Enterprise Software Success Myth #5

Doug Hadden, VP Products

FreeBalance is a medium-sized Independent Software Vendor (ISV) with considerable success competing against very large Enterprise Resource Planning (ERP) vendors. We are sharing 16 lessons learned by bucking conventional wisdom to encourage industry innovation and creativity.

Conventional View

Enterprise software companies scale by creating a services channel. This channel promotes ERP and other types of software while supplying labour for implementations. It is thought that labour cannot be scaled within a software manufacturer. Professional services staff at enterprise manufacturers should be focused on a small minority of implementations that extends products. In fact, high professional services revenue devalues enterprise software companies.

Symptom

Large numbers of expensive failures in enterprise software implementations may come from a disconnection between what companies build and how customers use products. This may come from a lack of on-site implementation knowledge for what Michael Krigsman calls the “devil’s Triangle” .

Emerging Trends

• Enterprise software companies are relying most heavily on after-sales maintenance revenue rather than meeting customer needs directly
• Maintenance model from ERP companies is increasingly under attack from 3rd party maintenance organizations and cloud computing providers.
• Failures in ERP implementations are regularly blamed on the victim – the customer .
• ERP vendors determine what new features will be presented in software, usually described at user conferences. These decisions are made with little or no on-site customer experience.
• ERP vendors rarely attend Public Financial Management conferences, and rarely participate to learn about new trends.

FreeBalance Approach

FreeBalance operates “in network” by seeking customer interaction. We realized that the “arms’ length” approach used by ERP vendors increases the likelihood of failure. As I described in 2011, software manufacturer involvement is critical to ensuring success for GRP in emerging nations and developing countries.

Other aspects of the FreeBalance approach include:

• Ensuring that FreeBalance is part of every implementation team. And, ensuring that implementation partners are focused on customer success .
• Using the venue of the FreeBalance International Steering Committee to change the product roadmap to meet customer needs.
• Deep commitment to PFM by participating in conferences and NGOs like FreeBalance ICGFM .
• Also, a commitment to share good practices and PFM futures with the broader PFM community.
• Local hiring to scale implementation commitments, build local capacity and improve insight into changing customer needs.

Doug Hadden, VP Products

We started a sharing good practices white paper series last month. This is part of our mandate, as a social enterprise, to share lessons learned in technology and good governance to the broad Public Financial Management (PFM) community. We published our latest good practice document to coincide with the upcoming East and South Africa Association of Accountants-General conference next week in Botswana.

The genesis for this good practice document is interesting. FreeBalance is a Canadian company based in Ottawa with a large installed base in the Canadian Federal Government. So, we have some insight into the complexities of public finances and human resources in a G8 country. Canada is considered to have one of the most advanced structures for governance in the world. And, our Canadian customers, thanks to the governance of “clusters”, have driven our products for almost 30 years.

Our mission has been to take our robust Government Resource Planning (GRP) software to less developed countries. Good governance, in my opinion, is not a “zero sum game.” It has a network effect in that improved governance in one country has positive effects in other countries.

The dark side of success?

In the course of events in the past decade, procurement cycles for government Integrated Financial Management Information Systems (IFMIS), as they are often called, were accelerated for post-conflict countries. The international community recognized that software and capacity building was necessary to rebuild government. And, FreeBalance software that was highly flexible for government, and only government with support for Canadian government requirements from decades past, gained a foothold.

The ultimate reward for success in fragile states – the only COTS vendor to have success under these difficult conditions – was an assumption that our software and expertise was only viable for “underdeveloped countries”. That’s certainly been the fallacy that major Enterprise Resource Planning (ERP) like to propagate.

There are some interesting observations that I have made from our on-going research into the PFM domain that may be of interest to you:

1. Many developing countries have leapfrogged developed countries particularly with the support for International standard, budget transparency and multiple year planning. Some might say that there is a double standard where donor countries and multilateral financial institutions expect better governance mechanisms in post-conflict countries than they support themselves.
2. The success rates for ERP software in government is meager. In very developed countries with high human capacity and good project methodologies. We encounter so many stories of ERP problems in government that we often fail to realize that ERP does not have a good track record in the private sector. The “enterprise” sector. It’s true that we update our ERP failure ERP Fail page as we learn about problems that have been reported publicly. We learn about far more failures and issues that are not reported. Which stands to reason – it is very embarrassing to country governments or IFIs to expose these problems.
3. The term “innovation” is frequently used in the enterprise software domain to the point where it’s lost all meaning. Vendors tout mobile technology yet the core of their software is client/server. They throw hardware (in-memory) to accelerate processing as if this hasn’t been done before by anyone. And, they continue to present the fiction that the larger the company then the lower the risk to customers.

