Software engineering salaries come from one of three budgets

Original Article ↗ Hacker News Discussion ↗

Software engineers are often paid from one of three implicit budgets: sales/marketing work that directly drives revenue, product R&D that aims at future profit, and maintenance or internal tools that keep existing systems running. Commenters argue this framing overlaps with the classic “profit center vs cost center” divide and shapes how organizations value engineers, affecting pay, job security, and career growth. Many note that maintenance and internal tooling tend to be underfunded and politically weak despite being critical to reliability, while the most attractive roles are usually tied to clear revenue impact or long‑term strategic bets.

Mapping the Three Budgets to Company Types

  • Commenters often map the three budgets to:
    • Consulting / selling hours → sales & marketing budget.
    • Product companies → R&D budget.
    • “Every other company” using software internally → maintenance / operations budget.
  • Many note companies can mix all three streams (e.g., product + consulting + internal tools) and devs may rotate between them.

Cost Center vs Profit Center

  • A recurring alternative framing is: are engineers seen as a profit center or a cost center?
  • Profit-center teams tend to get higher pay, more investment, and tolerance for risk.
  • Critics argue “cost center vs profit center” is conceptually flawed: every department ultimately supports profit, but leadership beliefs and KPIs make the model real in practice.

Career Strategy: Which Bucket to Target?

  • Some insist you should “always” aim for product/R&D or profit centers for pay, impact, and leverage.
  • Others strongly disagree:
    • Consulting (#1) can be lucrative and satisfying for people good at communication and rapid delivery.
    • Internal/“category 3” roles can offer stability, deep domain expertise, and long careers.
  • Fit depends on personality (intimacy with users vs scale, tolerance for politics, appetite for burnout).

Maintenance, Internal Tools, and SRE

  • Many say maintenance and tech-debt work are chronically underfunded, especially refactoring and internal tooling.
  • Others report the opposite: SRE/ops and compliance-heavy maintenance can be well funded and more protected than flashy R&D.
  • There’s concern that treating maintenance as low status creates organizational decay and repeated rewrites.

Consulting Economics and Billing Models

  • Hourly consulting often incentivizes maximizing billable time over quality or reuse.
  • Fixed-fee contracts are debated:
    • Pro: better-align incentives; can reduce adversarial client relations.
    • Con: estimation risk can destroy margins; requires strong scoping and negotiation skills.

Industry Examples and Fuzzy Boundaries

  • Many argue the three categories are blurry:
    • Big firms like Amazon, Netflix, banks, and defense contractors contain all three under one umbrella.
    • Within a single company, some software is core revenue-generating; other software is purely enabling/back-office.
  • Heuristics suggested: ask whether the business would halt without your system; observe who the “rockstars” are (sales, engineers, content, etc.).

Job Security, Layoffs, and Power Dynamics

  • Multiple anecdotes show layoffs cutting across “core” and “non-core” alike; financial targets and investor pressure dominate over local value.
  • Reporting higher up the org chart can provide protection, but only as stable as that leader’s own political position.
  • Being attached to clear revenue or uniquely hard-to-replace skills sometimes helps, but is not a guarantee.

Compensation Trends

  • Several note that very high total comp (~$500k) has become rare outside a few niches (e.g., certain FAANG roles, trading firms, specialized ML/distributed systems).
  • 2024 market is seen as less frothy than 2021–2022, with many companies capping around $200–250k even for senior roles.

Accounting: Capex vs Opex and Tax Changes

  • Discussion touches on capex (upfront, capitalized investment) vs opex (ongoing operating cost).
  • Cloud subscriptions and “pay monthly” models are often favored as opex due to flexibility and simpler decision-making, despite tax nuances.
  • New US rules (IRC §174) now require capitalizing many software development costs, affecting how “R&D” appears in financials.

Legacy Systems and COBOL

  • Being one of few who understand a legacy COBOL system is debated:
    • Some claim it should confer high value; others note pay is usually modest and work is often offshored or “in-sourced” to cheaper locations.
    • Organizations view single-person knowledge as risk and try to reduce dependency via replacement projects or team-building.