Loading today’s stories
Build1 distinct publisher3 min readUpdated
A dev.to survey puts European median total comp at €115,000. PayScale's German average is €54,006, and even the like-for-like German comparison leaves €35,994 unexplained.
The Engineer · Build desk
Compiled by The EngineerSomething wrong?How this is made
Take the comparison the survey stops short of making. PayScale's German average is €54,006 [5], and the German respondents in this sample report a €90,000 median [15]. Like for like, one country against itself, that is €35,994 apart [1]. Set the full European median against the same PayScale number and the gap widens to €60,994 [2], but that version mixes countries and should be thrown out. The narrower figure is the one a hiring manager in Munich would have to explain.
Neither comparison is clean. One side is an average, the other a median, and the survey's figure counts realistic bonus and equity on top of base [2] while the aggregator numbers are quoted as salary [5][6]. The UK reference is harder still to reconcile, since it arrives in sterling and rests on a sample that was 88% mechanically biased [cuk]. That statistic is the entire argument: the public data is describing the mechanical, hardware and electrical version of the role, which is what the phrase meant for four decades before software borrowed it [4].
The €45,000 in the headline is a third measurement, and it is internal to the survey [cswing]. Design Engineer on the contract carries a €115,000 median and Frontend Engineer €70,000, with Creative Developer at €50,000 on seven responses [7], among people who described their work before anyone looked at what their contract called them [11]. Fifty-eight of the 106 are on some other label [5].
The author is candid about the limits. The title gap holds when experience is matched, cannot be tested inside a single country because the cells get too small, and may partly reflect which kinds of companies use which titles [12]. The €25,000 spread between the code-leaning and design-leaning groups [17] sits on top of a code-leaning group that contains more senior people, and the survey says plainly that it cannot separate the two effects [c17b]. Pay steps up once at around four years and then flattens [16], so seniority is doing some of the work the title appears to be doing.
Watch what €70,000 does in this data. It is the Frontend Engineer median [7], the median for people at a company that is not a startup, which is below pre-seed [18], and the French median [15]. With cells of ten to twenty-four people, those groups plausibly overlap, and nothing in a 106-person sample can tell you whether the market is paying for the word on the contract or for the sort of employer that uses it.
The number is still better than the alternatives for anyone hiring the software version of the job, for the plain reason that it counts the right people [1]. The weakness is recruitment: a survey circulated among people who identify with a term will find the ones doing well under it, and the write-up concedes it is not the whole market [19]. What it has done is price an occupation the salary sites never counted. It has not yet priced the job.
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
An anonymous survey of 106 design engineers across Europe asked what they earn, what their contract calls them, how they work and what stage of company they work for.
The survey reports €115,000 median total annual compensation, defined as base plus realistic bonus and equity.
The middle 50% of respondents earn between €70,000 and €140,000; half earn above €100,000 and a third above €130,000.
PayScale puts the German average for design engineers at €54,006.
SalaryExpert gives €79,419 for the Netherlands.
A UK survey published this year lands on a £49,489 median, with 88% of its respondents mechanically biased.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Single self-selected survey, transparently caveated
All figures trace to one anonymous, self-selected survey of 106 European respondents (102 with a comp figure) published by the party that ran it. The author is unusually transparent — sample sizes on every cut, band-midpoint medians disclosed, small cells flagged directional — which raises the floor. But cross-tab cells run as low as n=7, there is no recruitment disclosure, no significance testing, no independent replication in the cluster, and the aggregator counter-figures are quoted without dates or methodology. That supports the descriptive numbers as reported while leaving the causal and comparative claims unresolved.
No adoption surface
This is a compensation survey, not a product, standard or release. The supplied source reports no deployments, releases, benchmarks, pricing or licensing events, and no evidence that any employer, vendor or aggregator has taken up the survey's numbers. Inferring an adoption level would require facts the sources do not contain.
Headline overstates a caveated small-sample result
The framing runs ahead of the data in two specific ways. First, 'Your job title is worth €45,000' converts a difference between two band-midpoint medians (n=48 versus n=12) into a causal price on wording, while the body concedes country and company-mix confounds it could not test. Second, the dek's €35,994 'unexplained' gap and the €60,994 Europe-versus-Germany comparison set a total-comp figure including bonus and equity against an undated aggregator average for what the article itself says is a different job. The underlying observation — that public design-engineer benchmarks may be pricing a mechanical role — is genuinely useful and honestly caveated in the body, which keeps this well short of pure hype.
No disclosed backing or commercial interest
The source is a self-published survey on a community platform. It names companies that hire for the role (Linear, Vercel, Stripe, Figma) and positions itself against commercial salary aggregators, but the supplied material discloses no sponsor, employer, recruiting business, product or funding relationship behind the survey. Scoring an incentive structure would mean inventing facts about who benefits.
Low — one unreplicated self-reported dataset
Confidence is limited by structure, not by internal inconsistency. The article is legible and self-caveating, so what it says it measured is clear; but a single self-selected 106-person sample, band-midpoint medians, subgroup cells down to n=7, one publisher, and no independent dataset in the cluster mean the descriptive medians are believable only as a directional first data point and the causal title claim is not established.
build
Three API calls, no rollback: the phone numbers your carrier bills and your database never saw1 distinct publisher
build
A Stripe SDK Major Bump Turned One Metadata Lookup Into a Silent Non-Delivery1 distinct publisher
build
Three ledgers, one payment: reconciliation breaks because the model picked a winner1 distinct publisher
build
A 20-digit ID went into a JSON repair tool and a different number came out1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
dev.to
1 article · August 24, 2026