Insight Community

Open Source is another mode of competition that allows one or more players to commoditize the complements to their offering. It is not a distinct threat, in that to offer any real threat it needs to be funded/supported by other vendors. The best way to deal with it is to attack the profit sanctuaries of the competitors that are attacking you.

SaaS is another mode of delivery that is actually more of a threat to traditional IT groups than it is to enterprise software vendors. This does to represent a threat to traditional vendors as long as they embrace it as a delivery vehicle.

The traditional enterprise software business is being attacked from two directions, open source technologies and hosted software-as-a-service (SaaS) offerings. Do these represent a real threat to the incumbent software providers, and if so, what is the best way to deal with it?

I've been studying both of these situations for some time now and believe both open-source technologies and hosted source-as-a-service are huge threats to enterprise software; they will completely remake the landscape.

Let's look at SAAS first. It is clear to me that the enterprise model of huge upfront licensing costs that are a capital expenditure rather than ongoing service fees that are counted as an expense no longer makes sense, and it is only a matter of time before all organization finally say 'no' to purchasing software instead of services. So the question is no if enterprise software vendors should switch but instead when and how.

That said Salesforce.com is the clear and obvious key player in the SAAS market. They make themselves known and their self-promotion shows no signs of modesty. However, they have weakness and one is they treat their customers with disdain as evidenced by their offerings and their contracts. One key example of how Saleforce.com does not accommodate users is they don't offer different levels of users access to the system for different pricing levels so a company has to pay the same for a secretary as they do for a salesperson. Task-based pricing would resolve this, and make it easier for companies to justify using the system

Salesforce.com also pursues contracts that lock companies in, and they don't let customers downsize as needed. Again, offering a flexible terminate-at-will policy and allowing customers to downside their usage based on seasonal needs or even after a downsizing can back off of pricing.

Sales managers will be aghast at these suggestions, but the reality is that this allows a company to focus its expenditures on improving the offerings it provides its customers instead of paying expensive sales people. In addition, once a company starts using a system there are not likely to move off of it unless it doesn't meet their needs. Better to focus on meeting the needs instead of locking in the customers.

In addition, Saleforce.com has backed itself into a corner by declaring a "No Software" positioning. However, that does not work for many companies. One of Saleforce's Achilles heel is in that positioning. My recommendation is to rearchitect offerings to be as modular as possible and using HTTP-based web services to connect the various components. Doing this will provide customers a choice of deployment location for each of the components allowing those customers that need deployment at their location the needed flexibility to do so. Salesforce.com does NOT offer this option and I doubt they ever will do so given their positioning.

Also, the architecture to use is the REST architecture that uses true HTTP web services as opposed to SOAP's tunneling-over-HTTP approach. Contract to popular assumption, REST-based architectures are enterprise ready and are infinitely more resilient to change than SOAP, making for an ability to be much more agile and react more quickly to market demand and competitive pressure.

Further, Salesforce.com has shown significant traction with AppExchange, but if you dig deeper you find that most vendors on AppExchange are not making much money and are disgruntled. Salesforce's policies are very one sided and Saleforce views their partners as both lead generators and revenue sources and are not viewed like a partner should be. Again, this is another Achilles heel that can be exploited by a competitor. Partner policies should be about ensuring the partners make money sooner than later, and make a lot of it. There are many ways to address this, and partner programs are probably one of my strongest expertise. But to cover that here would be out of scope and will cost a lot more than $100.

By the way, the modular REST architecture makes it much easier for 3rd parties to add value to, or even replace components in the core architecture.

As for open-source, encroachment of open-source into the territory of enterprise software is also inevitable. The open-source model should be embraced for its brings many benefits that a business person can appreciate, and yet still retain the commercial licensing aspect required of an enterprise software company. And a well thought-out modular REST-based architecture can make this work even better.

Basically the benefits to open-source are:

1.) Reduced sales cycles and sales costs; most customers come ready to buy.

2.) Reduced costs to facilitate evaluations; customers just download and try.

3.) Much greater ability to form a community that contributes functionality.

4.) Reduced development costs as the community contributes.

5.) Allows customer to gradually see value in paying as opposed to the all-or-nothing model.

Geoffrey Moore explained is best; separate the core from the context and use open-source for the context. Well, for the company trying to sell enterprise software, they sell the core modules and open-source the base level functionality. And over time, they move some of the older core into the base.