Good Practices in Government Resource Planning, Developed vs Developing Countries by FreeBalanceGRP

Doug Hadden, VP Products

Word came a little over a week ago about another Enterprise Resource Planning (ERP) failure in government. This time, a payroll implementation in the State of California. I was involved in some conversations on twitter that I storified below.

Relevant Details

• Project name: MyCalPAYS
• Functionality: Civil Service Management including payroll, position management, leave accruals, time & attendance, benefits, employee self-service, reporting
• Functionality not included: talent & succession management, training, recruitment & e-recruitment, compensation management, personnel cost planning
• Includes business process re-engineering to leverage “out-of-the-box” functionality
• Total number of employees when project was to be completed: 240,000 including 160 departments and 21 different bargaining units, State University campuses
• Total number of employees in first phase: 1,300
• 8 months of tests with 1,300 employees with no run without errors
• Original testing was with a sample set of 22,000
• Project started in 2003 with original contract provided in 2006
• Fired the original systems integrator in 2009 replaced by the software vendor
• $26M paid to original systems integrator and $50M paid to the vendor
• Reportedly $371M total budget of which $254M has been spent
• Data migration seen as a major problem
• Dedicated staffing by the State with a governance structure, project plan and change management processes
• Risk management identified of 16 of 34 attributes as higher complexity (3 or 3.5 out of 4) and overall complexity rating of 2.6 out of 4

Initial Analysis

• It is a misnomer to suggest that the amount spent was for 1,300 employees, the total $371M budget is for 240,000 employees
• The $371M cost seems to assume all hardware, maintenance and personnel costs including the State team: the full Total Cost of Ownership over a relatively lengthy period
• Saying that $371M is an order of magnitude higher TCO than we experience for similar scope Civil Service Management implementations for complete countries
• There has been sufficient time and budget to develop a full government COTS solution (we did in the same time frame at lower costs)
• Some analysts seem to think that it was a mistake for the software vendor to step into the project, but my view is that this is better to have the manufacturer in the governance structure than at arm’s length
• Payroll implementations without salary planning and management in the public sector, especially in a developed country, is probably a mistake because governments are budget driven
• There was a desire to use generic processes from the COTS software in order to reduce risks and accelerate implementation
• On the surface, project and risk management processes run by the State appear adequate when compared to other projects that I’ve seen

What could have gone wrong?

The “blame game” is in full force. The general consensus is that the State is probably at fault with systems integration, including the manufacturer, running a close second. Most observers believe that the ERP software is less at fault.

My perspective, coming from a Government Resource Planning (GRP) specialist, is somewhat at a polar opposite to the general consensus.

1. ERP software has been noted for high failure rates in the public sector – therefore it is unlikely that the victim should be blamed for all failure and more likely that the software is a significant contributor
2. Significant, if not overwhelming budget, is another clue about the difficulty of customizing software designed for the private sector for the public sector
3. Vendors have a moral obligation to advise public sector clients and augment project management processes and teams because it is public money

There are some significant but well-understood differences between public and private sector human resource systems. These differences often require significant code customization and therefore higher project risks in government. Some of these factors include:

• Some process reform to follow many out-of-the-box processes in private sector COTS is often not possible without changes to the law – like it or not, code customization is required
• Salary calculations including the modeling of collective agreements across one or more sets of salary scales shows how public sector payroll can be more complicated to articulate than in the private sector
• Secondment and shared personnel are more typical scenarios in the public sector
• Leave accruals and benefits calculation also tends to be more complex because of longer-term entitlements and complex flexible benefits
• Time and attendance including seasonal workers and overtime rules can also be more complex in the public sector

Research Sources

1. http://www.itworld.com/software/341305/california-terminates-contract-sap-over-massive-troubled-it-project
2. http://www.allgov.com/usa/ca/news/where-is-the-money-going/another-state-technology-project-goes-awry-this-one-costs-371-million-121226?news=846575
3. http://www.sco.ca.gov/21century.html
4. http://www.cio.ca.gov/government/it_policy/pdf/0840-070_spr3.pdf
5. http://www.lhc.ca.gov/studies/activestudies/infotechupdate/21stCenturyAug2011.pdf
6. http://www.sacbee.com/static/weblogs/the_state_worker/090406%20SCO%2021st%20Century%20Project.pdf
7. http://www.lao.ca.gov/handouts/state_admin/2011/21st_Century_Project_1_24_11.pdf
8. http://www.scribd.com/doc/51191244/TFC-Project-Charter-v2x

Tracing an ERP Fail Story

Interesting reaction to the lawsuit from the State of California related to a failed ERP project.

