If it can be targeted to another market group, do so. If you make, say, tensor analysis software, then your capacity to go after another group is dubious. In this case, it may make sense to go upstream. That is, find out what your customers would use the tensor analysis software for, and offer that service or make that product for your customer's customers.

And now for the best option of all, FAIL!

You and your startup team were mistaken about something as fundamental as the size of the market you were targeting. Please ponder that fact seriously. Why were you mistaken? Emotional attachment to an idea? "Irrational Exuberance"? Is there reason to believe that decisions taken from this point on will be based upon more factual foundations? Sober answers to all of these questions sound good in interviews for new jobs. Sober answers to these questions sound like music to the ears of Venture Capitalists or Angels when one is starting up a new venture.

Fail fast and learn.

Then demonstrate that you did so.

If you are failing anyway, you might as well tell us?