What should we do when an abstract gets accepted for a conference, but the material is not ready?

Someone has to say it: simply withdraw the poster presentation. Of course, no one wants to do this. But there is nothing wrong with simply yielding to reality: things just did not work out as expected. Then, next time you submit an abstract, bear this one in mind! Best of success with future submissions!

From my point of view - you already submitted an abstract so let it be (if your results still agree with the content). On your poster collect all the information that you have in the way that you can explain the idea and steps that were done. Leave the space to describe an outlook with everything what wasn't done yet and what you want to achieve.

Conferences are necessary for the scientific exchange. You will get one along with new ideas. Maybe you will hear some novel approaches that can boost your work and you can finish these results faster.

You may not have a complete set of results, but presumably you have a clearly defined research question and method. You may have already encountered and learned things along the way. Perhaps you have preliminary results. It could be an option to present these and look for feedback on your research question and approach and perhaps even get valuable input to improve your research. One of the things that is so appealing about conferences is to see updates about progress in various research projects, even when it isn't yet ready for publication in a journal.

Unless you intimated in your abstract that you had specific results which you don't have, or you do not have a research question or method, I would personally still present. Especially a poster.

I suppose that standards differ, but in my experience (CS) posters don't need to be all that refined. Unless your standards are different, I suspect that you could put enough together to inform people about your project, its current state, and its direction, if not the conclusions that aren't yet ready.

Many posters are "work in progress", rather than completed.

Alternatively you can withdraw if it is impossible to do enough to meet your own standards.