Well, first thing I’d say is that an accuracy of 88% is great by itself, so there is nothing to be concerned about! Many language learners aim for an accuracy of 85%+ with SRS as a good balance between the time spent and retention. The retention curve for accuracy is exponential, meaning it requires more and more time to continue improving it beyond the point of diminishing returns. At the end of the day, when all is said and done, my accuracy is high because I spend a lot more time than just WK which means I’m really not being as time efficient about it as I should be.
With that out of the way, if you still want some tips that might help, I made a post a while ago with some.
Since then, I’ve also started putting one of the items from each pattern of use for each new vocab item into Anki, so I have a lot more exposure to the readings than just WaniKani. Comparing the stats from the linked post to the ones in this thread shows my reading accuracy has gone up as a result of that.
That said, it’s not a recommendation I would make for everyone in general because it does increase the work load quite a bit though since it basically means every new vocab item on WK ends up with another average of 3 items in Anki. Given most days are 8-10 new vocab items, it’s an additional 24-30 new cards per day in Anki. It all adds up pretty quickly.
Anki’s FSRS algorithm has made it fairly sustainable for me in terms of the number of reviews, but it’s really close to being too much when adding that to the other things I do and I have been considering dropping it back down to maybe only adding two patterns of use for vocab item.