Because there is a recruiting department flush with cash but there is no retention department.
Retention is dumped on some engineering Director who dumps it on their engineering manager. Then it becomes a matter of report <-> manager rapport and the skill of the manager.
How many of us have seen incompetent managers? Yeah, they are not going to be good at retention either.
I’m not in those circles but I suspect many in management are not trained that “loyalty” is a two-way dynamic. Honestly I don’t think it’s all about compensation either, that’s a signal. I.E. the company is willing to invest in me. I once worked at a large company where the pay was average but they had an extremely healthy company culture to where I truly believed that the company cared about its employees and community. As a result most people there were happy and were (or were on track to become) lifers. The culture wasn’t for me but it was interesting to see that sometimes compensation is not everything with the right culture. However if the relationship is “transactional” then compensation becomes everything.
Because people don't like getting nothing for something. If I'm at the point of asking for a raise, the value proposition I'm offering as an employee is that I can do the same work for more money - alternatively I can threaten to leave, if my demands are not met - in other words, extortion.
I managed to get a reasonable raise only once in my career, otherwise if you want a higher salary, you must change jobs unfortunately.
The mental framework managers suggest that employees should adapt - that if you create greater value for the company, you will receive greater pay - up to and behind the point other employers are willing to offer you - is just a negotiating tactic they use.
for software developers it takes time to learn their way around a new code base (usually 3-12 months), so low retention is highly detremental.
economically a business can be better off giving a big raise than losing someone, paying recuiters, losing time on finding a replcement, losing time until the replacement gets up to speed, and eventually losing the replcement too.
if this whole cycle could mean for example the business only get 1 year of effective work out of 2 years, in which case doubling the salary to retain someone is just as effective.
also it isn't extortion if someone gets a better offer elsewhere, and the offer cant be infinitly high.
to conclude no retention means lack of second order thinking.
Except then word gets out that you’ll get a big raise at company X if you threaten to leave and they have to start paying everyone that amount. Even the people who would have otherwise stayed for 5 years at the lower salary.
This is why I’ve always assumed you have to change jobs if you want a raise.
I'm not advocating for this style of management thinking - in fact I despise it. But this is how many(most?) managers think, especially at big companies where productivity is not measured, or worse, measured poorly. In these circumstances, it's very very hard to get them to offer you a raise, and even if they do, it would be to prevent potential bad behavior (you leaving), than rewarding your good work and expertise. In fact, if you perform head and shoulders above your peers, the cause of the raise will still be the first reason - you'll get pissed off that your work isn't recognized, which will prompt you to consider leaving, which the management will try to prevent.
That's why the people getting raises won't be the ones who do the best work, but the ones, who complain about their salaries the loudest, or go to interviews and get counteroffers.
Remember - you are probably getting paid as little as the company can get away with paying you.
The employee who proposes leaving just has a good BATNA. We sell our labor and should let our purchasers know when we're raising the price. They can pay the new price if they choose. We don't need the moral weight of calling it extortion.
Maybe it's the people responsible for retention just don't have the time to focus on retention most of the time? They're busy putting out fires, going to meetings, worrying about numbers and other things that managers do. They just can't spend enough time on things that keep current employees around longer. Plus, at most places, retention isn't a number that is measured, or if it is, it doesn't matter much.
Since retention doesn't count, it gets pushed aside most of the time, so the people responsible end up spending more time on hiring/recruitment, which gives them less time to think about retention.
I think it would be nice to see some exit interview statistics, like how many employees leave because of money. I personally dont think that is the main reason.
Yes, if I want more money, i go to interviews, get a better offer and negotiate with my company a higher salary. Every single time they matched the value (sometimes asking 20%).
But every time i left a company it wasn't about money. It's usually culture, not believing in what they are doing. Incompetent manager or product guys.
There's a large cohort of people who don't list their real reasons for leaving for the sake of departing on good terms and to avoid burning bridges unnecessarily.
