It was said in the article, but I will reiterate: enterprise software companies are mostly made up of salesmen-not software engineers. The majority of their profit margin is from after-sale products such as warranties and consulting. These products have very little to do with the quality of the software product itself. In fact, it might even pay for the software to be buggy since a) more bugs = more consulting = more renewed warranties b) multi-year software contracts are not unheard of and c) ERP switches are extremely costly so companies stick with them for years, if not decades.

As a programmer, I wish that a quality ERP product could stand on its own, but sales and (company) size are key if you want to compete with SAP/ORACLE/MS. At the very least companies will ask for previous customer references, run a D&B on your company, and ask for a proof of concept install. This is a space so ripe with opportunity it hurts but for a small startup to succeed, it will need to pick off pieces of the giants before competing with the entire product.

[1] Yes, ERP software and implementation is that big of a deal. There are old stories floating around of botched ERP implementations costing companies millions of dollars and, eventually, sinking them due to the loss of business (even after reverting to pen/paper accounting).

There is tremendous scope for start-ups here and the market can be penetrated. One needs to think out of the box and study the problem. Most organizations see the need for ERP, when their business expands or when it is running through problems. By that time, their processes are not well defined and that 'confusion' in turn is passed in a sort of convoluted way into the software. At this stage most of them are in need of a good Business Consultant rather than software!

Most Businesses are still run in a 'fuzzy' way. ERP systems impose rigidity and inflexibility. The new generation software would free business from these constraints.

- They would be mostly web based. They will be made of 'websites'that can talk to each other. No need for EDI systems. RSS feeds for materials anyone?

- They would be evolutionary in nature as their proliferation will attract further development and cross-fertilization of good features.

- They would cost 1/100th of what existing systems sell for.

I think there is a great business opportunity here.

Please put your email address in your profile so that we can discuss that opportunity.

An ERP, in order to be effective, takes a lot of buy-in from all your business units, and getting that buy-in will cost the IT people some serious office political capital if the install fails. So what does the average person in a large corporation go with? The big guys who have a proven record.

It's similar to the aphorism "no one's ever been fired for picking [IBM, Microsoft, FedEx]" - sure, there's a risk that the thing might fail, but when you go with the largest company around who is almost exclusively focused on this, that risk should go down (I'm not saying that's true - only that that's the working thought process), and even if it does fail, you can fall back on saying "well, it's the industry standard, so I didn't screw up by picking it."

Start-up ERP will not go straight into large companies, and it probably shouldn't. Start with small local companies, and slowly work up in size - and then pick a size and stop there. And look at FreshBooks: they're a Toronto start-up that's got some decent penetration into tracking and invoicing sector, which is traditionally the domain of much larger companies. They do that for a while, get small-to-medium sized businesses comfortable with them, and then in two years maybe release a module that does payroll and integrates into the existing one? A year after, once their customers trust both of those, release an HR one? There's your start-up ERP right there...

The company has just paid a lot of money for the software, which was chosen through an evaluation process where the vendors showed that they could support the existing business processes, which made up the RFP. They chose one vendor and tied them up in a contract to say that the software would meet the RFP.

Then the project starts, and the company notices that the software doesn't really meet the RFP, so they get the vendor/consultancy to customize it. Otherwise they wouldn't be getting their contract's worth, you know. The earlier claims of not changing the software get thrown out the window. Plus you have the business people who have been seconded to the project, they don't want to tell their colleagues they are delivering a clunkier system than the old one so they fight for the customizations. Fair enough.

IMO, any software that is absolutely critical to your company's special way of doing business (i.e. your competitive advantage) you shouldn't customize an ERP product to do it - you should build it yourself. Anything that is generic and not a point of difference between you and your competitors, you should keep as standard as possible. The bit where they join up, make sure you abstract it properly.

As soon as you create barriers for upgrading your ERP, you lose the value of the maintenance (which is the ERP vendor's developers working hard to add new platform and application features, instead of your guys).

I guess this shows there is not one model to rule them all.

