The usual mantra is to 'try harder'. Trying harder is impossible when you're already trying as hard as you can.
But you can always try different.
Years ago, I was creating trivia questions for a product we built for Prodigy. We had a 99% accuracy rate in doing the questions. Which was great, except there were 1800 questions in a batch, which meant 18 wrong each time, which was totally and completely unacceptable. These were honest mistakes, made by smart people working as hard as they could.
No matter how hard we tried, we couldn't do better than 99%. So we switched our system completely and did it in a totally different way. Same number of people, same number of hours, 100% accuracy.
If it's not working, harder might not be the answer.