Exit surveys don't fix a rotten culture and incompetent managers. We can only hope the higher ups look at the numbers and start asking why certain managements have a higher turnover rate then others.
A few anonymous surveys targeted at those teams from people who can't as easily uproot their lives to leave the company will tell you what's going on.
It's easy to create hiring/recruitment metris but how do you create a good one for the retention? Why would you give a reward to a retention manager? How to measure retention manager's performance?
Because people get complacent and changing jobs is a lot of work. Updating resumes, writing cover letters, prepping for interviews, kissing butts in interviews, taking drug tests, onboarding..thinking about going through that makes my skin crawl. Most people won’t go through that until they are pushed to their limit by a company. And every job I have quit was because of culture and/or incompetence, had nothing to do with money.
5 6-month stints in a row is a red flag for me (as a hiring manager).
If you have 3 18-24 month stints (and ideally earned a promotion at any company [rather than only on a company change]), and a couple of shorter ones, it's fine.
People are overly worried about the second case, IMO. If you're at a company that sucks, the very best thing you can do is change companies.
They do. Job hopping every 1-2 years is pretty standard for software engineers, and it's pretty common for people to feel like the only way to get a significant raise is to get a new job.
I job hopped throughout my 20s, partially to increase my income but also because I was hoping to find a company that isn't dysfunctional. So far I haven't found one. The biggest problem with retention, IMO, is that the vast majority of companies are shiny-object chasing feature factories. The framework, language, and library churn is just exhausting and saps all the fun out of writing software.
Sometimes an employee that was necessary at a certain stage in the company's growth is no longer required once the company is in a different state, usually because of a core skill set that isn't in the roadmap anymore, or crisis management style that is more wartime than peacetime.
Those employees are usually aware they're in that boat though and sometimes given "severance", to the tune of a reduced workload at the same salary so they can interview at other companies.
Because retention is risk management and companies and managers aren’t great at it.
A related pattern is failure to promote from within the ranks, especially if a veteran employee is responsible for too many things.
One small benefit is that you might be able to hire a more relevant person for a specific position by looking outside, but ramp up to company norms can be as slow as ramping up an internal person to new skills, which is also hard to quantify for managers
I think it's because most companies know that change is difficult and painful to most people. There's an "activation energy" that it takes for you to switch jobs, particularly if you'll need to move to a new location.
I worked for a company that exploited this very successfully by having an office in a city where there were not many alternatives for comparable work. Anyone they recruited couldn't really change jobs without selling their house, uprooting their spouse's life/career, and sending their kids to a new school. The company had a strong incentive to pay you to move there, but it cost very little to encourage you to stay.
But if a bunch of companies are moving to Texas, doesn't that make it easier to switch between the companies that are moving to Texas? BTW, Austin has had a strong tech and start-up culture for at least the last 30 years.
I've only worked at GAFAM (GAMAM now?) and I'd say half the people I have worked with have stayed longer than 4 years at the same company past the 4 year cliff. I think it probably is actually cheaper to focus on recruiting to get the half who job hop while letting the other half be complacent.
A lot of times it's better to let unhappy people take their unhappiness elsewhere. I've had plenty of experiences where employees wanted reasonable changes to salary, working arrangements, or equipment and we accommodated them. However frequently people are unhappy for reasons that won't be changing or make unreasonable demands, and since these people will never be satisfied anyways it is easier to let them leave and replace them with somebody who wants what you have to offer.
If companies don't do what you think they should then start your own, and if you're right you should be able to out compete them.
Is it really true that companies spend so much more on recruitment than retention?
I happen to know the founders of Culture Amp [1] (just through being in the same city and hanging around in the same startup founder circles when we were all starting out 12+ years ago).
They've spent over a decade building a platform to gauge employee satisfaction and company culture health, to help companies with retention.
They charge a lot for it and they have over 4000 companies using it including some of the largest/biggest companies/brands, and including newer Silicon Valley companies and old/established companies.
They seem to be doing incredibly well.
Where's the evidence that companies don't take retention seriously?