An item for consideration is whether part of this customization disaster is driven by a misguided belief in software reuse.

Enterprise Resource Planning

ERP is best described as a massive unified software product that runs a huge portion of a business's operations, usually including two or more of the following areas:

  - HR
  - Accounting (Receivables/Payables)
  - CRM
  - Business Intelligence / Reporting
  - Business Process Management (e.g. form processing/approvals 
    across front- and back-office functions)

All that crap about HR and payroll and so on was bolted on afterwards.

The original acronym was MRP, Material Requirements Planning, a perfect candidate for business software. It answered the question, "If I need to deliver 9 helicopters on these 9 dates, then what components will I need on which dates?" Believe it or not, this was all hand calculated at one time.

MRP was very complex and difficult to implement because it required absolute precision and discipline, rare back then and still rare today. If your base data (inventory balances, lead times, quantities per, etc.) were the least bit off, the resulting automated explosions would be way off. So an industry of software vendors and consultants was born to attack all of these issues.

The problem with MRP was that it didn't work well at all for products with few components but complex processes, (think chemicals, energy, distilleries, food processors, etc.) So CRP, Capacity Requirements Planning was born to plan and manage factories with high capital expenditure requirements. (It doesn't matter if we have exactly the components we need if we have nowhere to work on them.)

Before you know it, everyone wanted in on the act of expensive software and consulting, even in disciplines that didn't require them (why should SAP make all the profits). So along came accounting, sales, HR, and everyone else, and now we're stuck with ERP, a cow that's ripe to be milked for a long time.

Which then they have customized to integrate with their ERP. :-P

1. No "next page" link. You have to determine which page you're on and then click the next one, a process made difficult by

2. A barely-apparent current-page indicator. Two offenses you might be able to forgive if it weren't for the

3. Lack of a printer-friendly/all-on-one-page option.

Why is pagination still so hard?

I inadvertently started a second debate by posting the content of the article here. I felt I had little choice. I'd rather skip the whole thing than send someone to that site. But I wanted to discuss the content, which was excellent.

If it wasn't for that excellent content, I would have never struggled with this horrendous site. It was as if they are daring you to finish the article. Some of the other horrors:

  - embedded video ads (incredibly distracting)
  - way too many words of the article were underlined links
  - many other underlined links that were *not* content and hard to differentiate
  - tool tips everywhere, that were a huge problem because
  - you need to mouse so much just to scroll the content
  - less than 10% of screen real estate used for content, which
  - brings up a new acronym, WTFA *Where's* the fucking article?

Yes the site has a lot of issues. It's a terribly design - long and skinny, so you have to scroll to read it.

But the article is good and by ripping off the content and posting here, you impact the traffic to the site and the quality of the site too. The less traffic, the more the publisher has to rely on advertising, the less chance of their being enough budget for a redesign. And, the more likely fall guy will be the writer that could face losing their job as a result of cutbacks.

I am generally a fan of YCombinator and the community so I was surprised and disappointed to see this flagrant disregard for IP.

If you don't present your ideas well, they will be missed. If your content is wrapped up in junk, your audience will move to another content provider.

This is the law of the land. Any tiptoeing just delays progress.

As my friend's grandpa used to say: "shit + ice cream = shit"

With the cost of software today, and the advanced platforms out there, agile methods, ERP systems, or not even an ERP system, but rather, the software a company needs to run the business can be built from scratch in less time and for less money than it costs to implement an ERP package from one of the big names.

It explains the problem about as well as any "ERP sucks" article I've read in years. It doesn't suggest any solutions, leaving it as one of the the most fertile opportunities for a B2B startup.

The website doesn't have a "Single Page" or "Print" option, so I'll spare the rest of you the torture I endured and reprint the whole thing right here. Definitely worth the read:

Why ERP is still so hard

ERP failures are commonplace, IT managers still struggle with upgrades and projects only have a seven percent chance of being completed on time. Why is ERP pain by numbers?