Storified by · Sat, Feb 16 2013 07:10:21

Doug Hadden, VP Products

Do government organizations reduce risk by acquiring Tier 1 ERP solutions rather than alternatives? Many consultants and purchasing decision-makers seem to think that there is less risk when buying from the “Big 2” ERP vendors rather than, for example, a Government Resource Planning (GRP) specialist like FreeBalance. Yet, a 2012 study by Advanced Computer Software in the UK found low satisfaction with ERP in government Over half of the respondents were using Tier 1 ERP software. The survey showed that Tier 1 ERP satisfaction is far lower than with alternative solutions. Users of non-Tier 1 ERP solutions were 4 times more likely to rate their solution as exceeding expectations on any of 5 dimensions and half as likely to rate their solution as worse than expected.

• 63%: ERP system did not meet expectations in at least 1 area
• 60%: Would choose a different company to implement the ERP if had to do it over again
• 50%: Tier 1 ERP implementation costs higher than expected
• 45%: Tier 1ERP implementation took longer than expected
• 43%: Would choose a different software than the ERP implemented if had to do it over again
• 40%: Tier 1 ERP ability to meet government requirements without customization lower than expected
• 39%: Tier 1 ERP system ease of use worse than expected
• 25%:Tier 1 ERP ability to meet government needs after customization lower than expected
• 20%: ERP system had negative impact on organization
• 20%: Would go through a similar ERP implementation again

Where is the Risk?

There are numerous areas of risk identified in The Advanced Computer Software study and elsewhere:

1. Over budget
2. Late
3. Not achieve expected results
4. Difficulty to use
5. High maintenance costs (TCO)
6. Difficult to adapt to changing processes

After reading about yet yet another ERP failure in government , analyst Frank Scavo recommended that governments adopt a risk-based approach His 5 points about risk management in ERP are well worth adopting.

The ERP in Government Open Secret

We’ve been collecting more and more public stories about ERP failure in government by updating one of blog pages. The reaction to these failures from the technology community tends to focus on consulting or customer deficiencies. Rarely is the software itself under question. That might be a result of assuming that software so widely adopted across the public and private sector must be effective. Not to mention the hype coming from the Tier 1 vendors.

There is an observation in the Advanced Computer Software study that Tier 1 ERP was developed for the private sector and is not applicable to the public sector. There seems to be growing realization that Tier 1 ERP is highly risky and overly expensive, in government.

And, ERP is wasting public money in an era of budget austerity.

Why is GRP Less Risky than Tier 1 ERP in Government?

The GRP approach addresses the 6 risk elements directly:

1. Budgets: GRP software is unencumbered with private sector functionality and requires less consulting effort and have smaller data centre need than the ERP software bloat
2. Time: GRP software is developed for the government domain with set blueprints to enable more rapid implementations
3. Expectations: GRP vendors, like FreeBalance, set proper expectations and employ PFM experts to help mentor government professionals
4. Ease of Use: GRP software is developed for government usage and is more intuitive for commitment accounting functionality such as handling multiple aggregate budget controls
5. TCO: the use of configuration rather than code customization accelerates implementation and reduces the maintenance burden making the software more financially sustainable
6. Adaptability: configuration also enables change, or what we call progressive activation

In addition, governments deal directly with the GRP manufacturer in our case. That puts us in the governance structure and ensures that we are committed to implementation success. We also take part in all implementations to remove the incentive to drive customization revenue.

How should Governments Protect themselves during Procurement?