Actually, I'd advocate a dual license model like MySQL uses where the customer needing commercial support and service-level guarantees pays for the license and the one who don't do not. But the good news is some of the non-payers will become payers and if they choose a competitive solution, chances of them ever becoming payers is slim-to-none.

The biggest problem is transition, and the way to achieve that IMO is as Clayton Christensen recommends: to create a skunks-workz outside of the current corporate environment and allow them to create the new partially SAAS and partially open-source solution w/o them having to deal with insiders that want to see the initiative fail for personal selfish reasons.

If you like this analysis and would like me to do in-depth work, I'd be open but only on a fully paid consulting rate.

Speaking from the perspective of an information integration specialist the drive for open source software in the enterprise will not come from a desire of the penny pinchers to save money but rather a desire of independently minded people to get something done with little or no money or support.

When it comes to the big budget projects and software evaluations the open source offerings will struggle against the smooth sales pitches of the software sales people. The core products tend to be very complex and proven over many years with hundreds of reference sites, the software comes with extensive manuals and training courses and consultants and scalability. With someone who has a big checkbook open source is out of consideration, they are looking to spend up on a low risk approach.

These large projects tend to be ponderous and lurching things. A lot of individuals in the organisation get frustrated at speed and scope of change. They see themselves in the mirror each morning growing old and building lines of sadness at spending each day using a five year old operating system, throwing shadow systems together in Excel and Access, waiting for browser based reports or a user-friendly data integration tool. They see cool things in IT magazines that seem to be far away on the horizon. In the end they either leave or make something themselves in a shady little corner of the network and use open source or MS Access because they have no money. They find a spare server that a big project forgot about or they hide a couple desktops under a spare desk where they whir away all day. These folks deliver exactly what they want because they have no political interference and their colleagues love it.

Thus open source BI, ETL, data quality and databases spring up and flourish in little corners of the enterprise. Traditional enterprise software vendors don't need to even notice these pools of open source goodness. They simply come in with an enormous budget and an enterprise standard and millions of lines of complex code and solid metadata integration and sharing between products and partner products. They trounce the open source pools with superior power or fall into implementation madness and spin wheels in the mud waiting for a better enterprise implementation to come along.

Software-as-a-service has similar flexibility as open source and similar barriers. It can be inserted into the enterprise through a back door thanks to its flexibility and low start up costs. It can be finely directed at a specific problem and become popular because of it. It suffers from the slow movement of the gears of a large enterprise, many of the core applications are not ready to interact with SaaS, they are not ready for SOA or XML or information as a service. Gatekeepers with a traditional software mindset fret at the security implications or the network performance.

SaaS is a stronger competitor as it can be backed by a premium vendor supplying millions of lines of code and a huge client list. Enterprise likes having a vendor bring them complex software with a large price tag on it. It lets them blame someone from outside the company if the project goes horribly wrong.

The best way for traditional vendors to answer the attack is to join in. Traditional information integration vendors need to have a SaaS offering and they need to interact with open source operating systems, databases and reporting tools. They have a million lines of complex code and a bunch of gurus sitting around thinking up companies to acquire and things to put in the next release. Let these guys loose on SaaS and the results could be impressive.

Judging by the strategic moves that traditional enterprise software vendors have made recently, it’s straightforward to conclude that hosted software-as-a-service (SaaS) and open source models are increasingly seen as threats to the traditional enterprise software business.

Consider SaaS.

Vendors, such as Oracle, SAP, and Microsoft, who provide traditional enterprise software, have already felt the SaaS-impact and have since made the move to offering hosted SaaS solutions themselves – Microsoft with its Windows Live and Office Live platforms and Oracle with its on-demand Siebel CRM offering. The entry by the #1 and #2 software behemoths into this space lends all the more credibility to the SaaS wave.

Hosted SaaS is definitely not a new concept. It has been around since the early days of the Internet. It was salesforce.com that was hugely responsible for making SaaS an industry buzzword with its hosted CRM offering. Perhaps the greatest advantage that SaaS provides is the ability for an enterprise to try a piece of enterprise software for free (read 30-day fully functional trials). Nothing could be more tempting for an enterprise as this approach involves no capital expenditure and is completely risk-free (well almost, security concerns apart). It’s a real-time proof-of-concept without having to touch the existing systems.

