Let the cobbler stick to his laast

Despite the fact that the axiom, by the definition, needs no proof, there again and again are skeptics, who try to deny it on their own experience and using the money of shareholders. They can’t put up with the fact that the development of management accounting systems, as well as any area of human activity, should be carried out with by qualified professionals only.

I remember the classic "no less than 70% of the people reading these lines have just tried to bite themselves by the elbow."

I wonder why it has not occurred to any company to write its own operating system for office computers, or, say, their own Bluetooth drivers for the CEO’s iPhone? Or, for example, why not to hire a research team and build a lab to develop their own formula of aspirin for the employees who have caught a flu.
It sounds funny. But aren’t the ambitions of some "Stroytorgtsentr" LLC to create an ERP-system for internal use even more ridiculous? A complete ERP system is just a little less complex than any other modern operating system.

Well, you have decided to automate their business on your own. Here and further we assume that your business has outgrown the scope of feeble 1C (which is discussed here).

Perhaps you:

• have good experience in managing large software development projects;
• are perfectly familiar with the current state of the market of ERP-systems, know thoroughly the advantages and disadvantages of the proposed solutions, their functionality and architecture and are able to design a system that would combine advantages and exclude disadvantages of your competitors;
• deeply understand the market trends and directions of the major players for the next few years (for the period of developing and implementing of your own solutions);
• you have a knit-up team of high-skilled programmers, especially qualified in the development of applications for modern DBMSs on modern platforms. Your competencies in production and distribution of tasks, assessment and motivation in particular areas are very high;
• believe that you will be able to ensure the proper quality of production code and documentation while the development — sufficient for the staff programmers who come up as a result of the inevitable turnover to seamlessly integrate into the system of design and / or support, the training of the new staff is well organized.

All "Yes"? — Congratulations.
And strongly recommend to consider the possibility of converting your business into the software development. Or, at least, assign this field to some subsidiary. After the announcement of a tender on automation of core business we would be delighted to participate in it together with your daughter and obviously win it.

No?
Well then, let us make some suggestions (already affirmed by hundreds of solid foreheads that have been there, and have done that):

• time of the development of your own solution should be multiplied by two or three. Do you estimate a year and a half? It’s three to four then, you may take our word on that. With a probability of 50% the project will never be finished;
• the uncontrolled growth of costs comes out of the previous statement. We should also note that in 100% of such cases, the labor demands increase during the implementation of the "project" — you may multiply by two for the beginning;
• for obvious reasons, you will have to freeze all changes in the business for the period of the development. Do you think your competitors will put themselves in your shoes and just wait?
• The decision will turn out to be dock-tailed and poor. It’s exactly the same as if you tried to compare the cobbled together in a garage car to a LADA, to say nothing of a Mercedes. There is a certain reason for pride: you have put so much effort into it. "But does it actually go? — Well, it’s not the point. " In our case every time you try to implement changes in business to the accounting system, you will constantly bump into obstacles like " we can not do it" "Well, theoretically we can make this in year and a half, but can’t you think of something better? "," No, connection of such type equipment was not taken into account at the development. " In the end you’ll have to adjust the real business to the absurd restrictions of non-professionally designed and implemented crooked system.
• you will inevitably find yourself strongly dependant on a few programmers and the head of the development project. Should I go deep into details here?

Well, and now comes the question " what for?"

An example from life — an interview with Sergei Galitsky, the co-owner of "Magnet", to "Vedomosti".
The following is an excerpt from the interview, which would be of certain interest in the context of our topic:

— Do you have your own software to monitor the network performance, or do you use ready-made solutions?
— All programs are our own, we have about 100 programmers who work on it. The existing solutions do not suit us, and it would be functionally wrong to use it. You can not really do anything to a finished product, but when you write the program yourself you have a chance to customize it for your business. There is nothing to be particularly proud of — it’s just a vitally necessary thing for us. How can one manage a retail company without monitoring the remnants? Although in the end it’s people who control everything. There is no such ephemeral programs that would do everything for you: you just press the button and it tells you what you have and where.
...Because you have three million customers every day, you make approximately 2,000 payments to your suppliers every day, you have automatic ordering of goods in each store, the automatic orders to our suppliers, maintaining of range of products in stock. All this with a uneven demand — first holidays, then it’s hot, and later it’s cold. We have a "KAMAZ" full of papers arriving to our head office every day, and this is only to the head office. Can you imagine how to control all this?

Well, it’s out of the question that Sergey is a skilled retailer, the results the "Magnet" company shows are the best evidence.
But this talented retailer does not understand a damn thing in IT. This can be simply by the number of programmers.

We do not hope that honored Sergei Galitsky will ever take time to read this page, we only mean him as an imaginary opponent.

Let go step by step:

— "The existing solutions do not suit us." ." Sergei have you looked at all of them yourself or was it your IT Director to show you around? He’s a person concerned in this case.

