Yes. Thanks, Ross. I'll answer your first one and then ask Bret to share some comments on the second one. But -- so just taking a step back, Tech & Shopping, obviously, is the challenge, was the challenge in Q4, will continue to be the challenge in 2026. Don't want to lose sight of the fact that the other 4 segments grew nicely in Q4 of 2025, and we believe will continue to grow in 2026. Within Tech & Shopping, the affiliate commerce piece is one of what I would refer to as 3 challenges within the business and worth describing and talking about the other 2 for a moment. So remember, we have the B2B business that's inside of the Tech & Shopping segment. You'll recall that our strategy in 2025 was to intentionally contract revenue at a rate that would be less than the contraction of expenses. In other words, we would cut more expenses than revenues, and we did that. So in 2025, the B2B revenues were down $11 million year-over-year, but the EBITDA was up close to $6 million and positive. So that strategy of shrinking the footprint of that business, cutting out certain products and service lines has worked, but shows up as a revenue drag. I just want to point that piece out. The last one is the published -- the Game Publishing business that's still a residual business that stayed within Tech & Shopping, which Bret pointed out, we sold, we're out of. That was like a $14 million, $15 million -- $14 million year-over-year bad guy in 2025 as well. So those are just 2 things to just point out as we think about '26 versus '25 that as we lap these things are going to be beneficial. But then yes, look, I think the belief that the pain that we're seeing on the affiliate commerce side in Tech & Shopping will start to improve in the second half, both because of comps as well as other initiatives, again, video monetization, licensing, building out traffic in both the RetailMeNot app and browser extension. And that collection brings the overall challenge of Tech & Shopping to being sort of more of a -- from a full year point of view, kind of a low single-digit decliner, but still a decliner.