Most companies sooner or later start to experience the issue of inter-division chaos.
A company usually starts small – let’s say 2 founders, 2 more employees. They invent a great online product and grow steadily – expanding and dividing itself into smaller divisions and subdivisions, employing more people, putting new ones in charge of smaller and smaller pieces of the same company.
Then, most companies start to internally fluctuate on those small differences. Let’s say – 2 databases used to store same user data in different format, 4 applications where each one partially uses one of those 2 databases. Next comes exceptions for international users and a new database created for that sole purpose etc.
And soon – we have a hard to solve puzzle that’s steadily growing with the company expansion itself.
Build upon foundation instead of prediction
When we start new business, most of the time we start with a vision. Let’s call that a prediction. Prediction of how we see the company working, what will our users want from us, what we will want from them etcetera.
At the beginning, this is good. But if we try and build up our company this way, sooner or later it’s going to cause trouble.
Reason for this is a growing numbers of bosses (let’s keep it simple and say one per division). They all have great visions and mostly no idea about the existing infrastructure. It’s reasonable to say that all managers can’t be not bound to understand all small aspects of the company. But when vision becomes a project, that’s where it’s time to build upon a foundation that already exists within the company – even if one has to ask his junior subordinates to help out understand how to best lay stuff out!
So instead of designing a completely new internal application infrastructure (let’s say for an accounting project) that will store our users in yet another database (!) with its own structure (!!) – normalize those 2 ones that already exist and build against the single normalized database!
My hammer weights more than yours
This is a common obstacle in many companies – the boss infrastructure.
Over time, each division will hopefully gain a responsible and reputable manager person that will bring much good to the company.
Then, the time will come to rule upon various bigger or smaller issues – and many of these reputable management people will turn into bloodthirsty vampires with huge silver hammers for their colleagues.
If company founders allow for this to happen – most of the time due to the ongoing expertise of those management vampires – it will ultimately lead to yet another round of building upon visions rather than foundations.
If your primary focus is on the welfare of your company, do it like Mark Zuckerberg - make your products for the people who will use them, not for the egos of your management personnel.