Few people want to take junior developers on a project that is bursting at the seams. However, they do. And I noticed that there are companies that are afraid to hire seniors. The company will prefer to hire several juniors instead of one senior. The idea of taking one senior instead of middle or junior, one might say, is not even considered.
Let's give a possible reason for building HR policy on the recruitment of juniors and their subsequent growth.
Disclaimer # 1 : Of course, the division into junior-middle-senior, used in the article, is extremely conditional and has little to do with work experience, and the gradation and salary levels strongly depend on the company in which the assessment is given. However, we live in a reality in which such terminology is used at hh in the vacancies of leading it companies, and we all, one way or another, understand what it is about, and approximately what set of requirements is presented in this or that case. And anticipating the further text of the article, I assure you that the absolute values are not so important for the conclusions drawn.
Disclaimer # 2 : By and large, the reasoning is not about the top companies in which everything is in order, but rather about the rest of the majority. We are also talking about technically complex products, although I don't really believe in the possibility of simple ones, in general. Simple products (remembering Paul Graham and his book "Hackers and Artists") are unlikely to successfully exist in a competitive market, otherwise they would be too easy to copy and repeat.
In the classic version, of course, there is a desire to save money.
Junior can be paid little. There are a lot of them on the market, they are easier to get. Yes, in the first months it will be useless, but after a few months it will start to be useful, and in a year it will beat off all costs. And you can also save, or rather lose a little, if the junior has not passed the trial period. This loss is not particularly painful compared to the loss of a higher-paid specialist. One senior can parallelize the work, loading three juniors with tasks of low complexity, taking on more high-level and complex tasks. Thus, we will make a product, but we will pay less, although it may not be so perfect and fast. And by taking a junior on a small salary, we do not undermine the established salary levels within the company.
Almost all arguments are untenable. Let's go in order.
, , , , , - . . .. , . . , , . — , — 3 . , 6 3 . 2 , 2/3 , 1/3 . 3 — 1/2 1/2 . .
. , . .
, , , . , , .
" -":
, . . , . 10:1, — 5:1. , , 20 . , 10 , , 10 . . . , . (, , .)
—
3 , 3 10-20% . , . , . , Skunk Works, .
— ,
. , . , . .
, . , , . , "" , . . , "" , . .
, , , .
, . . , .
. . . .
, . , . .
vs
. , , , . , . , , , .
, .
— . ? , .
, , , . , , .
. , .
, , . 3, .
:
? ?
, "". , . 5 1 1 . , , devops', .
, 550-600 ..
7 , .
:
.
1) - — 150 .. (8 — 4x ( ) = 4x)
2) ? — 110 .. (5 — 1 = 4)
3) ? — 80 .. (2x)
4) ? — 65 .. (1x)
5) ? — 60 .. (1x)
6) — 55 .. (0.2x)
7) — 50 . (0)
: 12.2x
, 5% . .. : 12.2x * 0.7 = 8.54x
: 66 .. 1x .
, :
- . . , . . , , , .
- . .
1) - — 170 (8x — 2x (. ) = 6)
2) — 160 (8x)
3) — 160 (6x)
4) — 100 (3x)
: 23x
:23x * 0.8 = 18.4x
: 32 .. 1x .
, :
- ,
- (, , , )
The recruiting strategy based on juniors is notoriously losing to those who will recruit seniors. A senior team is much more cost effective than a team with a large number of juniors and middle men at any given time. A more scrupulous calculation of efficiency, at least taking into account the factors mentioned above, will only widen this gap. A reason to take juniors and raise seniors, maybe a case when seniors cannot be found in principle, or as a way to intercept super talents, but keeping in mind that it will not be possible to return to the previous performance after hiring a junior soon.