Everyone loves to talk about innovation these days. Big companies are setting up “innovation labs” in droves. Startups in particular love to innovate. After all, who doesn’t love to innovate?

Innovation is generally a bad idea.

Innovation is hard. Really hard. Innovation requires lots of money, resources, time, and luck. Startups have none of the above. So startups should innovate as little as possible.

Most startups should only innovate in their product offering to serve their customers’ needs, and avoid all other forms of innovation. Moreover, since most startups are application layer companies (and not deep tech companies), they should abstract all material engineering challenges. They should focus exclusively on application layer problems - UX, features, stability, etc.

Outside of product innovation, startups should copy as much as they can. There is generally no need to innovate on any of the following, unless that particular area is the focus of the startup and the management team has deep expertise in the area:

Legal structure
Job descriptions
Invoicing and accepting payments
Internal communications
Investor relations
Org charts
Goal setting and planning
Enterprise selling
Inside selling
SEO/SEM
Sales training
Cold calling
Onboarding
Cap table management
Accounting
Budgeting

You should seek to actively commoditize every aspect of your business that isn’t your locus of innovation. Everything else should be emulated.