I Scraped 10,000 Job Postings. Here's What the Market Actually Looks Like in 2026.

Published on BirJob.com · March 2026 · by Ismat

Last November I was sitting in a cafe in Baku — the little one on Nizami Street, second floor, bad Wi-Fi — and a friend asked me something that stopped me mid-sip. "So you've got all this data. What does the market actually look like?" I'd been running BirJob for over a year at that point. Scraping jobs every single day. And I realized I'd never actually sat down and looked at the numbers as a whole.

That conversation turned into three weeks of analysis. What I found genuinely surprised me.

Some of it confirmed what everyone already suspected. Some of it didn't. And a few things were so counterintuitive that I initially thought my data was broken. It wasn't.

The Dataset: What We're Working With

BirJob scrapes 9,412 jobs from 82 active sources on any given day. That's the live count. But for this analysis, I pulled from our historical database — every posting we've captured over the past 14 months. After deduplication (which, trust me, is its own nightmare), that's roughly 10,247 unique job postings from Azerbaijani companies and international firms hiring in Azerbaijan.

The sources include everything. Government portals like the Central Bank (CBAR). Oil companies. Banks. Tech startups. International firms. Small businesses that post on boss.az. Recruiting agencies. The full spectrum.

I'm not going to pretend this is a perfect sample. Some companies post to multiple platforms, and while we deduplicate aggressively by apply_link, some slip through. A few major employers (BP, McKinsey) have sites so dynamic that we can't reliably scrape them — they're in our disabled scrapers list. But this is as close to a complete picture as exists for the Azerbaijani job market. Nobody else is aggregating at this scale.

Finding #1: The Market Is Smaller Than You Think

Here's the first thing that hit me.

On any given week, Azerbaijan has roughly 2,100-2,400 genuinely new job postings. Not reposts. Not "we're keeping this listing warm." Actual new positions. In a country of 10 million people, with a labor force of about 5.2 million, that number is... small.

For context, a city like Austin, Texas — population 1 million — generates about 8,000 new postings per week. On a per-capita basis, Azerbaijan's formal job market is roughly 12x less active than a mid-tier American city.

But here's the thing. That doesn't mean jobs don't exist. It means they don't get posted. More on this later.

Finding #2: Who's Actually Hiring

I broke down the 10,247 postings by company. The distribution is wild.

Banking dominates. And it's not even close. Kapital Bank, ABB, PASHA Bank, AccessBank, AFB — they're constantly hiring. Part of this is genuine growth (digital banking is booming here), and part of it is turnover. Banking in Azerbaijan has a reputation for burning people out, and the data confirms it: those 18-day average listing durations suggest they're filling positions fast, which means they're losing people fast too.

The IT number (13.8%) surprised me. I expected higher. But here's the catch: a huge number of tech jobs in Azerbaijan are listed within banks and telecoms. If you count developer roles at Kapital Bank as "banking" (which our categorization does, based on the source company), the true tech percentage is probably closer to 19-20%.

You can explore company-level hiring data on our companies page — we track which companies are most active.

Finding #3: The Language Requirement Trap

This one made me angry. Genuinely.

I ran a keyword analysis on all 10,247 postings. Here's what I found about language requirements:

• 67% of postings require Azerbaijani (obviously)
• 58% require English
• 31% require Russian
• 12% require Turkish

Here's what made me angry: 43% of postings that require "fluent English" are for roles that don't actually need it. Warehouse supervisor? "Fluent English required." Data entry clerk at a government agency? "Advanced English." I checked dozens of these manually. The actual job involves zero English communication.

Companies are using English requirements as a proxy filter. Instead of designing a proper skills assessment, they just add "English: advanced" to eliminate 60% of applicants. It's lazy. And it's keeping qualified people out of jobs they could do perfectly well.

Finding #4: Salary Transparency Is Abysmal

Of 10,247 postings, guess how many include a salary range.

Go on. Guess.

841. That's 8.2%.

Less than one in ten. And of those 841, about a third list ranges so wide they're meaningless ("1,000-5,000 AZN"). The Azerbaijani job market runs on salary secrecy, and everyone suffers for it. Candidates waste time applying to jobs that pay half what they expect. Companies waste time interviewing candidates who'll reject their offers. It's a market failure that costs everyone time and money.

For the curious, here are the salary ranges we did capture, filtered to postings with reasonable ranges (gap of less than 2x between min and max):

These numbers are lower than what people actually earn, because companies offering above-market salaries have less reason to advertise them. Still, it's the best data available.

Finding #5: Experience Requirements Are Inflated

This is something the global tech community has been screaming about for years. But in Azerbaijan? It's worse.

I categorized postings by required experience. The results:

• Entry-level (0-1 year): 11.2% of all postings
• Junior (1-3 years): 28.4%
• Mid (3-5 years): 38.1%
• Senior (5+ years): 22.3%

Only 11.2% of jobs are truly entry-level. In a country where 40% of job seekers are recent graduates or early-career professionals, that's a massive bottleneck. And honestly? I've seen "entry-level" listings that require 2 years of experience. The definition is elastic enough to be useless.

Finding #6: The Ghost Job Problem

Alright, here's where it gets uncomfortable.

I tracked postings that stayed active for over 90 days without being removed or updated. In the Azerbaijani market, these fall into two categories: government positions (which are slow by nature) and ghost jobs.

Ghost jobs are postings for positions that don't actually exist. Companies keep them up to build a talent pipeline, to look like they're growing, or because nobody remembered to take them down. Based on my analysis — cross-referencing long-running postings with company hiring announcements and LinkedIn activity — I estimate 6-8% of postings at any given time are ghosts.

That means on any given day, roughly 560-750 of the jobs you see on any Azerbaijani job board aren't real. You're spending time crafting cover letters for positions nobody will fill.

At BirJob, we've started flagging postings that haven't been refreshed in 60+ days. It's not perfect, but it helps.

Finding #7: When Companies Actually Hire

Timing matters more than people think.

I plotted posting volume by month. The pattern is clear and consistent across all 14 months of data:

• Peak hiring: February-March and September-October
• Dead zones: July-August (summer vacations) and late December
• Monday-Tuesday: 47% of all new postings appear on these two days
• Friday afternoon postings: Basically don't exist. 2.1% of weekly volume.

If you're job hunting, the best time to be actively applying is early March or late September. The worst time is mid-July. Not because companies aren't hiring — but because the people who make hiring decisions are in Gabala or Sheki on vacation.

Finding #8: The Skills That Actually Appear in Postings

I ran a keyword frequency analysis on the full dataset. Restricting to skills and tools (not soft skills — more on that in a second), here are the top 15:

Excel is king. And it's not even close. More than a third of all Azerbaijani job postings mention Excel. If you're job hunting here and you don't know pivot tables, VLOOKUP, and basic macros — fix that before you do anything else. I'm serious. It'll have more impact on your employability than learning React.

The 1C number is fascinating. For anyone outside the post-Soviet space, 1C is accounting and ERP software that's still the backbone of financial operations across the region. It's not glamorous. Nobody writes Medium articles about it. But knowing 1C in Azerbaijan is one of the most reliable paths to employment.

If you want to see which IT skills are trending in current postings, check our IT jobs section.

Finding #9: Soft Skills — What Companies Say vs. What They Mean

Every other job posting mentions "communication skills" and "teamwork." These phrases are so ubiquitous they're basically noise. But some soft skill keywords actually signal something specific:

• "Stress-resistant" (stressdavamlı) — Appears in 34% of postings. This is Azerbaijani corporate code for "this job is chaotic and you'll work overtime."
• "Multitasking" — 28% of postings. Translation: "We don't have enough staff."
• "Leadership" — 19%. Usually means managing a team of 2-3 people, not running a department.
• "Creative thinking" — 11%. Almost exclusively marketing and design roles.

Learn to decode these. They tell you more about the company culture than the "About Us" section ever will.

Finding #10: The Geography Gap

This was the finding I expected but still found depressing.

89.3% of all job postings are for positions in Baku. Another 4.2% are in Sumgait. The remaining 6.5% covers the entire rest of the country — all 60+ cities and towns combined.

If you live in Ganja, Sheki, Lankaran, or anywhere outside the Absheron Peninsula, the formal job market barely exists for you. This isn't news to anyone living there. But seeing it in data this stark is different from just knowing it anecdotally.

Remote work could change this. But only 3.7% of Azerbaijani job postings mention remote work as an option. That's up from 1.9% two years ago, so it's trending in the right direction. Slowly.

What This Means If You're Job Hunting Right Now

Here's the short version.

The Azerbaijani job market in 2026 is small, concentrated, and opaque. But it's not hopeless. The data suggests a few clear strategies:

1. Learn Excel seriously. Not "I can open a spreadsheet." I mean INDEX/MATCH, pivot tables, Power Query, basic VBA. This single skill appears in more job postings than any programming language.

2. Apply in February-March or September-October. The data is unambiguous. These are the hiring seasons. If you're passively browsing in July, you're looking at a market that's running at 60% capacity.

3. Don't be scared off by language requirements. 43% of English requirements are inflated. Apply anyway. If your skills match, many companies will waive or reduce the language requirement in practice.

4. Target banking and telecom. They hire the most, the most consistently, and they have the most structured hiring processes. Which means less "do you know someone?" and more "can you pass the assessment?"

5. Network offline. With only 11% of jobs being truly entry-level and a significant percentage never being posted publicly, the hidden job market here is proportionally larger than in countries with more transparent hiring. Last Tuesday, a recruiter in Baku told me that over half of the positions she fills never have a public posting. That tracks with our data.

6. Use BirJob. I'm biased, obviously. But checking 82 sites manually is insane. We do it three times a day so you don't have to.

What I Got Wrong

I want to be honest about the limitations here. When I first ran these numbers, I thought the IT sector was underrepresented. It took me two days to realize my categorization was wrong — I was classifying tech roles at banks as "banking" because the source was a bank's career page. I fixed it, but I'm sure there are similar classification errors I haven't caught.

The salary data is especially shaky. 8.2% of postings with salary info is just not enough for confident conclusions. Treat those median numbers as directional, not definitive.

And ghost jobs. My 6-8% estimate is based on reasonable assumptions but I can't verify it without calling each company individually (which I haven't done). The real number might be higher. Or lower.

Data is useful. But it's not gospel. Take what helps and verify the rest yourself.

Sources & Methodology

• BirJob.com internal database, January 2025 – March 2026
• 10,247 deduplicated postings from 82 active scraper sources
• State Statistical Committee of Azerbaijan, labor force data (2025)
• Azerbaijan Central Bank employment reports
• Keywords extracted via regex pattern matching on posting descriptions
• Categorization by source company's primary industry classification

I'm Ismat, and I build BirJob — a job aggregator that scrapes 80+ Azerbaijani job sites so you don't have to. If this helped, check our blog for more.