It seems self-evident to me that in a world where the top talent (particularly among developers/engineers) is so highly prized, companies that neglect employee satisfaction and retention will start declining quite rapidly.
I don't have a large sample set, but from what I HAVE seen is that most of these surveys are run with altrustic motivations/intentions, but simply do not change anything.
In the worst cases, providing critical feedback can be a CLM. Causing problems for your manager/s generally doesn't help you get more money. You end up seen as the problem, rather than a symptom or victim of the problem. This is partly, of course, because your average line manager doesn't actually have power to change anything, just to manage you, their peers, and their managers.
I don’t know personally as I haven’t been in a company the uses it, but my understanding is the Culture Amp platform is designed to get around these pitfalls, and that’s why companies pay a lot to use it.
Surely a company that just allows these issues to persist will lose talent to companies that can do better at culture and retention, even if just a little?
I certainly know the founders/senior execs are very smart and also very well intentioned, and they’ve been growing strongly for a long time, so they must be getting some solid results for their clients.
Why do companies not address CUSTOMER retention? It's the same reason.
The things you must do for retention are both "boring" and not revenue/reward generating, and you have to "bake in" those measures into your business core processes to have good retention of either customers or employees.
And most people starting and running companies largely achieved their current success by accident and despite their lack of skills. They see success as "repeating what already worked" and doubling down on that to attain growth. You have to "go meta" and revisit things they are "convinced" are already fixed.
You can't address customer retention by any of the things that generate leads or close sales - because retention is all about things you do AFTER the revenue is realized and thus it's accounted as a cost. Similarly employee retention requires actions that occur AFTER HR got the win of the hire.
Both retentions require lots of costs and actions that are "tedious" and "not rewarded" to the people who typically decide the things that must be done. HR, upper management, etc. And even direct managers are not "measured" on anything that helps retention.
My guess is lack of metrics. If they could tell what it cost and justify it as a savings then they would do it. Most leadership is as rational as they can manage.
The reality is that companies can't measure impact of individuals or small groups particularly well.
Certainly they can't tell the difference between good and bad employees so how can they tell who to attempt to retain?
More or less attempting to retain everyone at their market value is not on the table at pretty much every company even if they have the money.
Maybe some day someone will apply some science to the whole thing and we can find out which strategy is optimal and when.
> Certainly they can't tell the difference between good and bad employees
I have a much easier time telling the difference between "these are the people with whom I'd form a 3-person startup [vs those that I wouldn't]" than I do telling the difference between "these are the top 50% of my org who are diligently keeping everything running well [vs those who aren't pulling as much weight]".
Which is the category of "good vs bad" that matters for a company's success? I think the latter one would be more impactful if solved.
In my current role I am writing a promotion packet to go to the next level. The bar is relatively low, it must only demonstrate that I have been performing as an N+1 level. It requires extensive documentation and close to a dozen personal internal references.
The external interviewing process into that role must demonstrate that the candidate has been performing at a level that is higher than average for that role. The promotion process takes 1-2 orders of magnitude more time for a known internal quantity than it does for an external unknown quantity.
There is a marked focus on improving the recruitment process but little on improving the promotion and retention process.
May I ask if you're at Amazon? That particular bar raising process is fairly unique to Amazon, or at least started with Amazon.
One struggle as a result is that the levels themselves slowly shift upward; it's harder and harder to get promoted as time passes. Other companies tend to have level compression, which Amazon's approach attempted to address. Now Amazon has a problem in that level 6, accepted as the typical terminal/career level, has become larger and larger as it stretches upward.
This is an incredibly important contributor to all of this.
At a prior company, in order to be promoted from associate to senior staff, you had to demonstrate and be evaluated performing at senior level while associate for a year. In order to get hired as a senior, you just talked with a hiring manager and an HR rep, checked off some boxes based on a few blurbs about what you did at prior jobs, and boom, you were a senior. Same thing with getting a raise. To get a raise to 150K (just as a randomly selected round number), it was an enormous bureaucratic process going up 5 levels of management that would take months with lots of aggravation - and unlikely to really pan out. To get hired at 150K for the same role, it was a couple phone calls with an HR rep and hiring manager - about 15 minutes in total.
