Fireside Chat, Interview
A Conversation on Hard Tech with Eric Migicovsky
- Eric von Hippel is a YC partner and former founder/CEO of Pebble, which was one of the first hardware companies to join Y Combinator in Winter 2011.
- Pebble originated as "Impulse," a watch for BlackBerry devices; the name was chosen from a thesaurus, while "Pebble" was a pre-existing good name Eric had noted in Evernote two years prior.
- In the pre-YC phase (2008–2010), the team built the first 10 watches manually in a garage with a $15,000 budget, sourcing circuit boards from China and metal cases from a local CNC shop.
- Early prototypes failed due to RF interference (Faraday cage effect); the team iterated quickly by switching to plastic backs and adding foam to prevent electronics from exploding during shipping.
- The team continued one-by-one manufacturing up to 700–800 units to maintain iteration speed and avoid catastrophic quality failures across large batches.
- Eric highlights a critical error post-YC: investing $250,000 in inventory for 3,000 units with a contract manufacturer, which failed to launch for 6–8 months; during this delay, customers switched from BlackBerry to iPhones, leaving the company with unsold stock.
- The "frugality trap" is identified as a major startup pitfall: teams often become scrappy when cash-poor but lose discipline and speed after raising significant capital.
- Crowdfunding serves two distinct strategic purposes: (1) an early "validation play" raising $25k–$50k via simple messaging to fund a prototype, and (2) a later "sales/marketing channel" after achieving product-market fit (as Pebble did with 1,500 pre-existing customers).
- Eric advises against high-production crowdfunding campaigns early on, as they often fail to ship; instead, founders should use niche communities (Reddit, Discord, forums) for early validation.
- For hard tech and deep tech, the definition involves sectors where iteration cycles are slow due to hardware, fundamental scientific development, or regulatory constraints (e.g., medical devices, aerospace).
- MVP definitions for hard tech vary: for medical devices, it is peer-reviewed science; for aerospace, it is a scaled representation or demonstration of technology with early customer interest.
- YC's standard advice for hard tech is to prioritize sales and customer discovery before perfecting the prototype; a recent batch company secured $4–5M in sales contracts within three months by pivoting to customer conversations early.
- University researchers are advised to leave the lab environment to attend corporate conferences and speak directly with potential customers to test commercial viability, rather than relying on academic grants or long-term R&D cycles.
- Eric cites Relativity Space as an example of a "crazy" idea (3D-printed rockets) that succeeded by focusing on the unique value proposition of zero integration time through printing entire rockets in one piece.
- For software engineers entering hardware, the recommended path is to buy an off-the-shelf product that is 50% complete (e.g., via Alibaba or Amazon) and "hack" it to add their specific software or modifications, rather than building custom hardware from scratch.
- Founders pre-product-market fit should locate themselves closer to their customers rather than suppliers (e.g., Shenzhen) to accelerate software iteration and validate problems; moving to manufacturing hubs is only recommended once sales channels are established.
- Building a competitive moat is best achieved through network effects or data lock-in; patents are generally ineffective for 95% of startups due to high enforcement costs and the ease of circumvention.
- When targeting marginal markets, founders should start small to cheaply identify and validate a specific user base who will provide rapid feedback, then expand to larger markets with a clear roadmap.
- Pricing strategies should be based on the value delivered to the customer (cost savings or revenue increase), with experimentation recommended to find the optimal price point before scaling.
- Hiring for early-stage hardware companies should prioritize trust, reliability, and versatility over deep domain expertise to maintain speed; hiring a team of seven via friend networks in four days is cited as an effective strategy.
- Inventory management is a primary cash flow risk; founders should mitigate this by using pre-orders to fund production or by paying higher component costs (BOM) to reduce lead times and inventory holding periods.
- Eric suggests that "hardware-as-a-service" or subscription models work only if they align with the specific business model and customer willingness to pay; forcing a subscription model where a one-time purchase is preferred can hinder growth.
- For complex hardware concepts like holographic AI therapists, the advice is to overspend on existing off-the-shelf tech (e.g., HoloLens, Oculus) to simulate the user experience and validate the concept before investing in custom hardware manufacturing.