By Thomas Wailgum

Steve Berg knows what intense pain feels like: The man has been Tasered, in fact-not because he ran afoul of the law, but as VP of IT at Taser International he's partaken in a corporate rite of passage. "It's the worst five seconds of your life," he says. "You cannot move."

Like other IT execs, he also knows pain and suffering as it relates to ERP - from vendor selection and licensing negotiations, to implementation and change management, followed by upgrades and integration.

And as he and many other IT leaders have come to know, ERP-induced pain can last much, much longer than a mere five seconds.

Taser's attempt to wrap an ERP package around its corporate processes sounds eerily similar to most companies' experiences. The "before" picture: A mélange of disparate systems that didn't talk to each other and a good deal of "paper pushing" between the systems, Berg says.

"When you don't have a centrally managed technology environment," he says, "things can get overly complicated very quickly."

Executives had sought a unified system so that Taser "could do a complete workflow throughout the company without having to run redundant systems that don't communicate," he says. That was 2004.

Microsoft's Dynamics AX was eventually selected.

And again, like many companies, Taser decided to customise its chosen ERP package to meet the business processes that it already followed. "So rather than take an ERP system-which supposedly out-of-the-box has, say, an accounts receivable [process], with best practices that are inherent to the system, we decided...to modify AX to work like this other application because users were comfortable with it," he says, "and they didn't want to change."

But a funny thing happened on the next upgrade: Naturally, all of those customisations done to the initial AX rollout-which were "plentiful," Berg says-were going to have to be upgraded in 2009. Taser decided it didn't want to go down that road again. This time, Taser ERP users would change, demonstrating that vendor-purported "flexibility" has been both ERP's blessing and its curse.

"We're going to get rid of these customisations and go back to what the [Microsoft] AX best practices and recommendations out of the box," Berg says. "If we're going to be able to grow the company-we're at $100 million now and if we want to be a half-billion company in four years' time-the current processes are not allowing us to get to that point."

The upgrade took longer than expected: Testing and training issues, as well as certain customisations that were unavoidable, complicated progress along the way, Berg reports.

Executive sponsorship and interest never wavered, though. "It seemed like all eyes were on this upgrade and all eyes were on IT to make sure nothing could go wrong," he says. "Everybody understood the long-term benefits, but there will always be some teething pain at the beginning. We went live in May [2009] and now we're in July, and things are running smoothly. But May and June were pretty tough."

Taser's tale probably seems commonplace to IT vets. But the fact that Taser's story is so common, so expected, so universal, after nearly 40 years of all things ERP, makes it all the more significant.

ERP Pain, By the Numbers

The preponderance of corporate pain lurking throughout the lifespan of an ERP suite is unequivocal. To wit: ERP projects have only a 7 percent chance of coming in on time, most certainly will cost more than estimated, and very likely will deliver very unsatisfying results.

In addition, today's enterprise has a little better than a 50 percent chance that users will want to and actually use the application. Poor application design just adds to the turmoil. In sum, "ERP success" has become a very subjective metric.

As for costs, an ERP system from a Tier I provider isn't cheap: Total cost for an average SAP install runs nearly $17 million, Oracle at $12.6 million and Microsoft is relative bargain at $2.6 million. (Tier II ERP providers average in at $3.5 million.)

For all of that investment, today's enterprises surely must be basking in the glow of their fully modernised ERP backbones? Hardly. A Forrester Research survey of more than 2,200 IT executives and technology decision-makers in North America and Europe found that modernising key legacy applications is the top software initiative for businesses in 2009.

Making matters worse is that CEOs and CFOs are still trying to wrap their heads around the financial aspects of ERP, a most unusual piece of the corporate pie: the licensing, implementation, customisation, annual maintenance and upgrade costs. (More on that later.)

Not surprisingly, ERP has consistently remained among the top IT spending priorities in large corporations, growing at the rate of 6.9 percent each year and set to top the $50 billion mark globally in 2012, according to Forrester Research data.