Government RFPs for financial management systems include requirements about vendors. This includes a minimum amount for total turnover, numbers of implementations or numbers of users. Many of these requirements do little to reduce risk. Some are irrelevant. Others ensure that the government will take a high risk. Here are some suggested Request for Proposal (RFP) alternative requirement to protect governments from high risk:

• Total amount of revenue from PFM systems (because private sector revenue isn’t relative)
• Do not specify the need for “ERP” or call the project “ERP” (because that’s what you’ll end up with and you will have reduced competition)
• Experience in PFM under similar circumstances (because “success” in more developed countries, different levels of government or in the private sector isn’t relative)
• Willingness of the software manufacturer to commit to feature requests (to reduce or eliminate the high long-term cost of customization)
• Demonstrated success with multiple configuration changes, upgrades and addition of modules (because TCO needs to be measured over time)
• Turnkey costs and 5 year TCO in fixed priced contract (so that the vendor cannot bid low and overwhelm the government with change order costs
• Study tour: go and see the software in action – you’d be surprised that a high number of publicized ERP in Government “success stories” are far from successful

Overview

• Governments are increasingly adopting Commercial-Off-the-Shelf (COTS) software to replace legacy and custom developed software applications for financial, budget, expenditure, tax, treasury and civil service management.
• A major impetus for recent COTS projects is to replace multiple applications within a government organization with one integrated solution or to support numerous government organizations with a hosted shared service or private government cloud.
• Government organizations can chose to acquire Enterprise Resource Planning (ERP) software from large software firms whose software is used in multiple “vertical” markets or Government Resource Planning (GRP) software designed exclusively for governments.
• FreeBalance, with the FreeBalance Accountability Suite, is a GRP provider. FreeBalance does not build or provides software to the private sector.
• There are high incidents of failure of ERP implementations in government from late delivery to over budget to inability to achieve expected benefits. Many ERP implementations, even in the most advanced countries, fail. The purpose of this web page is to keep up-to-date with ERP failures in government.

Large ERP project failures in developed country governments

A large ERP shared services project in France country was estimated to be $200M over budget by the audit office and more than 1 year late and resulted in late payments of over $2.2B to defense contractors

The National Audit Office in the United Kingdom found that the use of ERP shared services added rather than reduced costs

In the United Kingdom, describing a Cabinet Office analysis found:

• Average cost to deploy a Tier 1 ERP is £160 per employee using the traditional method
• Theoretical cost to deploy via shared services is  £93 per employee
• Use of lower cost solutions at £52 per employee

Another analysis in the UK found inconsistent software license pricing in government for the same Tier 1 ERP package.

A 2012 study in the UK found low satisfaction with ERP in government Over half of the respondents were using Tier 1 ERP software. The survey showed that Tier 1 ERP satisfaction is far lower than with alternative solutions. Users of non-Tier 1 ERP solutions were 4 times more likely to rate their solution as exceeding expectations on any of 5 dimensions and half as likely to rate their solution as worse than expected. >/p>

• 63%: ERP system did not meet expectations in at least 1 area
• 60%: Would choose a different company to implement the ERP if had to do it over again
• 50%: Tier 1 ERP implementation costs higher than expected
• 45%: Tier 1ERP implementation took longer than expected
• 43%: Would choose a different software than the ERP implemented if had to do it over again
• 40%: Tier 1 ERP ability to meet government requirements without customization lower than expected
• 39%: Tier 1 ERP system ease of use worse than expected
• 25%:Tier 1 ERP ability to meet government needs after customization lower than expected
• 20%: ERP system had negative impact on organization
• 20%: Would go through a similar ERP implementation again

In the United States, reports by the United States Government Accountability Office (GAO) and the Department of Defense (DoD) Inspector General (IG) found that 11 of 13 ERP projects were over-budget costing American taxpayers Billions of Dollars with one ERP project resulting in $1B “largely wasted” and another project stopped after 7 years and $1B invested would require an additional $1.1B for about a quarter of the original scope

Major difficulties experienced in ERP implementations in the public sector in developed countries

ERP failures and cost overruns in the public sector have resulted in difficulties, contract cancelations and lawsuits, although lawsuits are rare because vendors would rather do what it takes to make the situation right than face potential public-relations damage from a high-profile legal battle:

• Australia: State government ministry
• Canada: non-profit, city
• France: city
• Ireland: 2 projects within an important ministry
• United Kingdom: university, councils (1) (2) (3) and government ministries (1) (2) (3)
• United States: non-profit, school districts (1) (2), universities (1) (2) (3) (4) (5), (6),  a police department, cities (1) (2) (3) (4), counties (1) (2),  state government (1) (2) (3) (4) (5),(6), federal government (1) (2) (3) (4) (5) (6)

KPMG found in 2011 some ERP in government patterns:

• Budget overruns particularly in implementation where additional software customization was needed: typically 6 months to a year
• Only 57% of implementation projects stayed in budget
• Many survey participants could not recall the original budget

Reviews of top ERP failures in 2010, 2011 and 2012 for IDG journalist Chris Kanaracus found that slightly over half came from the public sector despite representing only 20% of the market size.

