April 28, 2009

Develop faster and cheaper, from client/server to RIA to mobile to SaaS

by Gordon Graham, Editor, SoftwareCEO

If you're like most ISVs, retooling an existing app for smart phones or SaaS gives you shudders. All that recoding, and all that cost?!

It's tough to see how you'll ever get a payback on a new delivery model, especially these days.

Many ISVs desperately need a cost-effective way to develop for multiple platforms. And it would be ideal to pay-as-you-go, so your deployment costs grow with your customer base... instead of shelling out a big upfront expense.

But how can you do all this when you're developing in .net or Java, using a traditional IDE like Visual Studio?

Let's start by acknowledging the four main delivery models for software today:

• Client/server
• Rich Internet Applications (RIA) running over the web without a browser
• Mobile, chiefly for smart phones
• Software-as-a-Service (SaaS).

But most software is still being written for a single model in a specific language; rewriting an app for another model is a long, costly, and risky proposition.

When we chatted recently with Regev Yativ, CEO of Magic Software Americas, he said he has the answer for hard-pressed developers. Magic is a company with a 20-year history and 2,500 ISV customers, many having been with it for years.

Three questions to ask about R&D
Yativ speaks with software execs every day, and he's developed some fundamental questions for ISVs designed to probe the cost, effort, and risk of deploying apps in a new model.

Question #1: How much would it cost to move from your current model?
"What would be the real business damage to you?" asks Yativ.

In other words, how much would your R&D budget have to swell to support a new delivery model like SaaS? What would that mean to your business? Could you even find that much funding today?

Question #2: How much effort would it take to move a .net or Java app to SaaS?
"How much work would you need to do to scale up your business to provide SaaS? What technology do you need to move from one delivery model to the other, with as little pain as possible?"

Wouldn't that be "a humungous piece of work?"

Could you even do it with your current development platform? How much of your code base would have to change: 50 percent? 80 percent?

Question #3: How much risk would it take to rewrite your code?
"What would be your risk in taking your current environment and plunging into a big project of trying to rewrite .net code to create a RIA application?"

"Are you willing to take this huge risk for your business to move to a different delivery mode?

"The flip side is, how much money could you make if you could actually reuse your existing applications without touching them — without destroying them — but just modernizing them?"

These are tough questions, but Yativ says Magic has the answers in its cost-effective uniPaaS development platform.

The uniPaaS platform
Magic's uniPaaS is a "business application platform" that supports software development and deployment in four modes: client/server, RIA, mobile, and SaaS.

Applications developed in uniPaaS are cross-platform, independent of OS, database, and hardware. The system includes all the expected developer's tools, such as version control and support for remote teams.

Yativ calls it "the whole enchilada."

But the real benefits seem to be twofold:

• Faster application development
• Faster redeployment of apps in another delivery model.

"We are saying to ISVs, 'Guys, stop wasting money on experiments?! Take whatever you want to develop, and develop it quickly," says Yativ.

"Productivity-wise, we can compete with any other environment, and we are willing to put a scenario publically with .net or Java and prove it. We are doing that daily with proofs of concept."

Just how fast is uniPaaS?

Yativ says any new development scoped out at 100 person-years in .net can be done with uniPaaS in 30. Yes, 30. That's a savings of two-thirds of your R&D budget, and your time to market.
After a learning curve of a few weeks, Yativ says the productivity gains start to show up... and they never end.

"The productivity will be much faster than anything you know, so eventually this translates into how many people you need to accomplish the same task."

The other big benefit is faster redeployment of an existing app for a different delivery model.

Yativ says that any application developed in uniPaaS is ripe for redeployment. He estimates only 10 to 30 percent of the code will need to be touched to take an app from client/server to RIA, or to mobile, or to SaaS. That's another big savings of time, effort, and headaches.

"We are offering you — with less expense — a way to create a code base that will be reusable time and again."

The sweet spot for this platform
What sort of ISVs should consider uniPaaS?

Yativ says the ideal candidate has revenues of $10+ million a year — lots have $20+ million — and more than 50 customers, who has hit a wall in its .net or Java development.

It's also useful for any ISV with legacy code that wants to modernize and roll out apps in a more contemporary model.

"We have seen many software vendors having stagnant sales. We've seen vendors struggling with multi-platform, or with different environments where they don't feel they are experts... and this is where we shine. We are here to bridge the gap."

Magic can help ISVs expand their offerings for RIA, mobile, and SaaS delivery, and deal with each of these radically different business models in a cost-effective way.

"ISVs who have code in uniPaaS, even as client/server, can actually take this code, and with small modifications deploy it as RIA, and with slight modifications deploy it in a mobile environment. And then, another slight modification is required to deploy as multi-tenancy mode in SaaS... which means you can take the same code base and reuse it time and again," says Yativ.

"We have customers that started with us 20 years ago, and they are still selling the same application, only 20 years ago it was looking like DOS, and now it looks like web services.

"Our customers had web services before anybody heard about SOA. We already had SOA 10 years ago."

"That's what we call the power of choice." And that choice goes on and on, and applies to technologies that aren't even available yet.

Pricing is flexible
One of the interesting things about Magic's uniPaaS platform is the flexible pricing. Yativ says concurrent licenses cost $2,000 to $4,000 per seat.

"If you want to start developing a new application from scratch, you could probably do that with $10,00 in your pocket."

Then there's a deployment fee; but again, this is flexible.

"What we normally do with ISVs is sit down and analyze what would be the most productive business model for them."

This could be a perpetual license, or it could be performance- or transaction- or parameter-based pricing.

"Example: A company is issuing a specific number of airway bills for loading cargo on airplanes, or processing X million credit-card transactions a day. So we sit with the company and analyze what would be a benchmark worth pursuing, and then we agree on a long-term agreement," says Yativ.

"Once we agree, the initial investment is really negligible. If you are a new ISV and you want to join Magic today, the threshold is almost not there. If it is there, we will work it out."

The big objection? Magic's "small" size
With all these benefits, what's the objection to the uniPaaS platform?

Yativ's answer will resonate with most smaller ISVs.

"It's mainly our size," he says. "We are not Microsoft, we are not Oracle. We provide exactly the same level of technology, for double the productivity, at half the price — but we don't have marketing money to put it on billboards at the airports."

Unlike Oracle, where he worked earlier, he says Magic's pricing doesn't run sky-high to support a Porsche for every sales guy.

"What we want to emphasize is that we are doing business with lots of ISVs, and they are driving the Porsche in the end. If you look at my car, it's not a Porsche," he chuckles.

"I'm meeting ISVs daily, and they really don't want to discuss ROI, because that entails having to make an investment. So I coined an expression ROEI for 'return on existing investments.'

"You already invested so much in your technology. You put a tremendous amount of money and time and effort and attention into it, so let us show you how to get a return on your existing investment, how you can make things run faster and more efficiently, how you can save money on maintaining your code and on enhancing your offering.

"ISVs are more open to discuss this right now, in recession times, than any other time."

Magic is doing fine with that approach. Sales were up 6 percent in 2008 to $62 million.... precisely because its pitch and its tools have great appeal to software execs.

If this sounds good to you, you can find out more about uniPaaS from the product page which has white papers, a webcast, and a podcast to check out.

About the author: Gordon Graham is an award-winning journalist with 30 years in the software industry. And as That White Paper Guy, he helps B2B software firms tell their stories with persuasive, fact-driven white papers and crisp case studies.