On the subject of money - the other issue is that new hires, being certified as better than average in that role, are getting compensation packages in the top half of that band. The people getting promoted to that role are in many cases getting nothing or maybe a few thousand extra that vests in 24 months, because the stock appreciation on their N-level compensation package has lifted them into the bottom of the N+1 payband even before they were promoted.
Recruitment is a one-time cost per employee (or per candidate, is another way to look at it).
Retention costs include salary and other benefits, which are ongoing, recurring costs and can add up to quite a bit… so I would hardly say they aren't spending on retention.
Honestly I think it's simply because they don't think of the employee at all
Obviously they're a resource that needs paying, they exist, but overall I don't think companies notice people for people until they announce their intention to leave
My sample size is small so this might be way off but here is my guess:
Companies fundamentally want* to be able to treat software developers as completely fungible assembly line workers. The company's dream is to be able to fire and replace all of the software developers at any moment and lose little by doing so. This partly explains the heavy emphasis that tech companies place on process, project management, and so on even if the developers dislike all of those things. The company must pay for recruitment - there is no other choice because the company must get employees in. However, to pay for retention would contradict the fundamental company dream of making all of the software developers completely replaceable. Basically, retaining the employees is not one of the more significant company goals. Companies' actual goal usually is to make it so that the developers are as fungible and replaceable as possible. Companies want to be able to treat the engineering team kind of like an abstract arbitrarily scalable asset in a video game. Want to replace all of the engineers? Sure, push a button and it happens - you might lose 10% of this quarter's numbers but that's acceptable. Want to scale your team up and go from having 15 developers to 30 developers? Push a button! Want to go up to 60! Boom, do it - the point is that the company has the process that makes the developers fungible - to focus on retention would be silly, the company's goal is to focus on recruiting and then turning the new recruits into replaceable and fungible assembly line workers who are part of an arbitrarily and abstractly scaleable team. Companies don't dream so much about hiring amazing developers - they dream more about setting up a process that will let them hire developers at scale and plug them into a money-making machine that keeps going no matter how retention is going. As a developer working for a company, the end-users are not your real customers. The management and all of their politics are your real customers. What they want is to not think about you. If I go to a restaurant and order some food, generally I do not want to think about how many people are working there or what exactly they are doing - I just want my food. Same with companies and developers - what they want is to just be able to turn the development team into an abstract source of software development. Focusing too much on retainment would tie them too much to a given set of developers. They do not want that. What they want is to set up a process that lets them treat the engineering team as an abstract and arbitrarily scaleable machine that creates software.
*When I say that companies want something this is of course anthropomorphizing a process but what I mean is that the overall totality of the people in the company operating according to various incentives lead to things happening that might in short be described as the company wanting certain things.
One thing that seems to be missing from this discussion is employee performance. In my experience, well functioning organizations go out of their way to retain top performers (with better pay raises, bonuses, equity, project opportunities, etc).
Average performers are valued but replaceable. They probably will see nominal raises over time. Minimum effort performers are just a drag on the team, as they occupy a spot that might go to a better performer.
While it costs to hire and train someone new, you also get new ideas and possibly someone better than average (if the company has a good hiring process).
Companies asked spend a ton on retention- your salary, benefits, and VR ping pong tables are all in place to keep you there and working. That sgock vesting schedule? Retention technique.
If all those things aren't working and an IC has gone to the bother to get an offer somewhere else than all those things have failed.
Why don't companies just match whenever an employee gets a higher offer? There is a risk of uniformly increasing costs beyond that single employee, there is a quite real possibility that the current company evaluate the employees value lower than the offering company, and there is also a widespread and not entirely unfounded belief that an employee that has already gone through the bother of interviewing elsewhere is likely to have other issues beyond compensation and they are likely to remain a high risk of leaving, so the company will risk establishing a norm of higher compensation across the board while being unlikely to retain the person long term.
