The practice of running large projects from Oracle
Greetings.
In April of this year, by the will of fate, I was brought to the Oracle Commerce Conference in Lisbon. It doesn’t matter what exactly Oracle Commerce is like a product (if you're interested, I can write a separate topic), I wanted to talk about something else - namely, one of the largest suppliers of electronic commerce automation projects and sales industrial software.
In domestic practice, the overwhelming majority of integrator companies with whom I had to deal with one way or another, advocate an approach "and this is all, and this is all." That is, if we are talking about selling a project to automate any business processes, then a leather contractor climbs to prove its own global superiority in the market and a deep understanding of all possible areas of the customer's business. It would seem that this behavior is logical: the contractor thus collects the entire budget in one basket and seems to be signing up for the provision of a single integrated IT-landscape ... but there is a nuance.
A large trading business has a lot of requirements for each functional unit of the system. At best, the customer already has an ERP system with a host of business rules, including supply chain management, assortment management, forecasts and automated orders for suppliers, and a very sophisticated WMS fits in, somewhere alongside financial management, and all this is accompanied accounting and business intelligence. And now we are fastening the e-commerce system to all this, which, in accordance with current trends, should cover all sales channels, including the traditional ones.
')
In general, it is clear that when creating such a system without interfering with business processes that already exist in the enterprise management system, it is indispensable. That is, changes are necessary not only in front-end, but also in back-end, and the system integrator must have the appropriate competencies. So do you really think that one integrator company, albeit an experienced and large one, is able to ensure equal success at all sites?
I will even say more: now there is one very large project in Russia (about which I cannot disclose the details), and so there is one not the smallest western contractor with many years of competence in Oracle Commerce who, in turn, is working on it attracts to the project a team of offshore programmers and one of the largest consulting companies in the world. It is clear that you can talk for a long time about managing your own team, the advantages and disadvantages of this or that workflow model, but I want to emphasize that Oracle literally encourages the cooperation of partners when working on large systems. All participants in the process need good reference projects and clients who have achieved significant business success with the help of commissioned systems, and often the integrator should not engage in the development of a certain competence in his own company to the detriment of the timing and quality of work.
In a sense, awareness of this cluster approach has turned my understanding of project management. Having a certain experience in building such systems, I even didn’t even think of such a delegation of design work. We created even relatively large systems relying solely on our own strength, even if sometimes with the help of individual external freelance specialists, but not companies. Of course, it was cheaper for the customer, but in the end he received a system that did not fully take into account the best experience and was created by people who were not always well aware of the business development prospects. Obviously, a well-designed e-commerce system is able to fulfill the current set of requirements, but what will happen to it in a year? And in two? What, it will be necessary to endlessly end up with some kind of custom solution in order to finally rest on, say, a performance limit?
As a bright negative example, I know one rather rather big domestic trading company, which has its own staff of about 60 programmers who support the samopisny ERP-system. A special bonus is LAMP. You can imagine how this monster is updated and maintained.
There is a positive example, though, not ours, but the Dutch (if you need details, I will write, the project is interesting and, alas, absolutely impossible in today's Russia). The technological landscape is as follows: Oracle Commerce at the front, SAP ERP and Manhattan WMS backs up and manages all this with its own IT staff of 6 people, including IT directors.
That is why big business, counting money and working for the future, chooses industrial solutions: it buys not so much a software package as the notorious “best practices”, a guarantee of the absence of typical rakes, the existence of which he may not even guess at the current stage of business development. And in order to guarantee the result, Oracle insists on the participation in the projects of consultants (including its own) and other technological partners. This, it seems, helps to develop the most optimal solutions according to the principle “one head is good, but two is better”. That is, there is a hosting company with environments prepared for hosting a commercial application, there are consultants, there is a team of designers and project managers, and there are companies offering appropriate offshore programming. In sum, these several companies share the project budget and responsibility among themselves.
I wrote all this to what: colleagues, especially those who have experience in implementing truly large projects! What do you think, what approach in our realities is still more appropriate: “multi-core” or “single-window principle”? I will especially be grateful for stories from personal experience: this will allow me to put in order my own shaken ideas about project management.
