We just closed out the books on 2025, and the platform performance data is kind of wild.

TikTok went from $16 CPAs in January to under a dollar by July, then climbed back to $9 by December. That’s not a trend line, that’s a roller coaster.

If you had your entire budget there in Q1, you were bleeding. If you pulled out completely in Q2, you missed the recovery.

Google hit $16+ in November and December. Meta stayed boringly consistent around $8-12 all year. Reddit barely moved from $1.50-$3.30 range and nobody talks about it.

Here’s what matters if you’re scaling a DTC brand or managing a portfolio:

You need at least 3-4 platforms running simultaneously.

Not because diversification sounds smart in a deck, but because any single platform can spike 2x in cost with zero warning. We saw it happen twice this year.

Your Q4 budget models are probably wrong. Google costs jumped 40% from October to November. If you’re doing holiday planning based on Q3 numbers, you’re already underwater before Black Friday hits.

The platforms everyone obsesses over (TikTok, Meta) are not necessarily your best performers. Reddit delivered the most stable, lowest-cost results all year and I’d bet most brands aren’t even testing it.

For PE firms: if your portfolio companies are heavily concentrated in one or two channels, that’s a portfolio-level risk.

Platform volatility is no longer an edge case, it’s the baseline. Your brands need the infrastructure to shift budget weekly, not quarterly.

The winners in 2026 will be the ones who can move fast when the numbers change. And the numbers will change.

I encourage everyone who reads this to share it with someone who didn’t believe you when you said this was happening.

The proof is in the pixels.
…more