Can Enterprise-Class Solutions Ever Deliver ROI?
Jim HarrisThe information technology industry has a great fondness for enterprise-class solutions and TLAs (two or three letter acronyms): ERP (Enterprise Resource Planning), DW (Data Warehousing), BI (Business Intelligence), MDM (Master Data Management), DG (Data Governance), DQ (Data Quality), CDI (Customer Data Integration), CRM (Customer Relationship Management), PIM (Product Information Management), BPM (Business Process Management), etc. — and new TLAs are surely coming soon.
But there is one TLA to rule them all, one TLA to fund them, one TLA to bring them all and to the business bind them—ROI.
Enterpri$e-Cla$$ $olution$
All enterprise-class solutions have one thing in common—they require a significant investment and total cost of ownership.
Most enterprise software/system licenses start in the six figures. Due in large part to vendor consolidation, many are embedded within a consolidated enterprise application development platform with seamlessly integrated components offering an end-to-end solution that pushes the license well into seven figures.
On top of the licensing, you have to add the annual maintenance fees, which are usually in the five figures—sometimes more.
Add to the total cost of the solution the professional services needed for training and consulting for installation, configuration, application development, testing, and production implementation, and you have another six figure annual investment.
With such a significant investment and total cost of ownership required, can enterprise-class solutions ever deliver ROI?
Should I refinance my mortgage?
As a quick (but relevant) tangent, let's use a simple analogy from the world of personal finance.
Similar to most homeowners, I get offers to refinance my mortgage all the time. A common example is an offer that states I can reduce my monthly payments by $200 by refinancing. Sounds great, $200 a month is an annual cost reduction of $2400.
However, this great deal includes $3000 in refinancing costs. Although I start paying $200 less a month immediately, I do not really start saving any money for 15 months, when the monthly “savings” break even with the $3000 in refinancing costs.
Of course, saying only 15 months is ignoring possible tax implications as well as lost interest or returns that I could have earned since the $3000 likely came from either a savings or an investment account.
Additionally, refinancing might not be a good idea if I plan to sell the house in less than 15 months. The $3000 could instead be invested in finishing my basement or repairing minor damages, which could help increase its value and therefore its sales price.
How does this analogy relate to enterprise-class solutions?
The Business Justification Paradox
Focusing solely on the technical features and ignoring the business benefits of an enterprise-class solution isn’t going to convince either the organization's executive management or its shareholders that the solution is required.
Therefore, emphasis has to placed on the need to make the business justification, where true ROI can only be achieved through tangible business impacts, such as mitigated risks, reduced costs, or increased revenues.
However, a legitimate business justification for any enterprise-class solution is often relatively easy to make.
The business justification paradox is that although an enterprise-class solution definitely has the long-term future potential to reduce costs, mitigate risks, and increase revenues, in the immediate future (and current fiscal year), it will only increase costs, decrease revenues, and therefore potentially increase risks.
In the mortgage analogy, the break even point on the opportunity cost of refinancing can be precisely calculated. Is it even possible to accurately estimate the break even point on the opportunity cost of implementing an enterprise-class solution?
Furthermore, true ROI obviously has to be at least estimated to exceed simply breaking even on the investment.
Given the reality that the longer an initiative takes, the more likely its funding will either be reduced or completely cut, many advocate an agile methodology, which targets iterative cycles quickly delivering small, but tangible value. However, the up-front costs of enterprise licenses and incremental costs of the ongoing efforts and maintenance still loom large on the balance sheet.
Even with “creative” accounting practices, the unquestionably real short-term “ROI high” of following an agile approach could still leave you “chasing the dragon” in search of at least breaking even on your enterprise-class solution's total cost of ownership.
A Call for Debate
My point in this blog post was neither to make the argument that organizations should not invest in enterprise-class solutions, nor to berate organizations for evaluating such possible investments using short-term thinking limited to the current fiscal year.
I am simply trying to encourage an open, honest, and healthy debate about the true ROI of enterprise-class solutions.
I am tired of hearing over-simplifications about how all you need to do is make a valid business justification, as well as attempting to decipher the mystical ROI and total cost of ownership calculations provided by vendors and industry analysts.