In April of this year, by the will of fate, I was brought to the Oracle Commerce Conference in Lisbon. It doesn’t matter what exactly Oracle Commerce is like a product (if you're interested, I can write a separate topic), I wanted to talk about something else - namely, one of the largest suppliers of electronic commerce automation projects and sales industrial software.
In domestic practice, the overwhelming majority of integrator companies with whom I had to deal with one way or another, advocate an approach "and this is all, and this is all." That is, if we are talking about selling a project to automate any business processes, then a leather contractor climbs to prove its own global superiority in the market and a deep understanding of all possible areas of the customer's business. It would seem that this behavior is logical: the contractor thus collects the entire budget in one basket and seems to be signing up for the provision of a single integrated IT-landscape ... but there is a nuance.
A large trading business has a lot of requirements for each functional unit of the system. At best, the customer already has an ERP system with a host of business rules, including supply chain management, assortment management, forecasts and automated orders for suppliers, and a very sophisticated WMS fits in, somewhere alongside financial management, and all this is accompanied accounting and business intelligence. And now we are fastening the e-commerce system to all this, which, in accordance with current trends, should cover all sales channels, including the traditional ones.
')
In general, it is clear that when creating such a system without interfering with business processes that already exist in the enterprise management system, it is indispensable. That is, changes are necessary not only in front-end, but also in back-end, and the system integrator must have the appropriate competencies. So do you really think that one integrator company, albeit an experienced and large one, is able to ensure equal success at all sites?
I will even say more: now there is one very large project in Russia (about which I cannot disclose the details), and so there is one not the smallest western contractor with many years of competence in Oracle Commerce who, in turn, is working on it attracts to the project a team of offshore programmers and one of the largest consulting companies in the world. It is clear that you can talk for a long time about managing your own team, the advantages and disadvantages of this or that workflow model, but I want to emphasize that Oracle literally encourages the cooperation of partners when working on large systems. All participants in the process need good reference projects and clients who have achieved significant business success with the help of commissioned systems, and often the integrator should not engage in the development of a certain competence in his own company to the detriment of the timing and quality of work.
In a sense, awareness of this cluster approach has turned my understanding of project management. Having a certain experience in building such systems, I even didn’t even think of such a delegation of design work. We created even relatively large systems relying solely on our own strength, even if sometimes with the help of individual external freelance specialists, but not companies. Of course, it was cheaper for the customer, but in the end he received a system that did not fully take into account the best experience and was created by people who were not always well aware of the business development prospects. Obviously, a well-designed e-commerce system is able to fulfill the current set of requirements, but what will happen to it in a year? And in two? What, it will be necessary to endlessly end up with some kind of custom solution in order to finally rest on, say, a performance limit?
As a bright negative example, I know one rather rather big domestic trading company, which has its own staff of about 60 programmers who support the samopisny ERP-system. A special bonus is LAMP. You can imagine how this monster is updated and maintained.
There is a positive example, though, not ours, but the Dutch (if you need details, I will write, the project is interesting and, alas, absolutely impossible in today's Russia). The technological landscape is as follows: Oracle Commerce at the front, SAP ERP and Manhattan WMS backs up and manages all this with its own IT staff of 6 people, including IT directors.
That is why big business, counting money and working for the future, chooses industrial solutions: it buys not so much a software package as the notorious “best practices”, a guarantee of the absence of typical rakes, the existence of which he may not even guess at the current stage of business development. And in order to guarantee the result, Oracle insists on the participation in the projects of consultants (including its own) and other technological partners. This, it seems, helps to develop the most optimal solutions according to the principle “one head is good, but two is better”. That is, there is a hosting company with environments prepared for hosting a commercial application, there are consultants, there is a team of designers and project managers, and there are companies offering appropriate offshore programming. In sum, these several companies share the project budget and responsibility among themselves.
I wrote all this to what: colleagues, especially those who have experience in implementing truly large projects! What do you think, what approach in our realities is still more appropriate: “multi-core” or “single-window principle”? I will especially be grateful for stories from personal experience: this will allow me to put in order my own shaken ideas about project management.
Source: https://habr.com/ru/post/178993/
All Articles