Conferences mean very different things in different fields.
In some fields (e.g. most of pure mathematics), they are essentially just a venue for presenting/publicising your work to other researchers. In such fields, conference abstracts are not treated like full publications; there is no strong expectation of originality. So you can re-use your figures freely between conference talks/abstracts, and re-use figures from journal publications in conference talks/abstracts.
The only thing to worry about, in this case, is that if you re-use material too often, then people who see it multiple times may get bored of it, and think you are being lazy and not producing much new research. But this is unlikely if you are just re-using one or two figures, provided your talks/abstracts contain substantive other fresh material.
In other fields (e.g. much of computer science), conferences are a publication venue in their own right, on a par with journals. In such fields, there is an expectation of originality; so you should re-use figures sparingly, and usually explicitly note when you do so. Essentially, in this case, treat it as you would re-using material between papers. So it’s not necessarily a bad thing, especially in e.g. a background section, but it won’t count towards the original content that a referee will be looking for. And the re-use should be noted explicitly and prominently — e.g. “The relationship between fuzzles and woozles is shown in Figure 3 (reprinted from Jones 2014).” — unless possibly the figure is clearly just a very simple and standard illustration of background material.