When Microsoft Vista was announced I compared its cost to the GNP of Jamaica (resulting in a few interesting e-mails). Happy to say Microsoft has seen the light and Vista is now a wonderful upgrade. Just kidding. No doubt someone's head is on a stick outside the front gate at the Redmond compound. Lets not belabor the point since in the past year it seems everyone has taken a shot at Vista. What can we learn from the Vista launch? Vendors take heed- don't overlook a clear value message for the core of your launch. Vista isn't selling because even now no one can articulate the value in moving to Vista. The answer to "why upgrade?" can't be simply "because it's better." Had Microsoft taken a cleaner approach to the message, such as, "It's faster, easier to manage, syncs with everything, and eliminates viruses," it would have had a winner. Instead we got dubious "Vista Ecosystems" whitepapers. Hopefully the Edsel of tech will disappear soon and Microsoft will wow us with the next version of Windows.

Marc Songini, formerly of Computerworld, has been with Nucleus as an analyst for a few months. He has been covering SAP for a long time and launched our "SAP Watch" in February (sign up for our newsletter to get access or just search our Web site for SAP to download the latest). He recently looked at Dow Corning and its experiences with SAP. Now if you've seen any of our previous SAP coverage you'll know we're known for being the tough guys (and gals) in the industry and have taken SAP to task more than a few times (read this). So in that vein I'll give SAP credit. I was on the interview call with Marc and the folks from Dow and there's no doubt Dow's experience has been good. In fact it may be a poster child for a solid SAP experience and the value of being a member of the Americas' SAP Users' Group. If you've deployed SAP or are considering purchasing SAP give the case study a read (and join the users' group). But Bill (Bill Wohl, head of PR at SAP), don't be confused- this doesn't mean we're friends.

Well, that may be a strong statement, but it's not inaccurate. One of our clients recently had a problem getting a project approved by management. The return on investment (ROI) was good at 20% but the net present value (NPV) was negative. Not surprising and that's fine. It does highlight a common misconception about NPV. NPV is a useful tool for normalizing cash flows but for technology investments it should never be used. No, I'm not hedging here. Never use NPV. The problem is residual value. Without an estimate of the residual value of the technology at the end of the period, which for software is effectively zero, then the NPV will always be too low.

Lets look at an example. If you invest $100 in a project that returns $20 per year in benefits for the first three years then the average annual ROI is 20%. Sounds like a good project but the NPV calculation (at 10% cost of capital: -100 + 18.18 + 16.53 + 15.03 = -50.26) isn't so rosy. There's no way around this. Without residual value the NPV calculation is always low and with a guess at a residual value the NPV is, well, a guess. Forget NPV and stick with ROI.

If you want more on the metrics take a look the knowledge center at NucleusResearch.com

We recently published our top 10 predictions for 2008. Have a read and if you get a chance take a look at our analysts' predictions for previous years. It's a pretty good track record.

I was interviewed on Fox Business at the end of December about my views on a few key companies and trends. Here are my quick thoughts from the interview:

Apples everywhere

The iPhone was the launch item of 2007 and Mac sales continue strong. Microsoft Vista's struggles convinced many users to consider a Mac but there may be a storm cloud on the horizon for Apple. The company really needs to embrace the enterprise and prove to the world that the Mac can play in a business environment. If not this may be a 3-year bubble.

On-demand solutions make even greater inroads into the enterprise

It's so much easier to rent software rather than buy it. More and more enterprises are comfortable with the on-demand model driven by companies like Salesforce.com. Companies that provide on-demand solutions will continue to see growth and all technology vendors will need to consider true on-demand as another product to sell alongside the traditional software sales model.

Green is the trendy color

Every IT company will claim some green carbon-offsetting strategy but as nice as it sounds it's likely a one-year trend that will go the way of reengineering the organization and the paperless office. Give it a year to see if it has some staying power.

Google

This could be a make or break year for them. Like Apple, they need to recognize the needs of the enterprise. Google Apps is great but only if I can use those docs in business as well as home. Sure they claim lots of new companies signing up for Google Apps, but Microsoft is well down the Office Suite road and isn't likely to stay still. Time to recognize the need for stability and support rather than always coming up with the next trendy thing.

Microsoft

Likely 2008 will find them on the ropes. Their back office applications are fine and but Vista is a problem child. Time to take a real company retreat and reinvent itself.

Okay, I've been pretty quiet about the iPhone. I wasn't impressed when it first came out, but I thought I'd give it some time. Let's face it, the iPhone has the potential to be an A+ device but at best I give it a C-. Frankly, my list of dislikes is greater than my list of likes. The inability to take movies or send picture messages is simply unacceptable. ATT's ridiculous solution of directing you to a Web site to view a picture someone sent to you was clearly designed by someone with limited respect for the customer. Like a lot of IPhone customers, I was hoping future updates would include missing features, but Apple only seems interested in stopping users from modifying their phones or downloading their own ringtones rather than adding missing functionality. Bottom line? When asked I continue to strongly recommend NOT buying an iPhone. It's just not there yet. Sad really.

As Salesforce.com's announced at its latest Dreamforce event, the company's goal is to become a platform-as-a-service company - enabling IT to design, develop, and deliver applications on the Salesforce.com on-demand platform. Salesforce is essentially telling IT to focus their efforts on design and development and leave the support to them. Their claim is that this will free IT time for more strategic activities but is this just another way to encroach on traditional IT with a new proprietary platform lock-in?

At least one Salesforce.com customer, a large global wealth management firm, didn't think so. Deploying in mid-2006, the company will spend nearly $5 million over the first three years on its new Salesforce.com based client management system. In the process, it turned traditional thinking about on-demand spending on its head. With 45 percent of its total budget devoted to internal development personnel on the Salesforce.com platform. The benefits the group achieved included accelerated time to market (development team can create, test, and deploy new applications in 24 hours that used to take weeks to deliver) and improved client service by integrating customer, case management, and financial transactions into one single dashboard.

The ROI this organization achieved was 112% with a payback of 1.75 years - a bit long compared to typical on-demand solution deployments, but stellar compared to many traditional on-premise CRM deployments. Part of the low ROI is accounted for by the bank's conservative strategy of paying the first year of on-demand fees upfront (making them part of the initial investment).

As Salesforce.com tries to drive more of its customers to move CRM on-demand to platform on-demand, this case example is a great one that challenges some of the traditional thinking about the delivery options of enterprise applications. So now that we have to decide on both the best application and the best delivery method, we can rethink the risk-reward equation about on-demand. Although it may mean relinquishing some control, it may also in the long run mean better applications, less down time and disruption for end users, and lower IT staff turnover.

In my lat posting we looked at the IBM's mix of software to services. Now lets turn out attention to research and development. Customers depend on their software vendors to use a portion of software maintenance fees to both improve existing software and create new solutions to which they can easily migrate. Research and development trends indicate that IBM reinvests less of its revenues into new software development than its rivals:

Research & development as a percentage of revenues

2002 2003 2004 2005 2006

Oracle 11.1% 12.5% 12.3% 12.6% 13.0%

SAP 12.3% 14.2% 12.1% 12.8% 14.2%

IBM 5.9% 6.0% 6.1% 6.4% 6.7%

Here are a few things that IBM rivals are doing with their research and development expenditures that will benefit users:

- Oracle integration assistance. Oracle continues to invest in its platform for helping customers to streamline integration and reduce its costs. It includes pre-built, end-to-end integration applications, as well as reference models and standards that enable both developers and partners to rapidly integrate Oracle and non-Oracle applications.

- SAP's SOA integration. Although Nucleus continues to see SAP deployments that are too lengthy, complex, and costly, SAP has recently been investing in changes to its ERP solution and enterprise applications so that they are pre-integrated with the company's SOA solution, in order to accelerate changes to and integration among a users' applications.

- Oracle Fusion. Oracle is continually investing more in its Fusion architecture, middleware, and application strategy which enables users of applications acquired by Oracle to have a smoother upgrade path and more ready, SOA-based integration.

IBM's business model is drifting away from the creation and sale of software. With so much IBM revenue coming from consultants, and so little of it going into software development, IBM may have different objectives than the CIOs and CFOs to whom it sells. IBM has recently begun consolidating its products, brands, and development teams in an effort to simplify branding and marketing. But this may be too little too late. Oracle and SAP are further along in developing applications that help customers with integration, and they don't rely on partners for their customers' enterprise applications. This means that Oracle and SAP are better positioned than IBM to help their customers cost effectively purchase, deploy, and integrate their applications.

IBM is a critical vendor for many organizations so we weren't too surprised when we had an inquiry from a client who wanted our take on IBM's business mix. Our client's concern was the mix of services and software so our analyst team compared IBM to Oracle and SAP, two other key enterprise vendors.

It should be no surprise that while Oracle and SAP are becoming less reliant on service revenues over past several years, IBM has become more of a service company than a software solution provider. Here is how the revenue mixes of these companies have changed since 2002:

Services as a percentage of revenues

2002 2003 2004 2005 2006
Oracle 27% 24% 21% 20% 20%
SAP 36% 33% 31% 30% 30%
IBM 45% 48% 48% 52% 53%

It's true that IBM still sells plenty of software. In fact, it sells more than SAP and Oracle combined. IBM has tightly focused its product offering databases, middleware, and infrastructure, while relying on partners in order to provide customers with enterprise applications. But IBM's reliance on services may have some implications for its customers:

- Less turnkey solutions. With IBM's profits becoming increasingly reliant on the work of consultants, the company is becoming less motivated to create new solutions that deploy rapidly and simply.

- Costlier upgrades. The less focused IBM is on creating and selling software that installs quickly and simply, the less able it is to create smooth upgrade paths, which are key to minimizing the ongoing costs of operating a solution.

- Longer paybacks. When projects are complex and require customization, they tend to take longer to deploy and generate a payback. This means more uncertainty for CFOs and CIOs, who manage risk as much as they manage finance or IT.

IBM's software revenue mix is also a mixed bag. Much of its revenue comes from reselling other vendors' software, a lower margin business that makes IBM even more reliant on its higher-margin consulting business. Nucleus has seen more than a few low-ROI WebSphere deployments that required the purchase of additional solutions from non-IBM vendors such as Tivoli, Apache, or PTC.

Wouldn't it be nice if there were an easy way to assure value from a technology investment? A simple rule you can apply that guarantees your technology projects will deliver ROI both initially and on an ongoing basis? It turns out there is. By examining the breadth and repeatability of a project you can quickly assess the value the project will ultimately deliver.

Let's start by examining breadth. The more people impacted by the application, the higher the potential ROI. This sounds simple enough: touch a lot of people and you increase the impact of the application. Touch only a few people and the opportunity for value is limited.

You can use breadth in a two ways. The most obvious is to evaluate the potential impact of a project. If the project touches a large number of people, it has greater potential to deliver value. Projects that initially come up short when evaluating the business case may benefit from greater breadth. Try deploying to more people and see if that changes the potential benefit.

Repeatability is the other indicator of value. The more often you use an application the greater its ability to deliver value. Obvious, right? However, Nucleus has seen numerous examples of applications that violate this rule.

One common example is the 401K plan. Here's the problem: it may seem like a good idea to eliminate the paper forms used in this process and allow employees to submit changes using an intranet site " it helps every employee so it has breadth. What it doesn't have is repeatability. In practice, few employees make changes more than once a year, making the time spent re-learning the application greater than the savings. In most cases, hiring another employee in human resources can be more cost effective for the company than creating this type of application.

Breadth and repeatability aren't substitutes for a well-developed business case but they are indicators of potential. Be wary of applications with low breadth and repeatability that promise high returns. The odds are against it. Likewise, be sure to consider breadth and repeatability when evaluating a project that has not lived up to its potential. Increasing either breadth or repeatability can breathe new life into a failing project.

We work with many large enterprise organizations, and this time of year most of them are planning their budgets for 2008. The frustration associated with budgeting invariably brings comments: why bother with a business case? The pessimists point out that the budgets are usually cut anyway, the people who complain the loudest get their projects approved, and everyone inflates the benefits so no one believes the business case. The optimists would counter those points but most of them are still on vacation.

So why bother building a business case and calculating return on investment (ROI), total cost of ownership (TCO), and payback period? The CIO faces three business case challenges: selecting a solution, prioritizing the projects, and articulating the benefits. The metrics in the business case can help with all three - but only if all of your business cases are consistent.

Naturally, most companies use business cases to be sure they've selected the best solution for a particular project. For the project team, mandating a business case brings structure and focus to the project requirements.

The second challenge comes when prioritizing projects and budgeting. Here is where a business case process consistently applied across projects allows for fair and accurate prioritization. If every project has a business case built on a similar methodology then all projects can be easily evaluated for financial benefit.

Finally, at the senior management level the business case information provides the CIO with a valuable tool to both make the case for new projects and accurately convey the returns from past projects. Consistent cases across years and budgeting period helps build credibility with the CFO and it's a lot easier to get a budget approved this year if the CFO can review the benefits delivered to the organization from last year's budget. When budgets get tight, there's nothing stronger than bottom-line returns to free funds from a conservative CFO.

Not ready to mandate a business case for every project right now? There's a simple first step many of our clients have used successfully. Require every project team to list the three top benefits in financial terms over a three-year period from their proposed project. Sure, there may be more than three, but hold them to three only. You'll be surprised at how focused the team becomes.

If you've spent any time flying on commercial airlines lately, it's not hard to believe that reducing costs and streamlining processes are the watchwords of the moment. So is it really surprising that customer satisfaction is another area getting the cost optimization review?