The truth if companies in our industry all spent "more" on retention you'd just see salaries and compensation rise across the board. Given the high pay in our industry, I'd say that's already happened/happening.
If you are trying to build a product that will last, this is probably not a good approach. You'll need some people who are more senior to design your systems and to train the new grads.
The problem comes down to this: companies are run by children. I mean that literally, in the sense that the older I get, the less surprising it is that people younger than me do a poor job of management.
The older you get, the more appreciation you have of how many and how deeply the misconceptions about how to run companies are and go.
Retention of good people is very very important. But also the nature of that retention: you don’t want them in prison, you want them to be in paradise.
Manage people like beekeepers manage bees: they are always free to fly away, but they don’t want to, because this hive here is so nice!
70 comments
[ 4.4 ms ] story [ 140 ms ] threadRetention is dumped on some engineering Director who dumps it on their engineering manager. Then it becomes a matter of report <-> manager rapport and the skill of the manager.
How many of us have seen incompetent managers? Yeah, they are not going to be good at retention either.
Otherwise, it's hard to measure the success of the "retention department."
Only one group has the job of making new people show up: recruiters.
But if nobody leaves, do we bonus the 'retention department' or the regular manager?
I managed to get a reasonable raise only once in my career, otherwise if you want a higher salary, you must change jobs unfortunately.
The mental framework managers suggest that employees should adapt - that if you create greater value for the company, you will receive greater pay - up to and behind the point other employers are willing to offer you - is just a negotiating tactic they use.
also it isn't extortion if someone gets a better offer elsewhere, and the offer cant be infinitly high.
to conclude no retention means lack of second order thinking.
Except then word gets out that you’ll get a big raise at company X if you threaten to leave and they have to start paying everyone that amount. Even the people who would have otherwise stayed for 5 years at the lower salary.
This is why I’ve always assumed you have to change jobs if you want a raise.
That's why the people getting raises won't be the ones who do the best work, but the ones, who complain about their salaries the loudest, or go to interviews and get counteroffers.
Remember - you are probably getting paid as little as the company can get away with paying you.
Since retention doesn't count, it gets pushed aside most of the time, so the people responsible end up spending more time on hiring/recruitment, which gives them less time to think about retention.
Yes, if I want more money, i go to interviews, get a better offer and negotiate with my company a higher salary. Every single time they matched the value (sometimes asking 20%).
But every time i left a company it wasn't about money. It's usually culture, not believing in what they are doing. Incompetent manager or product guys.
Exit surveys don't fix a rotten culture and incompetent managers. We can only hope the higher ups look at the numbers and start asking why certain managements have a higher turnover rate then others.
A few anonymous surveys targeted at those teams from people who can't as easily uproot their lives to leave the company will tell you what's going on.
If they did so then we will eventually reach an equilibrium where companies would spend more on retention than recruitment.
we do; it's great :)
https://www.inc.com/business-insider/tech-companies-employee...
> Uber was at the bottom of the list, with a short average employee tenure of just 1.8 years.
> Netflix has broken the three year barrier, holding on to employees for an average of 3.1 years, to be exact.
If you have 3 18-24 month stints (and ideally earned a promotion at any company [rather than only on a company change]), and a couple of shorter ones, it's fine.
People are overly worried about the second case, IMO. If you're at a company that sucks, the very best thing you can do is change companies.
I job hopped throughout my 20s, partially to increase my income but also because I was hoping to find a company that isn't dysfunctional. So far I haven't found one. The biggest problem with retention, IMO, is that the vast majority of companies are shiny-object chasing feature factories. The framework, language, and library churn is just exhausting and saps all the fun out of writing software.
Those employees are usually aware they're in that boat though and sometimes given "severance", to the tune of a reduced workload at the same salary so they can interview at other companies.
A related pattern is failure to promote from within the ranks, especially if a veteran employee is responsible for too many things.
One small benefit is that you might be able to hire a more relevant person for a specific position by looking outside, but ramp up to company norms can be as slow as ramping up an internal person to new skills, which is also hard to quantify for managers
I worked for a company that exploited this very successfully by having an office in a city where there were not many alternatives for comparable work. Anyone they recruited couldn't really change jobs without selling their house, uprooting their spouse's life/career, and sending their kids to a new school. The company had a strong incentive to pay you to move there, but it cost very little to encourage you to stay.
Edit: never mind. It becomes very clear if you spell it right, as a sibling comment pointed out.
throw microsoft in there and it becomes MAAMAN
If companies don't do what you think they should then start your own, and if you're right you should be able to out compete them.
I happen to know the founders of Culture Amp [1] (just through being in the same city and hanging around in the same startup founder circles when we were all starting out 12+ years ago).
They've spent over a decade building a platform to gauge employee satisfaction and company culture health, to help companies with retention.
They charge a lot for it and they have over 4000 companies using it including some of the largest/biggest companies/brands, and including newer Silicon Valley companies and old/established companies.
They seem to be doing incredibly well.
Where's the evidence that companies don't take retention seriously?
It seems self-evident to me that in a world where the top talent (particularly among developers/engineers) is so highly prized, companies that neglect employee satisfaction and retention will start declining quite rapidly.
https://www.cultureamp.com/
In the worst cases, providing critical feedback can be a CLM. Causing problems for your manager/s generally doesn't help you get more money. You end up seen as the problem, rather than a symptom or victim of the problem. This is partly, of course, because your average line manager doesn't actually have power to change anything, just to manage you, their peers, and their managers.
Surely a company that just allows these issues to persist will lose talent to companies that can do better at culture and retention, even if just a little?
I certainly know the founders/senior execs are very smart and also very well intentioned, and they’ve been growing strongly for a long time, so they must be getting some solid results for their clients.
The things you must do for retention are both "boring" and not revenue/reward generating, and you have to "bake in" those measures into your business core processes to have good retention of either customers or employees.
And most people starting and running companies largely achieved their current success by accident and despite their lack of skills. They see success as "repeating what already worked" and doubling down on that to attain growth. You have to "go meta" and revisit things they are "convinced" are already fixed.
You can't address customer retention by any of the things that generate leads or close sales - because retention is all about things you do AFTER the revenue is realized and thus it's accounted as a cost. Similarly employee retention requires actions that occur AFTER HR got the win of the hire.
Both retentions require lots of costs and actions that are "tedious" and "not rewarded" to the people who typically decide the things that must be done. HR, upper management, etc. And even direct managers are not "measured" on anything that helps retention.
The reality is that companies can't measure impact of individuals or small groups particularly well.
Certainly they can't tell the difference between good and bad employees so how can they tell who to attempt to retain?
More or less attempting to retain everyone at their market value is not on the table at pretty much every company even if they have the money.
Maybe some day someone will apply some science to the whole thing and we can find out which strategy is optimal and when.
I have a much easier time telling the difference between "these are the people with whom I'd form a 3-person startup [vs those that I wouldn't]" than I do telling the difference between "these are the top 50% of my org who are diligently keeping everything running well [vs those who aren't pulling as much weight]".
Which is the category of "good vs bad" that matters for a company's success? I think the latter one would be more impactful if solved.
The external interviewing process into that role must demonstrate that the candidate has been performing at a level that is higher than average for that role. The promotion process takes 1-2 orders of magnitude more time for a known internal quantity than it does for an external unknown quantity.
There is a marked focus on improving the recruitment process but little on improving the promotion and retention process.
One struggle as a result is that the levels themselves slowly shift upward; it's harder and harder to get promoted as time passes. Other companies tend to have level compression, which Amazon's approach attempted to address. Now Amazon has a problem in that level 6, accepted as the typical terminal/career level, has become larger and larger as it stretches upward.
