I was discussing with a friend about the context of the place where he works. It is a medium sized company (a few hundred employees) with both "legacy" departments (e.g. monolith applications) and "modern" ones (e.g. continuous deployment, micro-services).

He told me that he noticed the following pattern and explanation:

  • the legacy departments slowly shrink over time, while the modern ones tend to slowly grow over time
  • when an employee from a legacy department wants to leave, it is regarded as a good thing because it helps reach the shrinkage rate. Basically they use the attrition as a "natural" shrinking mechanism.

We have asked ourselves why adopt such a strategy (to allow for "random" shrinkage). Actually our assumption is that this is not actually random since the best employees tend to leave (they go to higher salary positions) and the company loses talent.

Why not trying to transfer talents to other departments before they think of leaving the company and thus avoid losing them?

Question: What is the rationale of using attrition as the main mechanism of slowly shrinking a department (some departments only)?

  • 2
    What do you mean by "allowed to leave"? Isn't employment voluntary?
    – nvoigt
    Commented Sep 21, 2019 at 20:20
  • @nvoigt - yes, I have changed that part because it was indeed unclear. The bottom line is that it is not clear for us the rationale of using attrition as a shrinkage mechanism which heuristically seems to be far from optimum.
    – Alexei
    Commented Sep 21, 2019 at 20:30
  • Do they actually go to new jobs of that hot new technology after not being accepted for a transfer, or do they just leave to continue their legacy technology, but with a different company? Not everybody wants to transfer to the lastest, coolest js-framework-of-the-month team, and while I don't agree, I have seen people that were happy with older tools and technologies.
    – nvoigt
    Commented Sep 21, 2019 at 20:45
  • 1
    @seventyeightist - letting go of low performers sounds better, at least theoretically
    – Alexei
    Commented Sep 22, 2019 at 16:16
  • 3
    @Alexei, the process of removing "low performers" is well-known and is called "rank-and-yank". GE infamously tried it back when Jack Welch was CEO, it kept bean-counters happy for a while but also precipitated the downfall of GE from a mighty technology company to a rather sad financial services operation. Other companies have toyed with rank-and-yank, even microsoft. The practice is almost universally despised by all creative and technical people who care about their craft and not just quarterly performance.
    – teego1967
    Commented Sep 22, 2019 at 20:20

5 Answers 5

  1. At a company that size any amount of layoffs would be a big deal. If the cash flow is there it's not unreasonable to prioritize morale and productivity of the growing teams.

    This can be extra important if teams are geographically diverse and layoffs at a location could start rumors about the fate of the whole office or other satellite offices.

  2. Some companies/leadership teams look at technology skills as commodity skills that you just buy on the market for a given rate. This can be common in cultures or industries that are heavy on "credentialing" and formal qualifications/certifications.

    Or even if they do have some awareness of the vast difference in productivity between 'ok' and 'great' knowledge workers there can still be an attitude that if these people are really 'great' why would they still working here in a legacy department for whatever we pay and not off being expensive consultants.

  3. Layoffs in the team that support a legacy product can cause panic in the customer base for that product.

  4. Perhaps the growing teams actively don't want some of these staff. Maybe they actually have really good compensation packages the other managers don't want to take on. (If you had to choose between one ace Oracle PL-SQL expert or two new grads for your vue.JS project...?)


What is the rationale of using attrition as the main mechanism of slowly shrinking a department (some departments only)?

Firing employees, or laying off employees is a process that no manager enjoys. Each method of separation has costs to morale as well as monetary costs.

Some organizations do this by making a decision that unless the employee is really special if they are department X we will not try and stop them from leaving, and they won't replace them. Of course at some point some will leave because others have left and not been replaced.

Some organizations do this because the long term goal is to transform the company either in scope or in location. They may decide that location X will close next year when the lease ends, so they decide that nobody will be hired to work at that location. This means that when the time comes to close the location the staff that has to be enticed to move is smaller.

You will sometimes see this in government staffs. They are authorized 100 employees but only have 90. They are ordered to make a 15% cut in employees due to budget cuts. The reduction is from the the authorized level, that means that the 10 unfilled slots will never get filled, and the next 5 that leave will not be replaced. As long as they meet the deadline nobody has to get let go.

Sometimes they lose great or potentially great employees, but they have decided that is the approach they will take. They let the natural ebb and flow in staffing take care of the issue.


In your specific situation you are asking why would a company not make efforts to retain 'leavers' by offering them to move to another department which is working with more current technology. There may be a few reasons for that such as a "junior" level person with knowledge of a particular new framework is more useful than a "senior" level person with legacy knowledge but no knowledge of new frameworks. Cultural fit and things like that. Maybe they don't want those "Junior" positions but would rather carry on with the 'legacy' work in another company.

