No less than a third of inquiries from potential customers are requests to make an implantation over some existing software that they are already using. That is, they read a case and were in ecstasy over additional
functionality described.
Here’s a typical situation: “We already have a working ‘system’, it does our records and sales. What we need is logistics management like the one you have, and a ‘CRM’, and a call center, and online sales, please come and set them up!"
So, brothers and sisters, let’s wipe those rose-colored glasses of yours with sand paper.
Both now, and ever, and unto ages of ages.
I. A sin of modularization
To begin with, an intention to aggregate two completely different systems into one means taking the idea of a modular system and making an apotheosis of failure out of it.
We have already discussed modular systems, so let’s not repeat ourselves.
In your virtual case, my fair readers, the situation is even worse. If, let’s say, SAP at least invests a lot of effort into making initially heterogeneous ‘modules’ fit together, plus having a complete control over the situation, we, on the other hand, are requested to literally breed the unbreedable and obtain a ‘first class’ result in the end.
A dubious idea, even without jumping into comparing advantages and disadvantages of breeding.
But we’ll say more on this later.
II. The present tun of poison
To make reading simpler, we’ll once again repeat the mumbling of the potential customer: “We already have a working ‘system’, it does our records and sales. What we need is logistics management like the one you have, and a ‘CRM’, and a call center, and online sales, please come and set them up!"
Now, here’s a big question: if you already have a working ‘system’, then why wouldn’t you rather implement the functionality you need within that system? Things that you were so impressed by, while reading our cases, were described in a lot of detail, so there shouldn’t be any problems duplicating their business-logic.
And here comes the ‘Oops’, as Ms. Britney Jean Spears, a multi-platinum and world-famous American singer-songwriter, suggests us and our fair readers.
Next (invariably!) we learn, that the working ‘system’ is either one of the following:
- A 1C system, which for many years have been altered in-house and has come to the state, when any further work on it becomes simply scary – thank God, it’s working somehow, don’t touch it, or something will go wrong again. A large diversified business may have several different versions of 1C, whose users interact with one another in a completely unpredictable way.
- An in-house development, where full-time programmers have a one-and-a-half-year task queue.
- A systems’ zoo – characteristic of wealthy companies with enthusiastic managers, when a fancy CRM or hybris, etc., have been already installed over a leastways working 1C or an ancient in-house development. Millions of dollars wasted – nothing works. Except for the 1C system, but even its work is unsatisfactory, otherwise no one would be bothering.
- Otherwise – a hell of a mixture of everything listed above, "shaken, not stirred" ©
Guys, is THIS what you suggest integrating our solution with? Along with guarantying an outstanding final result?
III. An experiment of the century: integrating a tun of wine with a tun of poison
An efficient business runs smoothly as a single mechanism. Or organism.
It has different parts (organs), but all of them are interconnected and synchronized. A separate organ is not viable outside of the organism, while the mechanism is not functional without any of its parts.
Your (let’s say) trading company would not be able to operate, if we take out the warehouse – facilities, products and personnel. Likewise, this cutout warehouse won’t survive on its own as a sustainable business.
An information system automating your company’s activities will be only then fit its purpose, when it accurately reflects real business processes of your enterprise.
Anything else is rubbish.
Your call center is completely useless, if phones are disconnected and operators have no access to information on stock and prices. Same with a theoretical standalone CRM system – it’s useless.
As well as a logistic management subsystem.
Thus, however great is the Ultima functionality, which you’re interested in and which can’t be implemented in your ‘system’, this functionality is completely useless without a 100% integration with the latter.
Anything else is bullshit from unscrupulous salespeople.
“So integrate!" you might tell us.
We’ll answer with an illustration. Suppose, you’re building a house. You hire a team; they build a foundation and walls. Well, actually, the make quite a mess of it: the foundation is unsafe, the walls are curved and lopsided and could fall any minute. And when it’s time to build the roof, the team just stalls.
Well, you are not to be intimidated so easily: “If those ham-handed goblins can’t do it, we’ll find some others". Yesterday – remember? – you drove past a house with a fancy roof. All you need is to call the guys who did it.
But those guys you call just wouldn’t agree. They even begin to curse: “Are you out of you mind, old fool? You can’t put a roof on those lop-sided walls, and even if you do – it will all fall apart in a couple of weeks.
“We’re not the ones who built this crap, but when the roof falls, we’ll take all the blame! No, that won’t do: tear your damn Dol Guldur down, and then give us a call."
A full-scale deep integration of two different kinds of software is a very complicated task. Almost like transplantation of an organ from one body into another.
It’s a lot more difficult and risky, than implementing the functionality you need within a solution that is already in place.