Summing it up in an understated yet perfect way is Ray Wang, a former Forrester analyst and now a partner for enterprise strategy at Altimeter Group: "Business software is just not easy."

But, as far too many people have lamented over the years: Why does it still have to be?

Dawn of New Computing Era

To understand where we are today, it's critical to recall ERP's ascendency. In brief: Systems Applications and Products in Data Processing (SAP) forged the market in 1972. Businesses and their leaders in the 1980s and '90s bought into SAP's vision of a computerised mechanism to connect finance, operations, supply chain, HR and sales information.

"You were going to be able to be more efficient, effective and also lower your overall costs-that's pretty much what everybody was aiming for," says Manjit Singh, CIO of Chiquita Brands International. The market boomed. Other software vendors joined the mix. Oracle came along and bought up many ERP players, though SAP remained king.

Expectations for IT's omnipotence soared in the mid- to late 1990s. Credit flowed like the Mighty Mississippi, and companies thought nothing of spending millions on ERP installs-some of which were integrated, most of which were siloed.

As Y2K approached, fears of worldwide catastrophe created a panicked IT group desperate to replace all non-compliant systems. Companies were at the mercy of their ERP vendors and consultants-and both of those parties made a killing.

"There was a mass rush to implement these things, and therefore consultants were expensive and software wasn't discounted that much," recalls Vinnie Mirchandani, a former Gartner analyst and founder of Deal Architect, which consults with companies purchasing software. Companies bought suites of ERP apps-lock, stock and barrel. "This was pretty bad," Mirchandani says. "It was an almost irrational buying pattern."

The turn of the millennium ultimately proved two things: 1. Y2K was a non-event because IT did its job; and 2. Companies were now locked in to their ERP providers-not with a competitive advantage but with a competitive similarity-for the foreseeable future.

But now, businesses change at a pace at which ERP systems have trouble. "If business was still the way it was in 19-whatever, yeah, [ERP] wouldn't be a big deal," observes Wang. A CFO Research Services study of 157 senior finance executives succinctly addresses the situation:

"Companies grow and change, acquiring new business lines and divesting themselves of others. They open new facilities or consolidate operations, add partners or outsource functions, centralise or decentralise the back office. Reporting requirements increase as regulatory bodies heighten oversight and as companies expand across borders.... In short, businesses change, and as they do, so do management's information needs."

Of course, ERP applications can change. But it'll cost you. In customizations. In change and process management. In upgrades. A typical company, notes the CFO study, will spend an average of $1.2 million each year to maintain, modify and update its ERP system. ERP Costs Still Tough to Understand

The fine print and financial legalese contained within ERP application contracts can be alternately mind-numbing and head-scratching for the uninitiated.

"Of all the assets that an enterprise acquires, enterprise software brings with it the most unusual, onerous and restrictive set of constraints," writes Wang in a June 2009 Forrester report on software licensing. "In most cases, licensees may not resell, reuse or share their license. Licensees often encounter numerous grievances across the software ownership life cycle from selection to implementation, utilization, maintenance and retirement."

Oracle, for instance, will heavily discount license pricing upfront but will, rest assured, make that up on the backend - from its 22 percent maintenance and support fees, on which it does not negotiate.

Oracle President Safra Catz told analysts on a conference call that maintenance is "very profitable part of our business, and as the number gets bigger and bigger it's really impossible for us to actually spend our way through it, and so in general that's the sort of overriding thing that guides our margins."

Closing its most recent fiscal year, Oracle achieved nearly 90 percent profit margins.

Chiquita Brands' Singh understands and explains ERP this way when it comes to $1 million-plus purchases: "Your management team needs to understand that $1 million is not really $1 million. There are significantly higher costs as you look at the average lifespan of the purchase as well as resource implications," he says.

"The CFO and CEO need to know that because they're going to see IT costs go up as a result. And you don't want them constantly asking the question: Why are year-over-year costs going up?"

That's all assuming ERP implementations are reasonably successful. Not surprisingly, with all the risks and all the multimillion-dollar projects, ERP implementations, when they do fail, can be spectacular events.