We had one airline client who moved from a mainframe system to implement SPSS to analyze customer satisfaction information. Yes, they know you hate the meals, that's why they don't serve them any more. Here's the interesting part. With SPSS they were able to generate reports much easier and that gave them a free time to start analyzing other data. Say for instance just how angry you are when your flight is delayed.

They decided to correlate actual flight arrival data withy responses to customer satisfaction surveys. What did they find? Customer satisfaction remains fairly constant if a flight is delayed up to 30 minutes - because it's likely the customer will still make their connection. Not surprisingly, after a more than 30 minute delay customer satisfaction plummets. What is surprising is that if the delay stretches to three hours or more, satisfaction starts to rise again. Why? It seems that after 3 hours passengers are resigned to missing their connection, have already made alternate plans, and have likely gotten a voucher from the airline for food or a hotel. So for the airline, it makes no sense to push a 3-hour delay down to a 31-minute delay - customers simply aren't any happier.

We just released a report looking at the Internet and politics. We surveyed 383 people and the results are interesting. When we asked the common sources of political information, with multiple choices allowed, print media still ranked in the top spot with 56%. This was followed by network news then cable news. The Internet made a showing at 4th spot with media-affiliated Web sites but what was last? You Tube, with less than 5%, fell below the obviously biased political party Web sites. It appears that despite all the You Tube hype few are fooled into confusing the Obama Girl with Katie Couric. What's going on? I think many people recognize the difference between edited and vetted news outlets like newspapers and the free-for-all that is You Tube.

We also asked respondents to list who'd been doing the best job using the Internet. Clinton ranked first with 38% followed by Obama with 19%. Everyone else fell to 9% or below. Gore may have invented the Internet but he apparently didn't teach any of his colleagues how to use it yet.

It's somewhat common knowledge that approximately one-third of Microsoft's US employees are in the US working with either a visa or green card. That's a pretty high percentage of Microsoft's workforce and the company gathers a significant percentage of the 65,000 H1B visas that the US issues. Microsoft is lobbying to raise the total number of H1Bs issues to 300,000 in part because it claims it can't find qualified workers in the US. Apparently in Microsoft's world it's unbelievable that their campus recruiting desks are empty. It must be a lack of qualified candidates and not them. Before lobbying congress any more, Microsoft HR may want to take a walk past the Google table and see how they're doing. Microsoft's problem is not a lack of qualified candidates but a lack of candidates finding Microsoft an attractive choice. With strong competition, a stock that hasn't moved in a decade, and an uninspired product set held together by the few brilliant folks still there, it's time Microsoft take a hard look at itself. More H1Bs aren't going to help.

Few would dispute that employees make up one of the most valuable assets of an organization. but payroll, unemployment, Family and Medical Leave Act (FMLA), and countless other rules and regulations make employees very costly. If direct costs weren't enough, employees are probably your most temperamental assets. Employee morale can result in day-to-day productivity changes that would make even early versions of Microsoft Windows look steadfastly reliable.

It's no surprise that labor management applications continue to be a top concern for CIOs. When Nucleus Research studied the benefit of moving from manual timecards to an automated system we found that the payroll error rate for companies using manual timecards was 1.2%. That's 1.2% of the total payroll lost due to errors in calculating pay correctly. With costs like that it's not surprising that most large organizations have already recognized the need to implement a timecard and workforce management solution.

What is interesting is the added benefits organizations receive when they include other modules as part of their timecard solution. Minnesota-based Shooting Star Casino had been using a Kronos timekeeping solution since 2001 but did not have an automated process for monitoring employee absences and leave. Supervisors were spending time manually tracking and updating employee absences and leave in spreadsheets with no formal process for determining when employees were eligible for overtime or FMLA-related leave time, which created a number of problems. It's no surprise that supervisors often approved leave for employees that did not have time or approved FMLA time to employees who were not eligible. A more significant issue was the inconsistencies in granting leave and overtime, which led to employee morale problems. It's understandable that employees assumed favoritism was driving supervisors to grant some employees leave and overtime while refusing others.

Shooting Star Casino deployed Kronos Attendance and Workforce Leave as part of its previously installed timekeeper application. As expected, the solution resulted in increased manager productivity, reduced staff and overtime, improved and more accurate FMLA-related information, and yes, improved morale.

Bottom line? The Kronos Attendance and Workforce Leave application was added to Shooting Star Casino's existing Kronos Timekeeper solution and this minimal investment resulted in a 250% ROI with a 5-month payback. And that doesn't include the value of increased employee morale.