Will SaaS kill ERP? No, but it should

It's been a busy few days. First a post on ZDNet by Eric Lai invented a few problems for Cloud, or rather SaaS, and especially multi-tenancy: inflexible, less secure, less powerfull and maybe more costly - is what Eric claims multi-tenancy SaaS to be.
Thomas Wailgum neatly nailed that via a counterpost, as did Frank Scavo, to whose post Eric commented, and back again

The circle of Inspiration? Yes, it evolves hardest and finest on Twitter. And here are my thoughts: we need to grow up

First, now I've got your attention anyway, a minor disappointment maybe: ERP in this title and post doesn't stand for SAP, but for the epitome of enterprise IT as we know it

Good and bad arguments were made in all posts and their comments, but there was a lot of comfortzone involved - and I'm challenging that very comfortzone: enterprise IT is a multi-billion dollar industry where customers get very, very little value for money
Surely it employs many people and that huge discrepancy between investment and return gave a great boost to India and other low-labour-cost countries, but let's face it: this industry needs to mature

1. The foundation of IT

In the beginning, there was a bit.
Then, it was assembled into bytes.
And bytes formed a double-word - and then the story really started

A poetic attempt at sketching the genesis of computers, but I really want to stress the fact that IT all started with hand-crafting something out of nothing.
Over the past decades, however, standardisation and tooling slowly took away the burden of doing everything yourself. Compared to writing, you could buy ink and goose feathers in stead of making your own. You could go out and buy paper, in stead of making your own parchment

In IT, various "straight jackets" saw the light: hardware first, then software, and in between the wiring: network protocols. Choices became economically limited to a few dozen, and those became half a dozen. Standardisation always starts at the infrastructural level, at the very bottom of the IT foodchain. Compared to housing, the foundation is standardised exceptionally well across the globe as the choices are simply very limited when you have a cost-efficient solution in mind

Right now, in Enterpise IT, we have a few choices in hardware. Half a dozen choices in operating systems. Over a dozen choices in programming languages. A few dozen choices in program language interfaces, and over a dozen dozen (sic) choices in applications: the closer you get to humans, the more diverse the choices you have

2. The IT exception and rules

The odd one out there is ERP. There's SAP and there's SAP. And then there's Oracle. Wait a minute - we're 1-on-1 customer-facing and we only have 1.5 choices? That can't be right...
It isn't, in my point of view. The diversity fans out bottom-up.
One more time, with feeling:

• We're settling into the infrastructural machine layer all the things that we take for granted: the simple, static stuff that doesn't change, is highly structured, rigid even. The data, the smallest building blocks, bits and bytes. Order reigns here, where everything is subjected to business rules, and automation can thrive. It's boring and it should be: it's the foundation to our "homes"
• Facing the very end, humans, the opposite takes place: we encounter complex, dynamic stuff that changes all the time, unstructured, flexible itself and requiring the greatest flexibility at the same time. It's where knowledge and information flows freely, uncapturable. Chaos thrives here, exceptions are the rule, and automation usually is impossible. It's where Social sees the light

3. How the one-size-fits-all oxymoron survived

So how come ERP is where it's at? Well, only because it's customised to death. I've been told that you can buy ERP for amount X, and will need to spend an additional amount of 7 X's to make it work for you. That is caused by three phenomena:

• An enterprise-wide system is high up in the IT stack, yet if you want to ship & sell it off the shelf it needs to be as "flat" as possible: offer as little rules as it can (and that's rather easy to achieve, and achieved). There are no global rules, nor exceptions
• All kinds of industry-specific, national, regional, local, partner-specific, customer-specific and department-specific rules and exceptions, as well as regulatory compliance issues, are impossible to keep up with by one global system. Technically, as well as economically.
  Here's where Push is replaced by Pull (John Hagel and John Seely Brown can enlighten you on this one). There are many local rules, and exceptions
• Vendors, but particularly System Integrators, love to provide you with people on a Time & Material basis to achieve your goals, meaning: building in all the business rules and exceptions that you need. This might take many months, even years. The long payback time on ERP investment usually means any cost here gets sanctioned

Invest a few million in getting ERP, then being confronted with having to invest yet another few times that in actually getting what you want, slam onto that the annual "support" fee of 22%, and you'll be glad to pay off that initial investment before you die - it's an obscene lock-in. That initial 3-year horizon and payback time will slowly turn into a 20-year one, and keep you from further growth even.
A proper ERP implementation will lock you in for 15-20 years

4. The blame game and the solution

Is it the fault of ERP? Of SAP? Of Oracle? No. We still adhere to our Genesis: hand-craft from scratch.
ERP mainly is a, albeit rather huge, manifestation of our desire for one-size-fits-all customisation - which is an oxymoron by default

Look at housing: the fundament is standardised, as are the bricks, but everything else is up for grabs
Look at transportation: the roads are standardised, but you can pick a variety of vehicles to travel upon them
Look at the sky: the airports are fixed, but the variety of carriers is offering you all the choices you'd like
And then, look at IT: one choice of ERP is all it offers. Well, 1.5 maybe...

The message is clear: standardisation at the bottom allows for customisation on top of that. Customising the standardised parts itself? Ridiculous

Software as a Service comes to the rescue here: it will offer standardised solutions. Multi-tenancy, of course. They will offer you vanilla out-of-the-box and not much else. It'll be old-fashioned ERP as it came on the CD, yet pay-per-use
Where will the required standardisation come from? The need to comply with regulatory laws?

5. Turn the tide

From another SaaS, of course - or you can build it yourself, or maybe someone will offer it on-premise - or all of that combined. Do you need a business case for that? You could shop together your desired functionality online, just like a regular consumer. I hope Marc Benioff is paying attention...

SaaS will make this industry mature. It will straight-jacket enterprise IT buyers, and finally put an end to the ridiculous customisation on top of multi-million dollar ERP packages. Much like the car industry, it will force you to choose a brand, make and type, and allow you a few accesories - and that's it. If you need anything else, go get it and plug it into your Enterprise Integration Hub

SaaS your tertiairy business processes, or get them off-the-shelf.
SaaS your secondary business processes, or find a local shelf to pull them off from.
SaaS the boring core of your primary business processes, and create the specials yourself just like currently ERP makes you do so. But now, you can choose to buy local, Private SaaS them, or make - or any combination of that

That will leave you with a few disparate solutions to integrate - but SAP nor Oracle have solved that problem for you, now have they? At best, ERP's lack of strategy and consistency regarding Integration has cost you dearly.
And you know my answer and solution to that last question, of course

IT needs to mature, and become an adult industry. Pun intended. That's a multi-billion dollar business as well, yet I hear a lot less customer complaints from them