What is the evidence of ERP success in emerging economies?

A 2003 Study by the World Bank found a lack of success in government FMIS implementations whether ERP, GRP COTS or Custom developed:

• 43% delivered as specified
• 50% delivered on budget
• 21% delivered on time
• 25% unsustainable
• 69% likely to be sustainable
• 6% highly likely to be sustainable

A 2011 Study by the World Bank found that FMIS implementation sustainability has improved but that 18% remain unsustainable.

Examples of ERP in Government implementation problems in emerging economies including Azerbaijan, Costa Rica, Ghana, Kazakhstan, Malawi , Maldives, Russia, Rwanda, Uganda , Zambia and Vietnam.

Evidence of limited success rates with ERP implementations across multiple industries

Success Rates

• 25% indicated that organization had not recovered ERP implementation costs
• Between 20% and 35% of all ERP implementations fail
• 51% of ERP implementers judged their implementation as unsuccessful
• Nearly 70% of large projects were found to be improbable of achieving project success
• 41% of ERP projects achieved all or the majority of benefits
• 60% received 50% or less of anticipated benefits
• 40% of ERP projects failed to meet business case within 1 year of going live
• Average for ERP projects was 50% of intended benefits
• ERP projects saw the implementation of -59% of intended functionality
• 34% of respondents “very satisfied” with ERP project
• Fewer than 1/3 of decision-makers recommend their ERP vendor
• ERP vendors achieved a D+ in return on investment
• 75% of organizations experienced a productivity dip after implementing ERP
• 20% of survey respondents terminated ERP projects

Costs

• 53% of companies reported budget overruns
• 40% of all ERP projects will exceed estimates by 50% or more
• Average ERP cost variance was 182% of budgets
• Average ERP cost variance was 178% of budgets
• Average ERP implementation cost was 40% over budget
• Average ERP implementation takes 23 months, has a total cost of ownership of $15 million and with an average negative net present value of $1.5 million
• Users of Tier 1 ERP vendors will experience higher Total Cost of Ownership (TCO) than users of Tier 2 vendors
• ERP upgrades cost about ½ the value of the original license fee and 20% of the original implementation costs
• Experts warn that organizations should expect to pay as much as 3 times the original ERP software cost to upgrade to new technology
• Interviews with reference customers of a major ERP vendor found that 57% did not achieve a positive ROI

On-time Delivery

• 61% of projects exceeded timelines
• Project duration averages 17.8 months for first phase
• Average implementation time from the 2 major ERP vendors are 17 and 18 months with average variance of 2 and 4 months over schedule
• Average variance for ERP projects was 230% of schedule
• Up to 80% of ERP exceed time and budget estimates
• 70%  of respondents stated their ERP project timeline was inadequate

The FreeBalance survey comparing ERP and GRP experience in the Government of Canada

FreeBalance completed a survey at the Government Technology Exhibition and Conference (GTEC), November 5 to 8 2012 and the Financial Management Institute of Canada (FMI) Professional Development Week, November 27 to 30 2012.Both of these conferences were held in Ottawa. We received 207 responses.

Over 70% of respondents believe that FreeBalance software has a much lower TCO in government financial management than ERP. Almost 70% suggested that the FreeBalance TCO was between 10 and 50%.

Analysis of GRP vs. ERP 5 Year TCO

FreeBalance competes internationally against major ERP providers. Price quotations are often made public during bid openings in many countries. Most of the international requirements call for all costs over a 3 or 5 year period including software licenses, implementation, support, training, middleware and hardware. Although some of the proposals were not this 5 Year TCO, Tier 1 ERP prices average 191% FreeBalance prices.

Doug Hadden, VP Products

Large IT projects are not for the feint of heart. Or, the risk adverse. Despite our high success rates in Government Resource Planning (GRP) relative to other solutions, I can tell you that implementations are rarely problem free

It was with interest that I read last month Chris Kanaracus‘s article “The scariest US software project horror stories of 2012.” I found 2011 and 2010 year-end articles about the largest ERP failures of the year from Kanaracus.