At a prior company, in order to be promoted from associate to senior staff, you had to demonstrate and be evaluated performing at senior level while associate for a year. In order to get hired as a senior, you just talked with a hiring manager and an HR rep, checked off some boxes based on a few blurbs about what you did at prior jobs, and boom, you were a senior. Same thing with getting a raise. To get a raise to 150K (just as a randomly selected round number), it was an enormous bureaucratic process going up 5 levels of management that would take months with lots of aggravation - and unlikely to really pan out. To get hired at 150K for the same role, it was a couple phone calls with an HR rep and hiring manager - about 15 minutes in total.
Retention costs include salary and other benefits, which are ongoing, recurring costs and can add up to quite a bit… so I would hardly say they aren't spending on retention.
Obviously they're a resource that needs paying, they exist, but overall I don't think companies notice people for people until they announce their intention to leave
Companies fundamentally want* to be able to treat software developers as completely fungible assembly line workers. The company's dream is to be able to fire and replace all of the software developers at any moment and lose little by doing so. This partly explains the heavy emphasis that tech companies place on process, project management, and so on even if the developers dislike all of those things. The company must pay for recruitment - there is no other choice because the company must get employees in. However, to pay for retention would contradict the fundamental company dream of making all of the software developers completely replaceable. Basically, retaining the employees is not one of the more significant company goals. Companies' actual goal usually is to make it so that the developers are as fungible and replaceable as possible. Companies want to be able to treat the engineering team kind of like an abstract arbitrarily scalable asset in a video game. Want to replace all of the engineers? Sure, push a button and it happens - you might lose 10% of this quarter's numbers but that's acceptable. Want to scale your team up and go from having 15 developers to 30 developers? Push a button! Want to go up to 60! Boom, do it - the point is that the company has the process that makes the developers fungible - to focus on retention would be silly, the company's goal is to focus on recruiting and then turning the new recruits into replaceable and fungible assembly line workers who are part of an arbitrarily and abstractly scaleable team. Companies don't dream so much about hiring amazing developers - they dream more about setting up a process that will let them hire developers at scale and plug them into a money-making machine that keeps going no matter how retention is going. As a developer working for a company, the end-users are not your real customers. The management and all of their politics are your real customers. What they want is to not think about you. If I go to a restaurant and order some food, generally I do not want to think about how many people are working there or what exactly they are doing - I just want my food. Same with companies and developers - what they want is to just be able to turn the development team into an abstract source of software development. Focusing too much on retainment would tie them too much to a given set of developers. They do not want that. What they want is to set up a process that lets them treat the engineering team as an abstract and arbitrarily scaleable machine that creates software.
*When I say that companies want something this is of course anthropomorphizing a process but what I mean is that the overall totality of the people in the company operating according to various incentives lead to things happening that might in short be described as the company wanting certain things.
Average performers are valued but replaceable. They probably will see nominal raises over time. Minimum effort performers are just a drag on the team, as they occupy a spot that might go to a better performer.
While it costs to hire and train someone new, you also get new ideas and possibly someone better than average (if the company has a good hiring process).
If all those things aren't working and an IC has gone to the bother to get an offer somewhere else than all those things have failed.
Why don't companies just match whenever an employee gets a higher offer? There is a risk of uniformly increasing costs beyond that single employee, there is a quite real possibility that the current company evaluate the employees value lower than the offering company, and there is also a widespread and not entirely unfounded belief that an employee that has already gone through the bother of interviewing elsewhere is likely to have other issues beyond compensation and they are likely to remain a high risk of leaving, so the company will risk establishing a norm of higher compensation across the board while being unlikely to retain the person long term.
The truth if companies in our industry all spent "more" on retention you'd just see salaries and compensation rise across the board. Given the high pay in our industry, I'd say that's already happened/happening.
The older you get, the more appreciation you have of how many and how deeply the misconceptions about how to run companies are and go.
Retention of good people is very very important. But also the nature of that retention: you don’t want them in prison, you want them to be in paradise.
Manage people like beekeepers manage bees: they are always free to fly away, but they don’t want to, because this hive here is so nice!