Tech hiring is getting harder to read at the top of the funnel. AI-polished resumes and generic applications look alike, so a CV tells you less than it did two years ago. Teams that add structure earlier, with a sharper intake and a human look at context, waste fewer interviews.

This hits two groups. Hiring managers see more applications but fewer that fit. Engineers get lumped in with mass-produced profiles, even when their experience is strong. Both lose time when screening relies on keywords.

Application volume used to be a rough proxy for interest. Now AI tools produce a polished resume and cover letter in minutes, so volume says little about fit. The signal-to-noise ratio drops while the screening workload goes up.

Even large platforms say this out loud. A recent post on a major developer platform describes identifying a good-fit candidate at the top of the funnel as "exponentially" harder. That is their own wording, and it fits what we see in intakes.

The numbers in the Dutch market show the same gap between volume and relevance. Get U Hired counted 36,989 unique tech job postings in 90 days, roughly 2,869 new listings per week. Thousands of postings do not mean senior engineers are easy to reach.

A good resume used to take effort. Now it takes a prompt.

AI makes producing applications and matching profiles cheaper. It does not make judging experience cheaper. That judgement still needs a person who understands the role.

On the candidate side, tools help with phrasing, keyword tuning and mass applying. On the employer side, automated screening ranks profiles by keyword overlap. Both sides tune for the same surface, and the surface gets flatter.

The debate is already broader than resumes. Developer communities are also discussing AI-written job descriptions, which means vague requirements meet vague applications. Nobody reads closely on either end.

Our view is that this is a data-quality problem, not a talent problem. The people are still there. The signal about them is just harder to pull out of the text.

The Dutch market is growing and still hard to hire in. More people work in IT, and employers still fail to fill roles with the right skills. Both are true at the same time.

In the Netherlands, 63% of companies that tried to recruit IT staff in 2024 failed to fill at least one vacancy, according to IPMERC's labour-market analysis.

The same analysis puts the Dutch information-and-communication vacancy rate at 5.1%, against 2.4% across the EU27. And 56% of Dutch companies cite too few applications as a reason for unfilled IT vacancies, compared with 43% EU-wide. These are published analytical figures, not interviews.

Another market analysis says it plainly: hiring senior tech in the Netherlands is a sourcing market, not an applications market. The people you want are mostly employed and not applying. That is a published opinion and we could not verify the quote, yet it matches what we see.

So you can have a flood of applications and still have too few fitting candidates. That sounds contradictory. It is not.

A noisy process costs time first, then money, then good candidates. Every irrelevant interview takes hours from your engineers and pushes the vacancy open longer.

Estimates from Searchlab's 2026 statistics put average Dutch hiring cost at about €4,700 per hire, up 12% on 2024. Estimated time to hire is 52 business days for IT specialists and 61 for data scientists. Treat these as averages across many kinds of roles, not as a benchmark for your team.

Senior roles run longer. One market analysis reported 71 days for senior tech in Amsterdam. It also found that remote-only roles attract three times the candidate volume but close 40% less often. More volume, more screening, worse outcomes.

Candidates feel it as well. A good engineer who gets five generic messages and a slow process usually stops replying.

Look at what the person owned, not what the resume lists. A short, structured conversation about real decisions tells you more than a page of keywords.

In practice, a senior recruiter checks a handful of things:

• Did the candidate own an outcome, or contribute code to someone else's plan?
• Can they explain a trade-off they made and why?
• Does their experience carry over to a different stack or industry?
• Does their preferred way of working match the team's reality?
• Do their career moves show deliberate growth?
• Does their communication style fit the influence the role needs?

None of this needs a recruiter to be an engineer. It needs a recruiter who knows the roles well enough to ask the right follow-up. We wrote about the same idea in how AI fits into strategic recruitment: tools can widen the search, people judge the fit.

A small example. Two candidates list the same cloud platform and the same years of experience. One says she migrated a system and cut deployment time. The other says the team migrated it. A follow-up question about who decided what usually separates them in ten minutes.

A structured process makes decisions earlier and with more evidence. It starts with a clear intake and ends with fewer, better interviews.

At Doghouse that is the Delivery Sprint: intake, targeted sourcing, and interviews, with feedback after each step. The intake settles what the role really needs, which skills are negotiable and what the team looks like. A scorecard then keeps every candidate judged on the same points.

Volume sourcing focuses on activity: more messages, more profiles reviewed. Structured recruiting focuses on decisions. That is the difference between a long list and a short, explained shortlist.

It also helps with the practical limits specific to Europe. Language requirements, works councils, immigration rules and employment law differ per country. Qualifying them early cuts late-stage drop-off. The same applies in regulated fields, as in our piece on why cybersecurity hiring is a process problem.

We cannot promise a number here. Our own data covers only our clients, and every market differs. But fewer irrelevant interviews is a result you can check yourself.

AI makes it cheap to produce applications, so the top of the funnel says less about fit. Tech hiring now depends on judgement at the points where nuance decides the outcome: intake, assessment and feedback. In the Dutch market, where senior people are mostly employed, targeted search matters more than volume.

If you are working on a hard-to-fill role, or you are an engineer who wants a recruiter who reads beyond keywords, we are happy to compare notes at Doghouse.

1. Beyond the AI noise: How Stack Overflow is rethinking tech ...
2. Tech Hiring Trends in Netherlands 2026 - Get U Hired
3. 36,000 more people in IT, and the market still cools - IPMERC
4. Recruitment Marketing Statistics 2026 | 80+ Facts & Data - Searchlab
5. AI, JD, and other letters of the law
6. Kelly Rapmund's Post - LinkedIn

Written by our AI, read by a flesh-and-blood recruiter.