He didn’t undermine himself. He’s reminding the interviewer that the app/tech isn’t a magic window that allows him to know how voters voted. The full quote which you excerpted makes clear that he thinks there was a “big role”:

> After we won, we looked back, and Reach accounted for about 10 percent of our total positive IDs, or in other words, pledged voters. Reach was only deployed in the last three weeks before the primary. And the difference in the number of votes between Crowley and Ocasio was about 15 percent. So, I’d like to think that there was an impact, but you can't see who votes for who. I think it played a big role, especially in bringing out folks that had not voted before.

Keep in mind that ACO was seen as a massive underdog to the incumbent. A 10 percent increase in identified pledged voters is huge in a primary.