I am also tired of being told how emerging industry trends like open source, cloud computing, and software as a service (SaaS) are “less expensive” than traditional approaches. Perhaps that is true, but can they deliver enterprise-class solutions and ROI?
This blog post is a call for debate. Please post a comment. All viewpoints are welcome.
Reader Comments (9)
Good post, Jim. A few points:
Is it even possible to accurately estimate the break even point on the opportunity cost of implementing an enterprise-class solution?
I'm sure that you can approximate it, but it's much more art than science in my view.
I am also tired of being told how emerging industry trends like open source, cloud computing, and software as a service (SaaS) are “less expensive” than traditional approaches. Perhaps that is true, but can they deliver enterprise-class solutions and ROI?
I'd say that they can deliver ROI but I don't know too many learned folks who claim that these technologies will automatically result in the benefits that you described. If SaaS or OS is cheaper than on-premise software, then yes, the ROI should be easier to obtain. Easier and certain are hardly the same, though.
Fantastic question. I think the short answer of course as always is "it depends".
However, whats important is exactly WHAT does it depend on. And I think while the vendors of these solutions would like you to believe that it depends on the features and functionality of their various applications, that what it all depends on far more is the way they are installed, and to what degree the business actually uses them.
(Insert buzz words like "business process alignment" "project ownership" and "Business/IT collaboration" here)
But if you spend Gazillions on a new ERP, then customize it like crazy to ensure that none of your business processes have to change and none of your siloed departments have to talk to each other (which will cost another gazillion in dev and consulting by the way), which will then ensure that ongoing maintenance and configuration is more expensive as well, and will eliminate any ability to use pre-built business intelligence solutions etc. etc.... Your ROI is going to be a big, negative number.
Unfortunately, this is often how it's done. So my first comment in this debate is- If enterprise systems enable real change and optimization in business processes, then they CAN have ROI. But it's hard. And doesn't happen often enough.
Is enterprise DQ worth the price?
The scale of the project matters. You may be see situations where the return might be $30,000 on a $20,000 investment, but if the data has been sorely neglected and is an important part of a business process, enterprise projects can deliver phenomenal ROI. In some cases, the ROI you encounter is more on the scale of getting $200,000 on your $20,000 investment. In my years of working in this marketplace, I have no doubt that the 10X (and higher) ROIs are real, not so much as a function of the vendor solution, but as a function of the organization finally making the choice to properly manage data.
Are new, less expensive approaches like open source, cloud as good as traditional ones?
Some folks still like vinyl records for their music but I prefer an MP3 player for its portability, performance and convenience. The technologies you mentioned represent evolution in our field. In the late 1980s (when vinyl was king) most companies accomplished data integration and data quality with service bureaus because it was too expensive and difficult to set up the mainframe you needed to manage data. Today, we have powerful servers, unlimited storage and now cloud computing. In the age of MP3 players, we have easier access to data managment software (open source) and public domain reference data. It’s about evolution and we must embrace it.
When I was Director of BI at T-Mobile a few years back, we had to put business cases together for our projects and had to prove an ROI that was better than other IT projects over a three-year time period. The business cases were first vetted through our Corporate Strategy and Analysis team to ensure the numbers were reasonably accurate. Then those projects came to the table to "compete." OK....sounds good so far.
Where it does go wrong is that there is no measure of ROI some three years later to validate what the business case predicted. Hence no one is held accountable for what they put in the business case. The great missing link. If I were responsible for IT, I'd look for more ways to do smaller projects with quicker ROI. I'd also put high-stakes on those business cases - even putting your job at stake. Then see how those business cases shape up.
Terri
P.S. What always made me chuckle were all the claims to reduce subscriber churn. If we actually did implement all the projects, we'd see negative churn?!
What I have seen, particularly in the Telecoms industry, is that as the industry (and the people) moves so fast, sometimes an enterprise solution has been implemented, and no sooner been placed on a legacy roadmap, due to competitor acquisition, redirection of funding, departmental restructuring etc. Therefore it's often hard to demonstrate whether maximum ROI has been delivered or not. I like the point Terri makes about accountability above - it ties in nicely to this point.
