What we wrote. What happened.

How the claims held up

Every judged claim across the 85 posts, by verdict.

Eight things that changed

• The big calls came true late, and by a different route. The blog's 2017–18 predictions happened: coworking supply outgrew freelance demand, branches got smaller and more distributed, and service providers consolidated. But it took about seven years and came through distress and acquirers: COVID, WeWork's 2023 bankruptcy, Yardi buying Deskpass and Hubble, IWG signing a record 1,132 new locations in 2025, mostly in suburbs and small towns. "Sincere collaboration" was not how it happened.
• No numeric freelancing forecast came true. None of "Half the US workforce", "40% by 2020" and "80% of the global workforce by 2030" has come true, and the 2030 one is out of reach. The broad US share went from 36% to 38% (Upwork, 2023). The official BLS independent-contractor share is 7.4%. EU self-employment was 14.2% in 2025, below the 16.1% the blog quoted.
• Generative AI is the blind spot in every batch. No post saw it coming. It reversed the 2018 line that "writers are in high demand": job posts for writing and coding work fell 21% relative to manual work within eight months of ChatGPT, and writing projects on Upwork reportedly fell 32% in 2025, according to Mediabistro citing Vollna's count of Upwork job posts. It also changed language learning and copywriting, and arguably made the 2020 advice to learn coding the most dated on the blog.
• Remote work won, but as hybrid. About a quarter of US paid workdays are now at home. That's up from 7%, even though a December 2019 post said nobody would be working from home "in the upcoming few years". Employers then pushed back with five-day return-to-office mandates at Amazon, JPMorgan and the US federal government.
• Events outlived ventures. COWORK, Coworking Europe and Bits & Pretzels all still run. Most of the ventures the blog profiled are gone. Remote Year, the movement's poster child, was bought by Selina in 2020, then passed to Collective Hospitality when it bought Selina's business, and shut down under that owner in December 2024. COVID halted CU Asia, FreedomXFest, WiFly Nomads and Hubud.
• Many of the spaces in the city guides are still open. The survivors tend to be community-run or backed by an institution. Café hybrids and one-room spaces were the likeliest to go. Several became chains (aSpace 1→6 in Bucharest, Puzl's 8,400 m² site in Budapest). The pioneers shrank (Loffice, The Office Luxembourg), and some spaces that survived dropped or repriced pay-by-the-day desks.
• Governments adopted what the blog called grassroots trends. Rural coworking became national policy: Ireland has 400+ Connected Hubs and France about 4,050 tiers-lieux. Roughly 50 to 70 countries now offer digital-nomad visas, depending on the definition, Hungary among them with its White Card. Estonia's e-Residency, which the Tallinn guide backed, has 144,000+ e-residents who have founded 44,000+ companies.
• The details dated faster than the ideas. Prices, addresses, app names, free tiers and regulations all went stale, while offline, trust-based advice aged best: networking, volunteering, coworking itself, boundaries with clients. One piece of advice is now actively wrong: Bulgaria is not in Schengen zone. It joined fully on 1 January 2025 and adopted the euro on 1 January 2026.

Best and worst calls

A friendly word first. Every one of these posts was written in good faith by people who cared about coworking and wanted to help our readers, with the facts they had at the time. Grading them now is easy: we had eight years of hindsight, the whole internet and a lot of coffee. Nobody could have seen 2020 coming, and if they had, they would have been far too busy buying webcam stocks to blog about it. Most of the misses below are ours anyway. Read it as a toast to everyone who wrote for us, not a report card.

All 85 posts

Sorted by how well they held up. Each post ends with its verdict, claim by claim.

How we checked

Each post was researched in September 2026 against primary sources where they exist: company sites, official statistics, filings and news reports. On 3 November 2026 every claim was checked again: 415 were confirmed, 114 corrected, 136 hedged because the evidence is thin, and 11 removed. Anything that could not be established is marked "unverified", never guessed.

All outbound links were checked on 3 November 2026. Links that now lead somewhere dead or unsafe are struck through; six went to five domains now running gambling sites and were removed. Team profiles were left out because they are about private individuals.

Spotted something we got wrong? Every post has a Suggest a correction button.