If the source file contains images that's what you'll get when you export it to another format.
If you want to convert the text in the images to real text you need to run Text Recognition on it, but that requires using Acrobat.
Because you are referring to a service, I assume you are using Adobe's ExportPDF online service. ExportPDF will run OCR on text in e.g. scanned documents when you convert to Word or Excel, but you have to specify the correct document language, and, this only works if the input file contains a good enough image of your text. If the resolution is too low, or you are working with a bad image, then there is not much the OCR algorithm can do to recover your text.
Most of us here on the forums are users of Adobe's technology, just like you, and we cannot help with subscription related problems. If you want to cancel your subscription, you have to use the instructions here: Return, cancel, or exchange an Adobe order
Another possibility is that the page layout is too complicated and beautiful for you to make in Word. So Acrobat makes it into a picture to keep it.
If you must edit such a file, consider copy/paste of the text and remaking it.