How to read this post

Each prediction has a confidence level: HIGH (we would bet on this), MEDIUM (likely but not certain), LOW (plausible, one-in-three). Forward-looking predictions in an emerging market are inherently wrong; the confidence labels are the honest part.

1. Consumer humanoid prices drop under £10k for at least one model — HIGH

Manufacturing scale and commodity-LLM costs are both still trending down. At least one credible vendor will hit the sub-£10k list price in 2027, even if margins are thin and the spec is modest.

2. EU AI Act enforcement hits at least one skill marketplace — HIGH

The Act is now live; regulators are selecting early enforcement targets. At least one skill marketplace (not necessarily us) will face a public compliance action in 2027. Marketplaces that pre-empted with transparent risk-tiering fare better.

3. A cross-platform skill packaging standard emerges — HIGH

Today every manufacturer has a slightly different SDK. ROS-based standards, an industry consortium, or a dominant vendor’s format — one will coalesce in 2027. GeraSkills publishes against whatever wins.

4. Developer earnings distribution gets more top-heavy — MEDIUM

The App Store pattern repeats: the top 1% of creators earn 80%+ of revenue. Expect a handful of GeraSkills creators to hit six-figure annual revenue; everyone else earns pocket money. This is normal, not a failure.

5. Eldercare category unlocks — MEDIUM

Regulated, high-margin, under-served. The first credentialed eldercare skills pass regulatory review in 2027 and start earning serious subscription revenue. Credentials matter more than code.

6. B2B surpasses consumer — MEDIUM

Warehouse, hotel, hospital, and light-industrial deployments buy more skills than consumers do, at higher prices, with longer contracts. The home market gets the headlines; the B2B market gets the revenue.

7. Bundle pricing emerges — MEDIUM

Monthly “all-you-can-install” subscriptions alongside per-skill pricing — similar to Apple Arcade vs per-app. GeraSkills is architected for both.

8. One breakout skill clears £1m annual revenue — LOW

Not impossible, but it needs a compelling demonstration plus the right viral moment. Candidates: a genuinely-impressive meal-prep skill, a dog-training skill, or an accessibility-focused mobility skill. Most likely the third.

9. Mass consumer humanoid adoption — LOW

Ten years from now, yes. In 2027, no. Early-adopter households with disposable income remain the buyer base. Plan for that buyer, not for the mainstream.

10. A high-profile safety incident — MEDIUM

Probability-weighted, some skill somewhere will do something it shouldn’t in 2027. How the industry responds (transparent review, specific remediation, not-our-problem deflection) will shape public trust for a decade. We have practised our response playbook.

What this means for you

If you are a creator: build for the 2027 buyer, not the 2026 one. If you are a buyer: the catalogue gets better with time; you don’t have to buy everything today. If you are a manufacturer: cross-platform compatibility will not be optional.

Cross-links

See the state-of-the-market 2026, 5 myths. Compliance context via GeraCompliance; transactional action chaining via GeraNexus.