The single biggest argument in favor of the traditional enterprise software business was the ability to customize the solution to fit the enterprise’s needs. This too has been overcome with features such as salesforce.com’s AppExchange platform that allows customers to exchange customizations as and when required.

In a recent interview with Knowledge@Wharton, Adobe’s Chief Software Architect, Tom Malloy, said:

“We're big believers in the transition to hosted services and we're investing a lot of money in it. We don't have any $100 million hosted service businesses yet, but I think we probably will in the not-too-distant future – certainly in the three- to five-year time frame.”

SaaS is definitely here to stay.

Open source technologies are more of a mixed bag – but still a potent threat. For example, I don’t see enterprises chucking out Microsoft Office and taking the plunge to platforms such as OpenOffice. Microsoft’s Office is eons ahead in features, usability, and support. But enterprises could give mySQL or Postgre SQL a try. Windows is far ahead in reach and adoption for Linux to displace it. Microsoft made a killer move with the inter-operability agreement with Novell.

This is a classic case for the technology adoption curve. You would definitely see innovators and early adopters hugging these models right away (for example, Nokia embracing salesforce.com). The early and late majorities would adopt a wait-and-watch principle as they are not very keen on disturbing their existing functional systems (if it works, don’t mess with it).

For traditional enterprise software vendors, the best way to deal with these changes in the technology landscape is to embrace them. If you happen to be the best traditional enterprise software vendor, it’s time to be the best SaaS vendor too. Build on your reputation and offer your customers a wider choice. There’s no reason for not wanting to do so. At the moment, both SAP and Oracle offer customers on-premises and on-demand solutions. It’s up to the customer to decide what he wants. That, I believe, is a wise approach.

Continues Adobe’s Malloy,

“I think that's a phase shift. It's like the shift from paper to electronic, which is another area where we participated in that disruption. Paper hasn't gone away; I don't think shrink-wrapped software is going away, either, but a lot of the software business will move towards that service model.

So I think that we've been planning for it. Hopefully we will be ready for it and either we transition ourselves and our customers over to that model – those who want to move to that model – or, when the marketplace moves, we will move along with it.”

The short answer is 'yes'. These two threats are not necessarily coming from two different directions, but sometimes as one in the same direction when open source technologies are offered as a service, like what SpikeSource is trying to do.

Let's quickly examine each offering (SaaS, Open-Source and Traditional Enterprise Apps) for its strengths and shortcomings, and see whether enterprise customer needs, expectations and concerns are contributing to the rise of some (at the expense and erosion of others), or whether everyone will be able to coexist together to serve their segment of the market.

• Traditional Enterprise Software Vendors
• Companies like Oracle, SAP and other ISVs offering customizable software hosted and deployed in the customer environment.

• Strengths

• Customizability: Most enterprise software is only built and defined to a degree from which consultants can then customize to perfectly fit the solution to exact customer requirements. This approach is often very attractive to customers because (theoretically) they get exactly what they want with little compromise, and it is an easy proposition to get buy-in within an organization.
  

• Control: Solutions are installed and deployed in the customer environment, maintained and managed by trusted staff, policies and processes. This is an important consideration for companies who cannot afford any breakdown of their critical enterprise functions or breach of security over sensitive corporate data.
  

• Ability for Best-of-Breed: This is slightly different than customization in that the customer has options to buy components from different vendors that are pre-certified to work together, creating their own best-of-breed solution. For example, most enterprise customers might choose to use Oracle databases for a SAP solution, or use BEA app server to host an Oracle or SAP enterprise app. Requirements like these are common enough that most enterprise softwa plug-and-play into each other without much customization.

• Costly: It is expensive to select, customize, install, deploy, support and maintain your own enterprise applications. The shortage of qualified IT personnel is one major contributing factor.
  
• Risky: Enterprise projects are notorious for being late and over-budget. It is difficult to manage software projects with big requirements and multiple 3rd party entities. There is also the risk that requirements are mis-defined either due to poor analysis or rapidly changing business environments.
  
• Loss of Agility: Because of the cost and commitment in deploying your own solution, it becomes a barrier to change and evolution of your business processes which can be a critical disadvantage in a fast-moving market.

• SaaS
• Companies like Salesforce and NetSuite.

