Anki
Anki was a consumer AI robotics company behind Anki Drive, Cozmo, and Vector. Its shutdown shows that delightful hardware and real unit sales still need manufacturing margin, cloud continuity, support economics, and roadmap financing.
View original storyProduct snapshot
What it was
Anki built consumer robotics products such as robotic racing cars, Cozmo, and Vector, combining physical hardware with software, character behavior, apps, and cloud-backed features.
Who it was for
Problem / value
It made advanced robotics feel playful and accessible to consumers through animated characters and interactive robot products.
Core workflow
Buy a physical robot, connect it to an app or cloud service, play or learn through interaction, and receive ongoing software-supported features.
Core dependency
The model depended on hardware margin, manufacturing, inventory, cloud services, support, app maintenance, and enough capital to fund a long-term robotics roadmap.
Product form
Pricing model
Axios reported Vector went on sale for $249. Public sources do not disclose gross margins, cloud costs, support costs, return rates, LTV, CAC, or repeat-purchase economics.
Competitors or alternatives
What happened
Summary
Anki shut down after shipping popular consumer robots because the capital required for hardware, software, cloud support, and a long-term robotics roadmap could not be sustained.
Outcome
The company shut down, creating continuity concerns for connected robot owners and showing the capital intensity of consumer AI hardware.
Core risk
A delightful connected hardware product can have users and sales while still failing if unit economics, cloud dependency, support, and long-term roadmap financing do not close.
Timeline
- Anki introduced consumer robotics products including Anki Drive, Cozmo, and Vector.
- TechCrunch reported the company sold 1.5 million robots and raised $182 million according to Crunchbase.
- Axios reported Vector went on sale for $249.
- In April 2019, TechCrunch and Axios reported Anki was shutting down and laying off staff after failing to secure enough additional funding.
- Axios later reported concerns around cloud-dependent robots and continuity after the shutdown.
Before you build
Why it matters
Connected AI devices combine the risks of physical products and software services. A company must pay for manufacturing, inventory, returns, support, cloud operations, apps, firmware, and future R&D long after the initial sale.
Primary check
Before building AI hardware, smart toys, connected devices, or robot companions, prove the full business loop: unit economics, support cost, cloud dependency, repeat engagement, and capital needs beyond the demo.
Checklist
- Are customers paying for durable utility or first-week novelty?
- What is the lifetime service cost per device?
- How often do customers use the product after the first month?
- Can firmware, apps, and cloud services be maintained with existing revenue?
- Would a smaller software-only or accessory-light product prove demand first?
- What is gross margin after all hardware and fulfillment costs?
- What support and cloud cost follows each unit sold?
- What features still work if servers shut down?
- Can the current product fund the next roadmap stage?
- How much capital is needed before the product line becomes self-sustaining?
Relevant if
- You are building AI hardware, smart toys, connected devices, robot companions, IoT products, or cloud-backed physical products.
- Your demo creates emotional delight but requires manufacturing or cloud services to keep working.
- Your current product is meant to fund a much more ambitious future roadmap.
Less relevant if
- Your product is pure software with no device, inventory, return, or cloud-continuity burden.
- You sell hardware profitably without ongoing service obligations.
- You have a clear end-of-life plan that preserves core customer value offline.
Pre-build tests
- Build a full lifecycle cost model before ordering inventory.
- Test paid demand with realistic pricing that includes support and cloud costs.
- Prototype offline or graceful-degradation modes before selling connected devices.
Transferable lessons
- Validate margin after bill of materials, shipping, returns, and support.
- Measure repeat engagement beyond first-week demo delight.
- Design offline continuity for cloud-dependent features.
- Do not rely on today’s toy-like product to fund tomorrow’s much larger robotics vision unless it can stand alone financially.
- Model capital needs before committing to hardware inventory and long-term software support.