29 horror stories in total.

15 in the public sector.

That’s right: a touch over half.

In a market that is about 20% of the entire IT market.

So, if the stories are representative of IT failure, that means that public sector implementations are between 2 and 3 times more risky than the private sector.

Analysis

The culprits here are software and process. Software designed for the private sector and shoe-horned into government and custom-developed applications add significant risk.

The political nature of the public sector and the movement of public servants in organizations adds additional risk. FreeBalance has achieved success rates well above average with software designed for government AND business processes adapted to the government context.

Believe me, Project Management 101 is not enough.

Doug Hadden, VP Products

The evidence mounts – ERP has a high failure rate in government. ERP vendors, whose software is written for multiple industries, suggest that combining multiple ERP instances into a single shared service will result in “economies of scale” savings. And, it has a grain of truthiness, but, as I’ve written before, it is very much the promise of “free beer tomorrow”: this magic point at which ERP provides a better value than using best-of-breed like the FreeBalance Accountability Suite never happens.

KPMG found in 2011 some ERP in government patterns:

• Budget overruns particularly in implementation where additional software customization was needed: typically 6 months to a year
• Only 57% of implementation projects stayed in budget
• Many survey participants could not recall the original budget

A recent article from the UK describing a Cabinet Office analysis found:

• Average cost to deploy a Tier 1 ERP is £160 per employee using the traditional method
• Theoretical cost to deploy via shared services is  £93 per employee
• Use of lower cost solutions at £52 per employee

The Total Cost of Ownership (TCO) should be a critical concern in government technology procurement. Although “portfolio management” is a value proposition presented by ERP companies, our experience shows that up-front costs have little bearing on the total cost. Tier 1 ERP packages generate high costs for customization, maintenance, change and training. It’s not a good business case as I described back in 2009 (and tried my best to explain the value proposition of ERP in government as objectively as I could.)

A Study of the Literature

Doug Hadden VP Products

We received some interesting feedback from a recent blog post about cost overruns in ERP projects at the United States Department of Defense. I’ve been vocal about other examples of ERP failure within the public sector. It’s surprising how well major vendors are able to market solutions to government despite this lack of success. Perhaps there is so much marketing noise that it’s difficult for governments to uncover the evidence. Nevertheless, there is a significant amount of public information that supports anecdotal evidence. Some of the studies are dated – but if failure rates have been reduced by 50% in the last 5 years – that still means that 10% to 20% of all ERP implementations are a complete failure.

Let me know if you have evidence to the contrary or lessons learned in the comments section.

Should governments consider ERP or GRP?

• Developed country governments are increasingly adopting Commercial-Off-the-Shelf (COTS) software to replace legacy and custom developed software applications for financial, budget, expenditure, tax, treasury and civil service management.
• A major impetus for recent COTS projects is to replace multiple applications within a government organization with one integrated solution or to support numerous government organizations with a hosted shared service or private government cloud.
• Government organizations can chose to acquire Enterprise Resource Planning (ERP) software from large software firms whose software is used in multiple “vertical” markets or Government Resource Planning (GRP) software designed exclusively for governments.

There are many large ERP project failures in developed countries

• Reports by the United States Government Accountability Office (GAO) and the Department of Defense (DoD) Inspector General (IG) found that 11 of 13 ERP projects were over-budget costing American taxpayers Billions of Dollars with one ERP project resulting in $1B “largely wasted”
• A large ERP shared services project in France country was estimated to be $200M over budget by the audit office and more than 1 year late and resulted in late payments of over $2.2B to defense contractors.
• National Audit Office in the United Kingdom found that the use of ERP shared services added rather than reduced costs in one case and resulted in a negative net present value in another case

There are major difficulties reported in ERP public sector implementation in developed countries

ERP failures and cost overruns in the public sector have resulted in difficulties, contract cancelations and lawsuits, although lawsuits “are rare because vendors would rather do what it takes to make the situation right than face potential public-relations damage from a high-profile legal battle”:

• Australia: State government ministry
• Canada: non-profit, city
• France: city
• Ireland: 2 projects within an important ministry
• United Kingdom: university, councils (1) (2) (3) and government ministries (1) (2) (3)
• United States: non-profit, school district, universities (1) (2) (3) (4) (5),  a police department, cities (1) (2) (3) (4), counties (1) (2),  State government (1) (2) (3) (4), federal government (1) (2) (3) (4) (5)

Studies show a lack of ERP success across all industries

• Between 20% and 35% of all ERP implementations fail
• 51% of ERP implementers judged their implementation as unsuccessful
• Nearly 70% of large projects were found to be improbable of achieving project success
• 41% of ERP projects achieved all or the majority of benefits
• 40% of ERP projects failed to meet business case within 1 year of going live
• Average for ERP projects was 50% of intended benefits
• ERP projects saw the implementation of -59% of intended functionality
• 34% of respondents “very satisfied” with ERP project
• Fewer than 1/3 of decision-makers recommend their ERP vendor
• ERP vendors achieved a D+ in return on investment
• 75% of organizations experienced a productivity dip after implementing ERP
• 20% of survey respondents terminated ERP projects

Studies show significant cost and schedule overruns in ERP implementations across all industries

Costs

• 40% of all ERP projects will exceed estimates by 50% or more
• Average ERP cost variance was 182% of budgets
• Average ERP cost variance was 178% of budgets
• Average ERP implementation cost was 40% over budget
• Average ERP implementation takes 23 months, has a total cost of ownership of $15 million and with an average negative net present value of $1.5 million
• Users of Tier 1 ERP vendors will experience higher TCO than users of Tier 2 vendors
• ERP upgrades cost about ½ the value of the original license fee and 20% of the original implementation costs
• Experts warn that organizations should expect to pay as much as 3 times the original ERP software cost to upgrade to new technology
• Interviews with reference customers of a major ERP vendor found that 57% did not achieve a positive ROI

On-time Delivery

• Average implementation time from the 2 major ERP vendors are 17 and 18 months with average variance of 2 and 4 months over schedule
• Average variance for ERP projects was 230% of schedule
• Up to 80% of ERP exceed time and budget estimates
• 70%  of respondents stated their ERP project timeline was inadequate

Why is ERP so unsuccessful in government?

ERP software is designed for the private sector across many industries that rarely provide a good value for money to governments. Large-scale public sector ERP implementations additional time is required during the analysis and design phase to focus on the gap between the commercial process and the required process

1. ERP cost overrun and the failure to meet schedule are because ERP software requires significant software complex code customization (BPM scripts, call-outs & software development) to meet government requirements that extends implementation cycles. The ratio of services to software cost in the public sector is estimated to be three time that in the private sector or up to 15 times the cost of software.
2. High maintenance costs come from maintaining complex code through problem troubleshooting, and difficult upgrades that increases the Total Cost of Ownership (TCO). For example, the Government of Canada internal support for 15 different customized versions of a major ERP package received an award for saving more than $12M annually in “cost avoidance.” This is hardly a definition of IT success.
3. ERP functionality is often complex and hard to use in the private or public sector. 46% of ERP implementers characterized that their organizations were not able to understand how to leverage features to improve the way that they did business.

Why implement FreeBalance Government Resource Planning (GRP) solutions instead?

1. GRP software is designed for the government. It is possible to create software for a single “vertical market” that does not require code customization to support needs in most countries across all levels of government.
2. Governments can configure GRP software to meet unique requirements thereby reducing lifecycle costs and more likely meeting implementation schedules and optimizing benefits.
3. Software designed for government is easy to use in the government context.

Doug Hadden, VP Products

I’ve been reporting here on some major ERP failures in the public sector. I mentioned in the last entry about cost overruns in the US Department of Defense that this lack of success by major ERP vendors is an “open secret”. We don’t consider the FreeBalance Accountability Suite to be “Enterprise Resource Planning” because we don’t provide software to “enterprises” – the private sector nor do we provide software to more than one vertical market. (By definition, we sell to 3 sub-verticals: national government, sub-national government, international projects.)

My sense is that only a small portion of ERP failures leak out to the press. Some because of audits. Some because of lawsuits. Over the course of the past few years, I’ve heard some very interesting comments that I’d like to share.