— "There is no such ephemeral programs that would do everything for you: you just press the button and it tells you what you have and where." Sergei, you are mistaken. This is the basic functionality of any storage program (even without the proud name of ERP). If rated honestly, why does one need an ERP-system, not to mention the herd of programmers, if it doesn’t provide the chance to see what’s where with a click of one button? Absolutely all of our clients get the information about the warehouse stocks, transits and other states of their inventory this way. It’s absolutely clear that it takes not only the buttons in the program, but also the correctly aligned business processes to do that.
But as you know, "we do know how to do it."

— "...Because you have three million customers every day, you make approximately 2,000 payments to your suppliers every day..." How damn nice it is to show off the scale of the empire you built, right? But we do know that the properly aligned processes serve three thousand people or three million (in terms of software) — with the only difference in the power of the used servers. Especially, if you divide the "three million" by the number of stores, you get the same 2-3 thousand people per day. In general, it’s overall not so impressive, if each store operates on a single system (and this is most likely so, and then there comes the famous shops selling synchronization with the central database and pile of bugs as a result. So, Sergei, is everything fine with that? Not too tight?).

— You have automatic ordering of goods in each store, the automatic orders to our suppliers, maintaining of range of products in stock... Well, that sounds impressive, but only for those who have no clue about how it works. And since we do it again does not impress us.
All of your automation is certainly based on an extremely dull assortment matrix for each store, and it’s occasionally manually changed. And accordingly, the primitive output agent constitutes an invoice using the difference between the matrix and the actual stock. And, to say the truth, neither the heat nor the cold, or the uneven demand are taken into account. And the human factor is sure to lead to some skews.The dynamics of demand, the calculation of needs at integrated speed sales, the dependency analysis and a lot of other useful things are simply not taken into any account. You may come to us when you fire the 100 moochers you have for now and we’d share this information with you. The exchange will appreciate it if you manage to raise the revenue by 7-10% per square meter.

We have a "KAMAZ" full of papers arriving to our head office every day, and this is only to the head office. Can you imagine how to control all this? Of course, not.
And you can not either. On the other hand we can offer a solution that will relieve you from this paperwork. And if you think that this "KAMAZ" of papers (the "Chechen terrorist KAMAZ ") can be controlled by anyone in your company, we will disappoint you. Despite the fact that the money you spend on this "control" is a lot, it only results in another reason for pride for the amount of processed junk.

— We have about 100 programmers who work on it. Intuitively, it seems that instead of "it" you meant to say "work on all this crap." 100 programmers is a no laughing matter. This number can be a reason for pride for your IT Director, and at the same time, the reason for grief for you and your shareholders: a 100 of most likely well paid moochers.
You can imagine how much money has been altogether spent for nothing on this "software". And the result? Some kind of crap. The kind of crap that can not even show the current position of the goods, and the kind of crap that positions the primitive functional of autogeneration of invoices on the base of the assortment matrix, that will take any third-year student a half an hour to write in Excel, as a crown jewel.
every now and then). In the rare hours they may do something really useful at the request of other departments but only provided that these departments raise a hue and cry of the "If you do not do this week,
[it’s all over]..." So the waiting list of the tasks to be implemented is some months long, the order of implementation doesn’t depend on the actual need or the queue, but only on the loudest hue and cry, and your IT director is always complaining about the overload and a lack of human resources.
For your consideration: the largest domestic software company (in our market at least), 1C, employs about 200 programmers.
With all our prepossession towards the colleagues, we have to admit that their products are used by millions of people (compare it to the number of employees in your company, we bet not more than 10% actually work with the system).
And the diversity of even basic configurations, not to mention the customized ones is worthy of respect. That’s the most simple benchmarking, that answers all the questions about effectiveness.

Homo errare est, and people are also very persistent in their errors.
We are very sure that while Sergei is the head of Magnet, the company’s IT strategy is unlikely to change.
But the morning sun never lasts a day. Do you, Sergei, know what will the first undertaking of the new management (once it comes in) be? Yes, you’re right — the implementation of SAP.

In general, there is one claim, Sergei, we totally agree with: "There is nothing to be particularly proud of." You’ve said a mouthful. Of course, we mean the IT side of your business as the achievements of your business in the core competencies are needless to say.

Due to the size, "Magnet" company can boast the ability to spend some millions on such cute toys.

Russia is some decades behind the developed markets in terms of development of the food retail sector. Virtually no competitive environment makes it possible to cover losses created by the inefficient management by adding some five percent to the price or pulling an extra penny from the suppliers (as we all know, "one supplier — a penny, but a hundred suppliers — a dollar.") It’s also a good option to use second-rate meat for the ready-to-use BBQ or chop the faded greens down in salads.

And do you, dear reader, have some extra couple millions? Does your business enjoy the same hothouse conditions with always pliant suppliers and buyers, that have almost no choice?

If so, we are extremely happy for you. In this case we suggest you pay attention to something more hedonistic than the EPR systems, such as collecting modern art.

And if not — welcome. Ultima ERP is a system for efficient business in the competitive environment, rather than for an empire a la "KAMAZ of papers."