ERP, it seems, in one technology area in which a dose of Moore's Law does not apply. "It's sickening how ERP continues to be very expensive and very risky," says Mirchandani.

"There is no reason why it should be. The software should be heavily discounted to start with. The maintenance [plans] should have several different options, offered both by the vendor and third parties. And the implementations should be more brain-dead implementations."

ERP and CIOs: Complicit or Complacent?

The Chief Information Officer's slow rise to prominence inside enterprises is undeniably intertwined with ERP's climb. A significant part of tech leaders' career trajectory - from data processing to MIS to IS to IT - is the manifestation of ERP's impact on the wholesale digitisation of the innerworkings and processes of businesses, governments and nonprofit organisations.

In other words, there would likely have been no Chief of IT in the 1990s if there was no ERP - just an IS manager responsible for e-mail.

In that vein, some ERP analysts lay partial blame for ERP's spiraling costs at the feet of CIOs, who have aided and abetted vendors' addiction to maintenance fees, for instance. "I don't blame the vendors: They're doing what the market is telling them to do," says Wang.

"And they're doing what inherently the customers are telling them to do, and it's got them to this point. The problem is that just about everyone's been kicking the can a little farther, and I don't think we can kick it anything farther than now."

Maintenance and support fees, in particular, have drawn the ire of businesses scrutinizing balance sheets and trudging to make it through the 2008 recession. SAP's redesigned Enterprise Support plan would have increased its maintenance fees, but it wisely negotiated a détente with its global user groups who were up in arms over the proposed fee increases.

CIOs, too, can play a starring role in limiting costly customizations, by educating and imploring business managers and users why customisation, in the long run, is often not the better route. But that task is never easy. Singh contends that every company thinks its processes and products are so different that customization is absolutely necessary. "While in reality, there is no good, solid rationale behind that in the vast majority of cases," he says.

But the reality CIOs face when synching business processes with those in ERP applications leads to "internal arguments over how we are going to define something simple as a chart of accounts," Singh says. (A chart of accounts provides an overall view of items such as assets, liabilities and expenses.)

Each business organisation and unit will have a different view. "So all of the sudden, what looked like a very simple concept has exploded in complexity," he says, "and now you're into trying to get some very powerful people aligned behind one vision. In some cases, you can; in some, you can't."

All of this can add up to thousands of contentious processes when a company is implementing an ERP suite. "That's something [business managers] discount, and that's something the vendors don't talk about," says Singh, who says the CIO job can be more like a Chief Negotiator role.

"Vendors will say: "We have an out-of-the-box solution.' And they do. As long as you're willing to take what they're selling, it'll work. But as you try to deploy it, business leaders will say: 'We can't do it that way,' or 'That won't work for me.' And that out-of-the-box solution suddenly becomes heavily customised."

The era of mass customisation has had a contradictory effect on how packaged software was supposed to positively alter business software. According to Wang, "Packaged software was supposed to be: Let's all get together, we'll share requirements and what's going to happen is that we're going to have this wonderful mix of software best practices from all these different areas and companies."

In turn, businesses would have to spend far less on keeping IT workers in-house to maintain the applications; and thoroughly enriched vendors would use their R&D capabilities to deliver the best of the best. In this utopia, customizations would negligible.

"It didn't happen that way," Wang says. "And that's why we're where we are."

Does ERP Still Matter?

Yes. ERP still matters. Very much so. CIO magazine survey data published in 2008 showed that IT leaders and their companies were completely married to and dependent upon their ERP suites.

More than 85 percent of respondents agreed or strongly agreed that their ERP systems were essential to the core of their businesses, and that they "could not live without them." Interestingly, just 4 percent of IT leaders said their ERP system offered their companies competitive differentiation or advantage.

As the CIO profession has grown up, so too has the CIO's ability to manage all that comes along with ERP. Jeffrey Keisling, CIO of pharmaceutical-maker Wyeth, is presiding over a massive, multiyear ERP makeover, moving from multiple, global instances of J.D. Edwards to a single-instance SAP ERP suite.

