Everyone's all Shopify screenshots and growth percentages. Nobody posts about customer service. But CX is one of the most important parts of building a brand today, and it's changing faster than almost anything else we do. CX was rolled into marketing at Ridge 24 months ago. And year to date: Orders are up 50% First reply time down 77% — 22 hours to 5 (email & contact form) All-in CX cost per order cut in half, to <$0.30 CSAT up to 95%, a 17% gain in three years All while headcount has stayed flat. We've thought about it in a few steps — in this order — so we can get the fundamentals right before scaling with AI. 1. Self-service first. Most people don't *want* to have to engage with CX. So we invested first in giving customers the tools to resolve their own tickets faster: better onsite FAQs, dedicated warranty portals, and improved self-serve returns and exchanges. So while orders have grown 50%, tickets have only increased 10% and conversations per order fell 25%. And of the contacts that do come in, 89.5% resolve without ever reaching an agent. 2. AI-enabled agents. Before anything got fully automated, we wanted to make our current agents better. That comes down to documentation — AI assistance is only as good as the specs, FAQs and SOPs you feed it. It's why I moved us to @Richpanel two years ago (not a sponsored post). Their 2.0 platform launched in Q1 with the intelligence-first tooling that will not only eventually power our support, but immediately equip our current agents to be more effective. 3. AI answering tickets end to end. The final form of CX is obviously tickets answered autonomously, and Q3 was the unlock. Today almost 40% of tickets are handled outright, at a 93.6% satisfaction score against 94.2% for our human team. It took a lot of iteration before the AI could answer tickets effectively and without the really annoying AI interactions that plague most CX touchpoints. There's still a lot to improve, and I wouldn't call any of this industry-leading. But the path has led to real efficiency gains and, more importantly, happier customers.

Sweepstakes is one of our fastest growing campaigns at Ridge. With Tony Hawk onboard and over a half-million in prizes, we drove ~50% revenue growth YoY at a better MER — making it our first multi-8-figure sweeps campaign. The budget breakdown is below — and a few thoughts on what drove the channel shifts: 1/ The celebrity endorsement unlocked some new scale for us, with TV most notably gaining budget share. We scaled TV +71% YoY (exclusively linear) at roughly half last year's in-platform CPA — largely from significantly stronger response rates. Attribution is difficult with TV, but we validated performance with seeing "TV / Streaming" responses increased proportionately as a % of total responses. 2/ Meta gained share of budget (40% → 47%) as we spent 35% more YoY at a 14% better ROAS. I predicted earlier this year that Meta would gain share of budget for brands as advertisers embraced creative volume, andromeda, new attribution models, etc. We didn't see it in Q1 — and we'll see whether it continues through EOY — but interesting to see it take shape in this campaign. 3/ Improved Meta performance largely came down to creative diversity: partnership + affiliate ads. Not shown in the budget breakdown, but Sweepstakes 2025 saw ~3% of our Meta spend run through partner pages and 0% come from affiliate content. This year it was nearly 40% combined — creator and partner ads converting at a higher ROAS than our baseline account. All in, we launched over 2,200 ads during the campaign, +50% YoY. 4/ TikTok had the largest YoY % of budget increase (4.1% → 9.1%) as we basically doubled our investment in GMV Max. ROAS is still GMV-inflated so I'd take the improvement with a grain of salt, but this is the trajectory I was hoping for when I said TikTok was one of our biggest opportunities at the end of last year. Creators posted 5,269 videos on TikTok during the campaign — actually down from June–July, as we better aligned incentives around quality over raw volume. To the previous point, we are shifting some of our affiliate efforts to be more Trybe-based as we more directly incentivize creators to generate high-performing Meta ads vs TikTok Shop sales/GMV. 5/ The trims: Snapchat down ~75%, and X basically went to zero. After a strong run on X last year (and a prediction we'd spend more this year), we're letting performance set the budget — and those dollars went to the channels that earned them.