Working in a position where I am constantly 'herding the cats' toward enterprise solutions, much of the ROI can come from the ability for organizational units within an enterprise to be able to share data across/down and up the 'food chain'.
When each office has its own solution things can get a bit interesting. imagine one organization having WordPerfect, Open Office, Lotus notes, MS Office 2003/2007/2010, Google Docs, etc etc and there is *NO* common format that they can all be shared/opened in. Considerable time would be spent emailing documents back and forth requesting format changes from this product to that. While that time can be measured in terms of % of time * hourly rate - the actual ROI is much higher than that single dollar figure.
The chances a large ERP or DW project can deliver actual ROI in a couple of years are actually small. Nonetheless, even a moderately complex company can't do without. The basic value chain processes must be regulated within a common framework. All the rest is better served by small and specific applications. In my experience a following data integration project is far more welcomed by stakeholders than "doing right" and placing the process within the common framework. Management by Excel is here to stay.
Thanks everyone for your comments. As always, your feedback is greatly appreciated.
@Phil Simon Yes, ROI calculations are often more art than science because there are a myriad of complex variables. New technologies have some intriguing potential, about which I think your new book The Next Wave of Technologies provides a balanced viewpoint.
@James Yes, enabling real (necessary and meaningful) change and optimization of business processes is the key to delivering true ROI. And I agree that although the potential exists, it is sadly all too often not realized.
@Steve Although I agree that ROI is not so much a function of the vendor solution, but a function of the organization finally making the choice to properly manage data, I have to question the financial overhead of most traditional enterprise-class technology. I also can't help but wonder if we are witnessing an evolution or a revolution in data management since newer technologies have a paradigm that is highly disruptive to the revenue model of the leading vendors, which continue to consolidate into mega-vendors at least in part to try to fight the future - similar (philosophically at least) to the way the music industry tried to reject the unrelenting future of MP3 and online music.
@Terri The scenario you described adds another angle to what I called the business justification paradox. A lack of accountability for the business case is almost worst than not having a business case at all.
@Phil Wright Yes, the speed at which a new enterprise solution can quickly become a legacy application to be replaced is indeed troubling. I have witnessed many elegantly designed solutions that delivered the functionality mandated by the business - but the time lag between requirements and delivery changed the requirements so that the solution was no longer applicable. Agile development continues to be emphasized, in part, for this very reason. Although an agile approach can make the enterprise more adaptive, it doesn't guarantee ROI - especially when the the up-front costs of enterprise licenses and incremental costs of the ongoing efforts and maintenance are factored into the equation.
@Glenn Yes, I have witnessed very large organizations compounding the ROI challenge by different business units being allowed to implement separate (siloed) enterprise-class solutions - and then another "solution" is sought to consolidate them.
@Augusto You raise an excellent point about focusing on the basic value chain processes and delivering them as an enterprise-class solution within a common framework, and then serving the rest of the business needs using small and specific applications, where leveraging tools like your Viney@rd product can help with this approach.
I think folks in the C-suite like to bandy about terms like project ROI and the like, but have little or no clue about what actually goes into coming up with a number.
To further your mortgage analogy: I know a lot of people who think they are making a lot on real estate or they expect to make a lot. They buy a house for $300,000 and when the market value goes up to $320,000 next year they are ecstatic! They have made $20k! But have they really? How much interest did they pay? How many repairs and maintenance amounts were paid out? How much in property tax? Legal? What's the transaction cost on a sale?
Typically, none of these externalities are considered and, in my view, the same is true for most Enterprise implementations. A colleague of mine likes to point the situation that sunk the jewelery company, Shane Co. In their bankruptcy filing, one of the reasons cited was a failed SAP implementation. The thing took 2 years and $30M more than was estimated. Whoops. Money pit.
I'm bias on the SaaS argument being that I work for a SaaS B.I. company, but I can tell you one distinct advantage regardless of whether or not ROI is actually calculable. That is, no sunk costs. It's a HUGE advantage. SaaS is quick to prove whether or not it will be a viable solution for a business. If it doesn't, you cut the cord and don't feel the least bit bad about it. If you've sunk $5M into something, you might feel "pot committed" if I may mix metaphors a bit here. Who goes "all in" before the flop?
Great post Jim! Cheers!