Do you make money per client? Why are you artificially constraining that? Growth != loss of trust reputation if you have a good product that your users trust and want.
We're learning. Move fast & break shit doesn't work on Wall St. We knew how to build the best financial sentiment NLP in the world, but we had no clue about the space we would eventually sell into. It's cautious, slow to act, skeptical, very sophisticated and secretive. We had to find a client who knew exactly what to do with our data without any handholding and was a bit forward thinking, even experimental in that they take chances on startups. We found that perfect client (actually they found us!) and they could see the value of what we built right away, probably better than we could. They loved us so much they eventually invested in us so we knew we were on the right track and they've been helping us ever since. Growing client #s required us to become experts in a number of different areas so we took our time getting them right.
Did they also sign a paying contract with you? That's the best of all :).
I agree with you that you can't just break things when you're working with financial firms. That doesn't mean you have to artificially constrain growth. It's probably a matter of degree, and I don't have a good enough sense of where you sit to have a more informed opinion.
How are you actually constraining yourself? Are you keeping out users who want in? Are you limiting the number of companies to whom you are trying to sell?
No we are not keeping out clients that want our data. But there are precious few clients that proactively seek data sets like our first client did with us. We are however (up until very recently) retraining outreach. We needed to encourage proper academic research which took time to acquire, create solid marketing collateral which took time to get right and even understand what clients wanted to see, we needed to harden our infrastructure, improve the technology, and we needed to understand what client concerns would be and how they could use our data. It's like having electricity, knowing it's valuable, but not knowing how to get your clients to use it to turn on a light bulb. If we screwed up at any stage it would burn our reputation. The sentiment space is similar to the blogging space. Literally almost anyone can hang out a shingle, use generic software and start a sentiment business. Just like blogging. It is, however, extremely hard to get right and our clients know this.
