Two Over 35 in a 38-Person Front-End Team: The Age Cliff and Its Survivors
The red line is a cost calculation: a 10-year front-end at 30-45K monthly is replaceable by two 18K juniors who are also more obedient and overtime-tolerant, which is why survivors all hold a niche so deep or a business so specific that losing them breaks something. Western teams should read this as a warning that age vulnerability tracks perceived business value and barrier to entry, not raw seniority.
The department audit is blunt: 36 of 38 front-end developers are post-95s, the youngest born in 2004. The over-35 count is two, and the only other survivor is a 37-year-old who switched from testing to Vue on business familiarity alone. The article maps where the rest go: management (a title-only change in most cases), layoffs (a hidden HR red line at 35), career shifts (product, operations, milk tea shops), or emigration.
Why does front-end age so badly? The article points to two structural forces: managers see front-end work as "drawing pages," so even a 3-second-to-1.2-second LCP optimization earns zero business credit, while backend architecture and algorithm gains get visible praise. And the entry barrier is low: a three-month trainee can ship a Vue + Element Plus admin, a six-month self-taught PM can compete, so supply is endless and replaceability high.
The survivors share three traits: a deep, non-replaceable technical niche (low-code engines, WebGL, cross-end rendering, front-end infra), business knowledge deeper than product managers, and a calm acceptance of the rules. The author's own counter is to keep adding new lines to the resume each year and to write articles as a self-audit tool, but the closing is unsentimental: age unfriendliness is the rule, and the only defense is gathering trump cards before the screening mechanism triggers.
The article frames age discrimination as a rational ROI decision: replacing a 10-year dev with two cheaper juniors is presented as normal business arithmetic, which makes the market's cruelty seem inevitable rather than a policy failure.
The misalignment between technical investment and business perception is structural: front-end optimization is invisible to managers, so it never builds the political capital that backend or algorithm work accumulates, and that is why front-end loses its bargaining power by 35.
The claim that anxiety peaks at 28-29 suggests the industry's age bias is anticipatory: people internalize the cliff before it hits, which explains why the survivors at 35+ are calmer than those approaching the threshold.
The survivors' common trait is not raw skill but irreplaceability, either through a deep technical niche or through owning a business-critical domain, implying the market does reward specialization even if it punishes generalists.
The author's own response, learning a new technology each year and writing to expose gaps, treats age survival as a continuous portfolio management problem rather than a single career jump.
Front-end as a standalone role is dismissed by one position, which insists the field has already become full-stack development. Company context counters that, with a hardware-driven firm retaining older engineers for software maintenance, and a 39-year-old teammate becomes a benchmark — reaching 40 in coding is widely regarded as a win. The split is between seeing age anxiety as intrinsic to front-end versus situational, tied to the employer's business model.
Bro, you live in your own mental model. Where is there still front-end nowadays? It's all full-stack development now~
Yes.
Our company is mostly 30+, including one Java old-timer born in ’89. Most of the software developers are 32 and up. Of course, our company doesn’t make money from software; it’s all hardware. So if the boss wants to cut, it would be an all-or-nothing cut: either drop software entirely or keep people for maintenance. Right now we’ve kept 4 front-end devs, 3 iOS (one moved to product, one to CFO assistant), 3 Android (one moved toward hardware development), about 5 Java, and the rest are mostly embedded software and hardware.
There's a 39-year-old guy on the team, and everyone envies him. The general understanding in the team is that doing programming until 40 is already pretty good, even considered a win.