"There's always a lot written about examples of drastic overruns or miscues or re-dos around ERP," Keisling says. "But generally, the people I know in large enterprises have been much more effective in understanding the levers to pull to mitigate risk in these large programs."

Most notable among those levers: the ERP implementation has to be a priority for the company - from CEO to users, Keisling says. "If this is something one person is trying to push up the hill or one division or function is trying to push up, we would reject that," he says. "Without that level of understanding, sponsorship and expectation of value, I wouldn't take the bait."

ERP has its place, too. "The measurement stick for me and my team is not how well we did SAP," he says. "The measurement is: Did we improve the company's performance or our ability to get products to patients? [Our SAP rollout] doesn't dwarf the need for innovation for new products and for working with patients to get those products in their hands."

On the vendor side, the rise of the enterprise software supervendor (dubbed "MISO" by many) has been an unyielding force: Microsoft, IBM, SAP and Oracle have and continue to centralize their positions as software juggernauts.

"The traditional boundaries between integrated ERP and best-of-breed vendors have disappeared," notes a recent Forrester report.

"Over the past few years, leading vendors have significantly extended their portfolio via acquisitions and in-house developments to offer both: integrated packages for core enterprise processes and best-in-class horizontal solutions for procurement and sourcing, supply chain management, CRM, and other cross-industry application software," BI being most notable.

According to Mirchandani, consolidation might actually backfire for the supervendors: Whereas companies might have dozens and dozens of different IT spends with various vendors (which are easier to overlook, since there is not a combined view of the total dollars), a unified Oracle spend could be detrimental to Oracle's revenue streams.

"If anything, Oracle, by consolidating this much, has made itself a target, because now you can have Oracle-wide strategy," he says. "Previously, companies didn't worry too much about their PeopleSoft, JDE, Hyperion [spends, because they were separate]. Now they've got a big bull's eye on their back."

Even so, Mirchandani says, vendors probably aren't losing sleep over their customers walking away any time soon. "The vendors are counting on inertia," he says.

So, are CIOs and their companies really, for lack of a better word, stuck with ERP? AMR Research Chief Research Officer Bruce Richardson was quoted as saying, "You do ERP once, concrete it over and hope you never have to dig it up." Without question, ERP has been a career-enhancing or career-limiting endeavor for many CIOs. Chiquita Brands' Singh terms the ERP vendor selection process as choosing "the lesser or two evils."

Altimeter Group's Wang believes CIOs have figured out the rules of "the ERP game." But the business side is still confused.

"The business sees the slick demos and possibilities, and then keeps forking over the money for this, and they don't understand why they are still paying all this money," Wang says. "Why is it so hard to get a simple report? Why is it so hard to add a new product or build a new product line? Why is it so hard to get consolidated financial information? Isn't that the whole point of ERP?"

http://www.copyright.gov/fls/fl102.html

I think that a full reprint here is clearly against the spirit of #3 and #4 -- this is the entire piece and it stands alone (#3), and by reprinting it here the market (advertising revenue on the original site) is affected (#4).

Edit: Since this is a UK publication, this is probably the fair use doctrine that applies (similar in principle to fair use in the US): http://www.copyrightservice.co.uk/copyright/p09_fair_use

In addition, today's enterprise has a little better than a 50 percent chance that users will want to and actually use the application. Poor application design just adds to the turmoil. In sum, "ERP success" has become a very subjective metric.

There must be some sort of perverse incentive working here. 7 percent?! How much will it cots? How much do I need to quote to close this deal? How long will it take? How long do I need to quote to close this deal?

ERP has always interested me. Vendor/customer relationships are often complex however the software side shouldn't really be the case - if anything, software should be the easy part.

Anyone starting up in the space obviously can't (successfully) build an integrated solution. You would have to develop something better than current SAP/Oracle/SAAS solutions for one or two of the operating units described by joshwa.