• Strengths

• Quick, easy and cheap(er): This is an obvious strength of SaaS, leveraging economies of scale to reduce costs associated with provisioning and managing enterprise apps.
• Low commitment and switching costs: Since SaaS can be provisioned relatively quickly, you can turn on and off features you need or no longer need. You can also very quickly move to a different SaaS apps provider if desired without too much overhead other than data migration.
  
• Reliability that comes free: This is especially true for smaller companies who cannot afford to build higher-end reliability and security into your IT environment. With SaaS, you automatically get an IT backend infrastructure that is built for large enterprise deployments at an affordable price.
  

• Fixed feature set: There is little to no customization available for most SaaS apps. What is provided covers best-practice processes that most companies will find adequate.
• Loss of control: Despite SLAs and guarantees, there are companies who rather have their data in-house and use their own staff and processes to deal with reliability and security. SaaS providers serve more than one customers and the chance of a breach (no matter deliberate or accidental) is a real concern.

• Open-Source
• Projects like SugarCRM, JasperSoft, OpenBravo, MySQL, JBoss etc.
  

• Strengths

• Cost-free licensing: No cost to license the software. Of course, that is only part of the cost associated with an enterprise app, but nonetheless reduces the overall cost of ownership in many cases.
  
• Customizability: With access to source code, you can do whatever you want with it.
• Free support: Strong and high-quality community support is often a feature of open-source software, primarily because you have free access to (a potentially large) community with strong vested interests in squashing bugs and helping gaining greater adoption for the project.
  
• Quicker evolution to quality and security: The open access to source code allows the community of users to contribute improvements and fixes in quick cycles. This also means that any security holes will be exploited quickly and consequently, fixed more quickly as well in most cases.
  

• Accountability: An open-source project is a cooperative among many entities and does not indemnify against any defects that cause damage to businesses that use the software. Trusting a community of strangers without accountability can be scary thing for many businesses.

• Analysis
• The aforementioned list of strengths and weaknesses are also important concerns and requirements of customers. If you segment the customers according to size and industry, some concerns become more critical than others.

• Smaller companies are more price-sensitive and less prone to invest for the long-term.
  
• Enterprise functions of young companies are mostly blank-slates, so they often adopt best-practice processes which are already built into most enterprise apps. Older companies have more entrenched idiosyncratic processes that can be costly and disruptive to change.
  
• Companies in certain sectors have critical data (eg. hospital patient info, credit card info, airline reservations etc.) that can cripple the business if compromised and inaccessible for any period of time. Others can tolerate downtime and loss.
• Customizability is critical for businesses whose competitive advantage is the way they operate their business. (Companies like Fedex, Walmart, Amazon.com, and the airline business come to mind.) They will undoubtedly need highly customized solutions. This is an inevitable consequence when a sector matures and gets commoditized, and large companies dominate the space.
• Startup companies often compete on product innovation, and require speed and agility in their business processes to accomodate changing conditions. They will seek the lowest-cost solutions for functions that are not contributing to top-line growth.
  

• What is the best way to deal with it?
  
• Traditional enterprise apps need to acknowledge that SaaS and Open-Source models are non-superficially different from traditional apps. SaaS is not just traditional apps served over the web, and Open-Source is not just a traditional app with their source code published openly. They are very different in the way an application is sold, provisioned and supported. Fixed pricing, self-provisioning, mostly self-supported vs. negotiated pricing, consultant installed and phone/on-site support. A single organization to service all three models that can leverage economies of scale is difficult and less likely to succeed very well.
• Traditional enterprise vendors should spin-off or acquire separate SaaS and Open-Source focused companies and leverage shared customer leads and references, while having management and processes that are completely tuned for each model.
• Traditional vendors should focus on optimizing their service and software offering (and margins) for big customers, attacking the major problems associated with big custom solution projects. Reduce the cost and risk in implementing customized solutions and not worry about the SaaS and Open-Source undercutting the lower-end segment. In the long term, more companies will graduate out of SaaS and Open-Source into the higher-end market which traditional vendors will be primed to serve efficiently.
  
• Traditional vendors should help SaaS and Open-Source succeed and stop competing in something they are not built to do. They shoud offer their base software packages to SaaS and help them transform the apps into services. Since traditional enterprise apps often contain fragments of open-source software, traditional vendors should actively feed patches and improvements back to the open-source community.
• Offer painless transition from SaaS and Open-source to traditional apps if and when companies are ready to do so. Partner with SaaS and Open-source companies to incentivize and optimize data migration, training and IT operations transition.