5% engagement on abandon seems decent. As you said this can vary wildly dependent on industry or consumer behavior. Is your survey optimized for a single answer? Or is it an invitation to answer multiple questions? A best practice is present one question with expected/common answers, but with option for custom.
A Beginner's Guide to Abandonment Surveys - SaleCycle
Another option is soliciting deeper engagement with those who do. Consider an incentive to participate in a short interview. You could make the case that fewer but higher quality responses can be as representative as a larger sample.
Best wishes.

Thank you so much, Colby. I will take a look through this guide.
We do currently have the survey as a single-select question and we have the question auto advance once the respondent has answered; we theorized we might be losing people who were not submitting their answers.
I did go back and look at engagement from a survey completes perspective (survey complete per impression) which knocked us down to an actual response rate of roughly 2.5-3%.
Thank you so much for all of your help!
Christie

That's a really good industry resource - what a wonderful find by ColbyA