This startup preferred to remain anonymous, so I've obscured some details about them.
The open-source model seems to me to be high risk; ADS could spend a lot of effort on building something great, but find only at the end that users may not be willing or able to convert into customers. I recommended instead that they consider selling consultancy services, something they are already doing a bit of to keep alive. They can sell these services without making a big deal of the product, then deliver the result using the product and get a lot of feedback about it and (they hope) a referenceable customer using it. This won't scale of course but seems a viable means of pursuing customer development.
ADS have found one customer through friends but don't think direct referrals like this are likely to produce enough people who want to try the product, and I agree. Financial services seems a natural data-hungry industry to go after but, as I often tell everyone who will listen, it is a closed club and very difficult to break into if you don't already have a phone full of numbers - and unfortunately A and M do not have that sort of contact list. We talked about many industries but all seemed to require heavy-duty enterprise sales, unlikely to suit hackers like A and M and as they are bootstrapped, unlikely to come from a sales hire either. I would prefer to find a mid-range niche, populated with companies small enough to make buying decisions quickly but still large enough to need the product. I didn't have any bright ideas except that maybe delivery companies would suit - often family-run and relatively small at the head office, but with a lot of lorries and containers on the go and therefore probably data-hungry especially if they have real-time sensor data on delivery locations.