Over the last few days, I spent time testing through HubSpot’s Prospecting Agent(beta) use case. The exciting part is not just the agent itself, rather what happens when AI sits on top of a clean RevOps foundation and starts working with the right context. What I worked through was not just “turning on a feature.” It was building the foundation that makes the feature useful: + Cleaner & Enriched data in segments (use smart contact & company properties smart properties are great for answering questions such as - Is this company still operational? - Have they merged or been acquired?) - Are they using the tech stack we care about? - Are there fresh research signals that actually make this account worth attention? +Deeper AI context (setup of context & knowledge vault) +Better-structured plays. That’s where this gets exciting. If you set up the Agent well, it’s not just “another AI feature.” It becomes part of a compounding GTM system that can get smarter, cleaner, and more commercially useful over time.

@HubSpot add-on in Marketing - AEO (Answer Engine Optimisation), pretty impressive and simple start with AI marketing-ops from an operator's perspective. And add-on next SEO, #AEO would be a window for realtime impact tracking, adjustment suggestions and reporting. I'm sure there is much more in depth, but with a quick test and play highlights key AI enablements in nutshell : + Tracking brand visibility score across AI engines (@ChatGPTapp , Gemini, @perplexity_ai ) + Shows sentiment are mentions positive, neutral, or negative? + Surfaces which prompts are triggering your category and whether you're in the answers + Reveals which domains AI is citing in your space (your competitors likely already know) +Turns all of that into prioritized, actionable recommendations Why it matters operationally: AI visibility is becoming a pipeline metric, not just a branding one. This would one way to connect the dots of direct or indirect impact of AI citations on your brand, and sentiments at the earliest stage of the buyer journey. Before intent. Before your SDRs ever get a shot. (attached- test views)

Operator hat-on : the pattern I'm tracking that teams build a lead/ engagement score, feel good about it, and never touch it again. That's not a scoring model That's a false sense of pipeline clarity Here's what a real model looks like: - ICP Fit tells you WHO deserves attention - Engagement tells you WHEN to act - Decay scoring kills false positives automatically - The matrix (A1 to C3) tells your sales team exactly where to spend human energy In this build: 23,995 contacts. Average score: 8 out of 200. 9,992 sitting in C3 -> low fit, low engagement. That's not a data problem. That's a prioritisation problem in disguise. Fix the model → fix the pipeline → fix the Monday morning Sales vs Marketing argument. It's never a fixed model. It evolves as your buyers evolve. Treat it like a product, not a setting.

A few weeks ago, I tried something simple on a client project. When someone was about to leave key pages pricing, features, comparisons we didn’t push harder. No discounts. No aggressive “book a demo now”. Just a different path. A relevant guide. A softer entry point. Something that matched where they were in their thinking and something interesting happened. People stayed. Engaged. Played with ROI tools It made me rethink something. Most B2B websites are still built like brochures. Same message. Same flow. Same CTA. But buyers don’t behave like that anymore. They come informed. They’ve already compared options. They’re closer to a decision than we assume. So the question is not: “How do we get more traffic?” It’s: “How do we adapt the experience in real time?” What I’m seeing now in personalisation: It’s not about adding names to headlines. It’s about reading intent and responding to it. Not just pages - but journeys Not just segments - but behavior One signal should trigger multiple actions across the funnel. What we’re testing next: - show a different journey based on: • page visited • ICP fit • intent signals Example: High-intent ICP visitor on pricing - show ROI angle or strategy call Early-stage visitor - show learning content or use cases Personalization is shifting from: “change the message” to “change the path” Still early, but this is getting interesting.

Rebuilding Growack quietly. Traffic turns into revenue when the system is real - No ads - No VC - No fake momentum Growack pipeline: €242,530 total €139,890 weighted €137,520 closed Numbers don’t need hype Boring is beautiful when it pays the bills.

I don’t chase virality. I chase people who stay Growack last 30 days: • 13,450 pageviews • 3,079 users • 4m+ average session • 33.9% bounce Attention compounds before revenue does...

HubSpot + LinkedIn users, mark this down. Jan 19, 2026 is the deadline. If your LinkedIn personal account is connected to HubSpot, you must reconnect it to grant new API permissions. Why it matters: • Reconnect before Jan 19 → HubSpot backfills 365 days of follower data • Reconnect after Jan 19 → Data starts from zero, only going forward Actually useful if you track LinkedIn performance in HubSpot instead of posting into the void. Don’t be exporting LinkedIn analytics manually in February wondering why your dashboards are empty.