There is always value in having a store of data that is unique or differentiated in some way. The trick is figuring out what that is and who might be intereste

There is always value in having a store of data that is unique or differentiated in some way. The trick is figuring out what that is and who might be interested. For a while I ran product for a social monitoring company and our traditional user base was brands, agencies, etc. who would use our giant database of public content to do market research, etc. At various points we would get inbound requests from someone with a unique ask - I recall: - a military historian working on a government grant who wanted to analyze the social media activities of various militias in a particular part of the world - several pharma companies looking for adverse drug reaction reports online - hedge funds looking for deep sentiment trends in particular areas for perception of certain businesses - some company looking to find properties where women made announcements that they were pregnant. And then there’s always the requests for X but in Y language/country. “There’s a Twitter like service in Bangladesh, can you get that data?”. All of these people had money to spend and specific interests - we couldn’t help most of them as the economics didn’t work out in terms of building a scalable business, but if you can find a niche and run things lean, there’s a real long tail of opportunity there.