Feels like the disruption could be identifying a subset of the 1000s of features used that a cover a large enough part of the market and solely focusing on that.
Instead of 1000s of features, have 40 good features. Limits how many customers you can get, but should lower development costs as well. There's probably a sweet spot that works.
> identifying a subset of the 1000s of features used that a cover a large enough part of the market
Deceptively hard. The only way to make this happen is using telemetry, and users don't like that, understandably. My belief is that Excel and Word are so successful because EVERYONE uses a different subset of those thousands of features. A one-lawyer office will differ from a dozen-person firm, a big general contractor will use different features from a small one, etc.
You can't survey users about those features either. Users don't know what your surveys mean half the time. They will answer according to how they think you want them to answer instead of what they really do. Or they will be satisfied but name features they would never use because they heard about them at a conference once. They use different terminology from you, so signal gets lost either way. Ad nauseam.
Well, rather than telemetry, I suppose excellent domain knowledge may help? Experience in a particular field might help you choose a particular workflow that could work for a lot of users.
Instead of 1000s of features, have 40 good features. Limits how many customers you can get, but should lower development costs as well. There's probably a sweet spot that works.