• Salesperson from major ERP vendor admitted to 50% success rate in government
• IT professional in central government agreed that ERP is too complicated with a high cost even though the government was pushing to implement more ERP
• ERP software enabled “overspending the vote” – exceeding the budget
• Major ERP vendor in use by at least 2 national governments for accounting – but not budget controls (which is why one buys financial software for government)
• Country implements major ERP vendor product, first phase, in 2.5 years. (Neighboring county implements FreeBalance, first phase, in 1 month.)
• Senior member of government audit organization admits that the ERP implementation has sapped the government
• Former consultant admits to problems trying to implement major ERP vendor product at sub-national level in G7 country as overwhelming
• Another sub-national government in G7 country spending millions of dollars every year for consultant to maintain system
• Developing nation government has been attempting to implement major ERP vendor product – for 13 years
• National ERP shared service for G7 country now more than 6 years in the making with 20-35% cost overruns and more than 1 year late
• Presentation in conference by government attempting to implement ERP in government begins with “normally you hear what went right at these conferences…”
• At another conference, government presenter speaking of an on-going ERP in government implementation had nothing to say when it was suggested that the project was a failure. [This conference was sponsored by the very same vendor. I was asked during a full session what my thoughts were. I was as diplomatic as I could be. Later, one of the vendor representatives commented to me that there was enough "room" for many vendors. I suppose some governments can afford the luxury of cost overruns and high cost of ownership.]

Doug Hadden, VP Products

The Inspector General for the United States Department of Defense published “Enterprise Resource Planning Systems, Schedule Delays and Reengineering Weaknesses Increase Risk to DoD’s Auditability Goals” on July 13th. This comes on the heals of the Government Accountability Office “DOD Financial Management: Reported Status of Department of Defense’s Enterprise Resource Planning Systems” from March 30th. (And, the UK National Audit Office report and the Panorama survey that found ERP software project overruns ‘distressingly common‘.) The Inspector General report provides more financial details to five of the projects listed by the GAO and adds information for an additional two projects. Six projects were addressed by the GAO but not the Inspector General.

This new information requires an update from a previous blog posting.

Cost vs. Budget Statistics

• The 13 projects have acronyms of: AF-IPPS, DAI, DEAMS, EBS-EC, EBS-Procurement, ECSS, GCSS-Army, GCSS-MC, GFEBS, IPPS-A, IPPS-Navy, LMP, Navy ERP
• 4 of 13 ERP projects are expected to cost between 100% and 120% of original budget
• 4 of 13 ERP projects are expected to cost more than twice the original budget
• All 13 projects are using leading COTS ERP software from the 2 leading manufacturers including the leading HR package acquired by one of the leading ERP manufacturers
• The 3 HR projects are using the same or similar configuration which may explain why the projects are roughly on-budget.
• The average lifecycle estimated cost for the projects was $1.355B and the current estimated lifecycle costs (because most projects have not yet been completed) is $2.316B. That’s an average of almost $1B per project – although it should be fair to note that two projects
• According to the Inspector General: “DoD experienced life-cycle cost increases of $8.0 billion and schedule delays ranging from 1.5 to 12.5 years during system development and implementation.”

Inspector General Conclusions

The inspector general recommended “the DoD DCMO and CMOs of the Army, Navy, and Air Force, develop procedures to review business processes and limit funding to programs that are not demonstrating adequate business process reengineering.” The report provides good evidence in the lack of articulating existing and future business processes and having these approved.

Blaming the Victim

There seems to be a general assumption in the market that there is nothing wrong with implementing Enterprise Resource Planning software designed for the private sector in the public sector. Project failure is chalked up to project management failure.

We really need to consider the real value and the real Total Cost of Ownership of using ERP in government. And, we need to stop blaming the victim for project failure. Of course, my viewpoint is biased as a provider of government-specific Government Resource Planning (GRP) software. Here are my takeaways:

• ERP manufacturers, in projects in the 100s of Millions of Dollars ought to be part of the governance structure and actively adapt software to meet critical government needs. Systems integrators and government IT specialists should not be on the hook for significant customization.
• So-called “best practices” for government are clearly not implemented in these ERP packages. If these practices were embedded in the software, (not to mention the supply chain and logistics “best practices” from the private sector), then “business process reengineering” would have been an exercise of tweaking existing templates. For all the marketing spin on “vertical markets”, it appears that ERP manufacturers have little interest in truly understanding the needs of the Department of Defense.
• Transformational projects that attempt to change business processes and integrate multiple systems ought to be led by domain and change management experts rather than IT – that’s a big lesson learned in our public financial management practice