 2003

J - which of the activities I mentioned requires access to a central database? ;-)

However, agreed that enterprise client/server apps can suffer the same problem...

Philo

Philo
Tuesday, April 22, 2003

There are other advantages to being able to use an ASP-model. Some components may be extremely expensive, but licensed by the CPU. For example, say you made everybody buy a copy of "The Super Special Widget" to run your application. It costs $5000/CPU/year. In an ASP model, you can spread the cost of the software across hundreds of users, but you make it very expensive to bring the app "in-house".

Brad Wilson (dotnetguy.techieswithcats.com)
Tuesday, April 22, 2003

I think microsoft are being very successful at moving people over to a subscription based model.

Three are plenty of people using the MSDN subscription to purchase Visual Studio.net.  They seem to think that it is some kind of loophole that MS don't know about.

When I point out that they only get it for a year, they smile slyly and say 'I can still keep it after my subscription has run out.'

I think Microsoft are going to get quite a few hooked using this method.

Ged Byrne
Tuesday, April 22, 2003

I haven't checked the license recently, but as of 2000, there was nothing in the MSDN license preventing you from continuing to use the enclosed software after your subscription lapsed.

IMHO, the model of success for MS was to make it easy for developers to do their thing, and charge for corporations to do *theirs*. MSDN supports this model, Ballmer apparently doesn't get it.

Philo

Philo
Tuesday, April 22, 2003

Aha! Microsoft's Licensing 6.0.  They were successful in moving to a subscription model only because they are monopolists and the only other option that customers had was to pay full retail price for every copy when they want to upgrade - or to leave Microsoft's products altogether, which only a few were able to do even though they wanted to.

T. Norman
Tuesday, April 22, 2003

What is the alternative to rent/subscription based software? I do not want to know about schemes that reduce the software to be a loss-leader to sell some other stuff (advertisments, support, books ...). I want to know about options for selling software.
While we were in a growth market with barely mature products, sure, you just sell to virgin customers and the version you are selling is just not completely there yet, but already usefull.
But as we move towards a more mature market, where most customers already have the basics taken care of new features take a lot of customer education, what are the options?
Classic "buy the box" models in a mature market require that you develop the realy new killer app/must-have feature every single time. There is no gradual improvement path here. You hit the jackpot or you are outcompeted by your own release x-1, which the customer already has. Can this be done? Even trying will mean spending boatloads on R&D investing deeply in 50 lumps of coal to find the one nugget of gold. Furthermore, every time you succeed you have just made the next round that much harder. This is not even a red-queen race, it's the red queen on an accelerating treadmill!
Subscriptions are a way out. The danger everyone will immediately point out is that without the "must improve on x-1 or die" incentive companies will just go weak and stop innovating. I do not think this is nescessarily so. There is still competition between the different companies, just as in the box model. But in subscription there is more a gradual improvement path possible. You fall behind, you will gradually start to loose subscriptions to the competition.
Now can we move our industry towards such a model? My guess: No. The reason: It requires too much faith and cooperation, and a kind of smarts humans do not possess.
You see, as mentioned by many before, there are alternative models when you take in the broader picture. Software does not have to bring in the money. It can just be turned into a loss leader. Hey man, it's free, have a sniff.
This is where the whole TCO type debates come in. It might be free but if you look in the long run and add in all the factors you will see it is more expensive ... dooooo ... mistake against human nature.
If you want to do TCO for humans, take into consideration that accounting for even the least bit uncertain future costs just does not register that way. Take at least a 50% reduction for every budget year you move a cost into the future.
So while you may argue that y1: 500$ + y2: 500$ + y3:500$ =1.500$ is less than y1:0$ + y2:1.000$ + y3:1.000$ = 2.000$, what you should actually compare against is y1:0$ + y2:1.000$/2 + y3:1.000$/2/2 = 750$.
What is even worse: People that got burned in the past do not seem to learn this. So showing after the fact that actually 500$ was lost will not get you an Aha! from the customer. He'd rather take another tun at the craps table.

Great for the guys and girls running the stuff software is the loss-leader for, bad for the software people. You are no longer the tree that bears the fruits that bring life, you have become a cost that needs to be minimised.

Just me (Sir to you)
Wednesday, April 23, 2003

How short human memory is.

Does no one remember the anti-trust decisions on IBM regarding leasing hardware and software and charging on 'overuse'?

If you rent the software and/or hardware with which to run the Command, Control and Communication of your enterprise then you are at the mercy of the supplier who can change the conditions of use at their whim.

Given the very advantageous terms of write off for IT equipment and software (especially in the UK), there is absolutely no reason to convert this to an Expense overhead.

Simon Lucy
Wednesday, April 23, 2003

I just have to wonder how anyone who says "unless we lease our software we'll die" thinks the electronics industry manages to keep going? Apparently Sony didn't wither and die after "the last person" bought a Sony Walkman.

Seems to me that one answer is diversify. Look at Winzip or CuteFTP. They haven't just rested on one product; each company is working on new solutions. nSoftware hasn't just sat on IP*Works - they're now offering IM, Zip, and EDI tools.

Then there's competition to be had - make products that are better and cheaper.

I don't think there's a shortage of opportunities in the COTS software world. Just a shortage of initiative and work ethic in some companies.

Philo

Philo
Wednesday, April 23, 2003

Philo,
not that it is all doom and gloom in software land, but your examples are a bit unfair.
Walkmans wear through usage, so the limited lifetime is an implicit feature of the product. Software does not "wear out" (except maybe due to security exploits, which could become the major reason for upgrades, but as I stipulated in earlier treads I do not think this can save a big software shop since it only has to do with one dimension of the product that could be covered by specialized competitors).
While yes, you "own" the walkman you just bought, the thruth is the thing is engineered to last about just until the warrantee runs out. If bits would degrade through usage or over time we would not have these problems (nor the upportunities, but that  is a different story).
As for WinZip and the like, don't get me wrong but models that work for one man and his dog do not nescessary scale to the 100+ COTS developer shop.

Unrelated: In your CAMEL Saga if all the client machines can handle the .NET framework why even consider ASP.NET?

Just me (Sir to you)
Wednesday, April 23, 2003

Well that's an excellent question, Just Me.
How about "Why are they so gunshy about deploying the .Net framework, yet they've rolled out two IE updates in the past four months?"
Or "if there's already another winform application approved that uses the .Net framework, why is this an issue?"
Or "how come we have to explain that winforms are more robust then web forms to a management team that just had a webform project fail because the IE UI couldn't handle the user needs?"

All good questions. I have no answers. :-)

Philo

Philo
Wednesday, April 23, 2003

I doubt if Walkman's are built to run out just after the warrantry.

Look at the mobile phone market. Nokia has almost become a monopoly in Saudi, which is one of the most discriminating markets, simply because its phones are so much more reliable than the competitions.

About eighteen months ago one of my classes decided to liven things up by putting one  in a cupboard in the classroom and then phoning it continously. As soon as I got near they stopped calling it and it stopped ringing. This went on for about twenty minutes until I clicked, opened the cupboard door with a great flourish and threw the phone on the floor from a great height whereupon it bounced, nearly broke some innocent lad's ankle, and smashed against the wall. I then picked it up again, dropped it on the floor , and jumped up and down on it half-a dozen times before putting it in my pocket and handing it in to the office. The secretary was just locking it away when it rang!

Stephen Jones
Wednesday, April 23, 2003

kjh

dd
Friday, May 09, 2003

*  Recent Topics

*  Fog Creek Home