And if you clearly understand, that even the simple solution won’t work, than what’s point of hitching your wagon to a star?
IV. United we stand
Let’s look at the problem of integrating your working ‘system’ and our solution from the management theory point of view.
Forget complex wordage; simply put: a very long tunnel in the mountains is being drilled from the two sides. In order for the drills to meet, both sides have to perform perfectly (!).
A short remark: all co-ordination is good intentions.
It’s clear the drills won’t meet, while finding out what happened will come down to finger-pointing, some of it more persuasive than other.
This situation is clear to anyone who had anything to do with construction: if there is no prime contractor on site, than there this situation has no solution.
It’s exactly the case in the scheme that you propose: there is no prime contractor, and the most complicated tunnel is being drilled from two sides independently.
If it’s being drilled at all. As for our side – we’ll be drilling, but your staff – why should they even bother? You can’t control them, you can’t check whether their excuses are good or lame. Because if you could, there would be no need in calling us.
In short, the outcome is a bit predictable.
You want us to become the prime contractor? Remember, what the roofers said after looking at the walls and the foundation? The situation with software is even worse: while you can tell that walls are lop-sided right away, it will take several months only to figure out what’s in that multi-megabyte code that causes the mess.
V. How much the cake is worth
In the end: Is the cake worth the candle?
No.
The cost of integration (without even mentioning the 95% probability of failure and an unpredictable duration of the process) in any case will be higher than both an attempt to implement the functionality in question within the working system and a complete migration to the new platform.
Anyone, who has seen a Russian car, customized into something ridiculous, would twirl a finger next to their temple, meaning: “What an idiot! He could have bought a decent car with all that money."
Remember that episode of Top Gear, where they found a guy, who drove a 400-dollar Lada? Folks from Lotus Sport spent two weeks and 200 000 dollars – and made something worth driving out of that piece of scrap metal. Funny. However, from the boring standpoint of a profitable business – for 200 000 dollars they could have bought a Porsche 911 Turbo (and spend change on beer), and get a car with a way better customer performance – from any point of view.
It would save us two weeks’ worth of waiting, along with Lotus Sport professionals’ time, and the old VAZ-2107 car would be left alone. Let the shadow of a Soviet panzerwagen make its neighbors scream at night.
VI. Who’s to blame and what is to be done?
Who’s to blame – a human temptation (so very characteristic of human nature) to make easy, simple and wrong solutions. Priority of short-term results over long-term goals.
What is to be done?
Options:
- Nothing. Be patient. You somehow managed before, you can manage for some more time. At least, you’ll save your money, time and nerves on dubious projects.
- Try and put the working system in order and implement new functionality within it. The prospects of this idea are uncertain – it all depends on specifics.
- Make a deal with a software provider of contentious integrity, who will promise you both a superb functionality and a seamless integration. In this case, as they say, ‘no comment’ and good luck – you’ll need a lot of it.
- Admit the obvious – and finally migrate to a decent platform, gaining all at once: reliability, functionality, speed of implementing modifications, high quality support.
It’s clear, that we’re more interested in the last option.
However, it’s also the one where you profit the most. More precisely, it’s the only option that lines with long-term interests of your business.
It’s clear, that along the way of realizing, what the only right solution would be, you’ll be facing many
obstacles, mostly psychological ones.
Really, so many years and so much money spent in vain?
Well, folks, not completely in vain – you have eventually come to the right decision. And the faster you come to terms with the reality and focus your efforts in the right direction, the sooner you will overcome your psychological issues and earn a lot more money.
P.S. Some readers, who are unfamiliar with Ultima’s commitment to the ‘we make money together with you’ principle, might think that we’re selling an allegedly well-bred puppy to our naïve readers who’ve come for a fishing rod.
Harrow and alas!
We do understand, that only one out of ten (hopefully!) will choose the right path.
A bit less nicety in these matters – and we could have been a lot more wealthy (like many of our famous colleagues) selling imaginative fishing rods.
However, the Force
- of the great principle “we make money together with you", as well as
- of the practice of taking money only from satisfied customers and only after the work is done, puts us off — facilis descencus Averni! —
puts us off making easy money on a pipe dream.
P.P.S. In the meantime, the cash for imaginary fishing rods (90% of cash-flows in the industry) goes to the abovementioned guys, who are not so particular.
Their name is Legion system integrators.
Tabloid press, such as CIO and CNews, is heavily marked with press releases, reporting cases of charlatans’ successes, i.e. of breeding the unbreedable. The seamy side of it – watching those improvised huts collapse – is visible only to insiders.
A pessimist, as we know, is a well-informed optimist.