'Legacy' technology often (not always) carries a salary premium if there are not that many people still available to work on it. (e.g. there are relatively few COBOL programmers nowadays and many have retired, but there are still many many COBOL based systems that need to be maintained!)

Here are some possibilities I thought of, as an answer to the more general question of "why use attrition as the main method of shrinking a department [with the implied 'rather than layoffs']?"

  1. Morale - if you actively lay off members of a team you want to 'shrink' it's likely to have a detrimental effect on morale of the people remaining in the company. This could be for various reasons e.g. "will I be next?", perceiving that they will have to cope with more work using less resources as a result of losing people from the company, wondering if the company is in financial trouble and what that means for their own future at this company, etc. Lower morale often brings problems like less productivity, a negative working environment that finds it harder to recruit new people, etc.

  2. (as a consequence of the above) - after layoffs, resignations of other people from the company, who you would have preferred to keep. As mentioned in another answer this would often be the 'best' people, who have the most options that they can readily jump to (at another company). This is probably less likely if the department shrinks 'naturally' as it can be spun as something like "Jane has decided to move on to another role at XYZ industries, and we will see how it goes with the workload for the moment rather than recruit a replacement for Jane".

  3. Following the proper process and administration associated with that - When laying off employees (but not when they resign normally) there are some processes you have to go through which vary by jurisdiction and number of people but may include things like a consultation period, paying out severance, ensuring fair processes are followed (e.g. selecting people based on objective criteria), notification to the appropriate labour authorities, etc. And of course the financial implications of paying severance, time taken to deal with layoffs that would otherwise be used for normal business activities and so on. This one could be reduced to cost.


TLDR - New employees have fewer expectations.

Someone who's been supporting an older technology with the company for 10 years likely gets paid exponentially more than someone who's being brought on at the same junior/non-experienced level solely to become codemonkey for the new technology....

If you brought on an entry level person at $40k per year, and it took them 2 years to become really, extremely proficient with that technology.... or potentially moved that older tech professional to the same position, same timeline, only they're making $120k per year.... in that same 2 year time frame, you're likely getting similar levels of productivity as they both have to learn the new product, however with the new employee, you suddenly have the budget for an extra 1-2 team members compared to the older one.

Additionally, someone who's been there 10 years might get annoyed that they're not being moved up to management, whereas a new hire likely thinks this will take a few years. Employees like to feel like they're moving forward - Changing companies feels like a step forward, and if you're in the same industry vertical, you feel as though you're becoming more of an expert in your field. Transferring to a different technology feels as though you're starting over... it's not preferable.

The people that leave are probably going to other companies that desparately need to find workers to support this older technology, in a market where skilled developers on that platform are shrinking. This means it will potentially become more lucrative for those employees - That's good. That's a step forward for them. They may eventually switch to that other technology.... but only when they realize it will become necessary for their career / sustainability in their industry.

Tech professionals' salaries can fluctuate a bit when you're dealing with highly specialized niche's / technologies... going from being a 10 year expert to entry level is something that can / will happen, but it's usually best, salary wise, to find a way to keep moving forward in your space than to start over.

Of course that person would be a fool not to start learning about other technologies in their free time for the day that they do need to switch


I don't know that there is "an answer", and I suspect there will be votes to close this question as "mostly opinion based". But I'm going to give you my opinion.

In my experience, people stick with legacy products because they enjoy the work, and don't leave until there is either less work, or less interesting work. This decline in new or interesting work also seems to track, more or less, customer demand for the product.

I think the second half of your question is the more interesting one, and again my experience is that employee departures tend to be self-selecting, as you noted. I do believe that your conclusion, that higher-talented employees leave rather than transfer laterally, is incorrect. Employee acquisition ("hiring new people") is a very large expense for companies, but the employee must have the correct skills for a lateral transfer. Employees who have those skills are more likely to be successful than those who don't. Which raises my next point.

One of the things I learned very early in my career is that "specialization is for insects". By which I mean, being highly skilled at an obsolete technology is not a viable solution to remaining employed, and all "high technology skills" are destined to become obsolete. Thus there may be a point where the most highly skilled workers in a legacy department are underskilled in the areas needed for new product development. A savvy developer would begin working on improving their skills in the new product development department while continuing to work in the legacy department. Sadly, it is also my experience that this happens less often than it should, and an employee with high skills in the legacy area is at a great disadvantage even to new hires.

While it may seem inefficient, it does appear to have natural sorting properties -- those who can change, do, and move on. Those who can't or won't, don't, but have the needed skills.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .