Why do some solutions cost $10,000 and others $1 million?

After years of incrementalism and “band-aiding,” many large growing e-Tailers are starting from scratch and redesigning their entire e-commerce website from the ground up.  Like any typical selection process, they document their vision, their requirements, their decision criteria, etc.  They then go on to create a short list and begin the vendor parade.The PowerPoints decks are impressive.  (Many are too large to fit in your e-mail inbox.)   Like most mature software categories, e-commerce vendors tend to copy each other and end up sounding very similar.  But unlike other software categories, e-commerce software solutions vary greatly and selecting the wrong vendor directly impacts the top and bottom line of a company.  You can pay as little as $100/month plus 1.5% of online sales for Yahoo! Stores and up to several hundred thousand dollars per processor for the best-in-breed.  The key questions: What are you really paying for?…and is that large premium in cost worth it? 

I can’t say it enough: It all depends on the size of your business.  You are indeed paying for quality and uptime in these high-end solutions, but you are also paying for “growth” tools.  Today’s high-end e-commerce solutions include very sophisticated tools for behavioral targeting, AB testing, real-time merchandising, searchandising, managing dynamic content, analytics, collaborative filtering and many others.  In fact, Omniture just paid $65 million for Offermatica, a vendor specializing in AB Testing, illustrating the value placed on just one conversion feature.  These new tools can easily improve your conversion rates by 50% to 100%.  A 10% increase is almost guaranteed.  The question is whether this increase in online sales pays for the investment. If you sell over $20 million online, the answer is most likely yes, invest in a best-in-class e-commerce Suite.  If you are much smaller, stick to a basic catalog and shopping cart.

One caution….beware of “smogging”.  I happened to come across the term “smogging” while reading. In sales, it means “blowing smoke” — a hype-laden pitch that makes promises the salesperson can’t keep. You’ll get your fair share of vendors claiming they can “do it all at half the cost.” Caveat emptor…do not take short cuts in your selection process. Get your hands on the software and even do a proof of concept. It’s only then that you’ll appreciate what you’re truly investing in. It’s an outdated cliché, but in software it’s almost always the case: you get what you pay for.

Zoli Erdos had a post earlier this week summarizing research that shows that SaaS is not just for SMEs. The post references reports from McKinsey (via Nick Carr) and Nucleus Research that each demonstrate increasing SaaS adoption in large enterprises.

We couldn’t agree more!

Technorati Tags: SaaS, software as a service, nick carr, Nucleus Research, zoli erdos, McKinsey

Bob Warfield from Callidus Software has a great article on the Sand Hill Group’s site, bringing some rationality to the SaaS discussion. As we’ve been saying for quite some time, there is no one “right” model (on-demand or on-premise) for software deployments. I mentioned in my last post that we’re seeing great adoption of ATG’s OnDemand offerings from some very large companies. Likewise, some smaller companies prefer a traditional licensed application model installed on-site, despite the “conventional wisdom” of SaaS/on-demand being THE model for SMBs.

Now more than ever, customers should have the flexibility to chose the model that is right for them. This should be self-evident and alone should dictate a hybrid model. But even for companies that are not motivated solely by their customers’ best interests, Bob makes a really good point:

“For established software vendors, the transition to a pure on-demand model can be daunting. The financial, engineering and customer service issues aren’t anything to be taken lightly.”

The hybrid model is best for customers and best for vendors - I can’t see why any software company would choose one over the other.

Can you?

Technorati Tags: Saas, software as a service, on-demand, Callidus Software, Bob Warfield, Sand Hill Group

eWeek today is reporting on a new Forrester study “The State of Enterprise Software Adoption.” The line that caught my eye is “SaaS spending will continue to increase among enterprises. Although medium and small companies (defined as those with 100 to 499 employees) lead as the current users of SaaS, 45 percent of Global 2000 and 32 percent of very large enterprises remain “somewhat interested” in adopting SaaS in 2007.”

This is very consistent with what we are seeing. Our SaaS business is flourishing among larger companies. They choose to outsource the operations for a variety of reasons, but the bottom line is they choose to do other things with their own resources, and allow us to run their systems.  Some very large companies, who obviously have the wherewithal to run their own commerce or customer care systems if they so choose, choose instead to buy the service from us: Symantec, Coca-Cola, Airbus, T-Mobile, AARP, New York & Company, Road Runner Sports, Orange (of France Telecom) and Cingular, to name a few.

I expect this trend to accelerate.

Technorati Tags: SaaS, Software as a Service, Forrester

When talking to retailers, reporters, and even industry analysts in the eCommerce space, we sometimes run into an issue where the person we’re speaking with is absolutely convinced that every company should—and does—want a hosted model of software. Far be it from us to knock SaaS—we offer our eCommerce solutions in both license/on-premise and on-demand deployments. But still, there are skeptics who just don’t “get it” that a one-size-fits-all approach is faulty.

Last month, Gartner released findings suggesting that in certain implementations, the Software as a Service model is not as cost-effective — or even effective — as earlier projections and experiences have led some companies to believe. Sure, it can be cheaper and offer an impressive speed to launch. For some retailers, this makes perfect sense and helps them get the best out of technology without having to put their hands in it. Yet, they say for complex, larger implementations this just isn’t your best bet.

It’s interesting, but there doesn’t even seem to be a consensus on the definition of SaaS. Forrester has a very rigid definition, calling out SaaS as “built from the ground up to be multi-tenant at all layers of the stack: database, server, and application.” Yet the more common view found in wikipedia is in line with our view that SaaS is “a model of software delivery where the software company provides maintenance, daily technical operation and support for the software provided to their client.” The key here is that there must be flexibility—the strictly multi-tenant approach does not work 100% of the time.

Serious merchants do not want to suffer the work or risk of downtime on someone else’s schedule — and the risk of upgrade problems is real. How does multi-tenant work in a situation where a single store needs more than one server—which is most serious merchants.  It doesn’t—even the multi-tenant players must use dedicated tenancy in this case.

We are multi-tenant in the parts of the stack where it makes sense, such as in data storage, and dedicated where that makes sense. The oldest rule is at play here—“treat customers how you want to be treated.” That means we should be treating each implementation uniquely, and offering what makes sense. Otherwise, you’re simply doing a disservice on the hope that you make a sale.