Barry O'Brien

Barry O'BrienSep 10, 2026

🤖 ➡️ 🤝 The job-seeking candidate journey is changing. We’ve just published the first article in a new series exploring Employer Branding and AI Search - and what it means for how candidates discover, research and compare employers. Increasingly, the journey can start like this: 👉 Ask AI → shortlist employers → investigate companies → apply Which creates a new question for Employer Branding teams: What is AI telling candidates about your company before they ever reach your careers site? In this first article, we look at how AI is reshaping the candidate journey and why Employer Brand teams need to start paying attention. We’ve also included a free "AI Search – Employer Brand Audit" - a practical guide and checklist that you can download to help you find out what AI is telling jobseekers about your company today. (Link to article + free audit in the comments. AI assisted image below.) Thoughts? Malin Sandell, Lina Kvizikevičiūtė, Anders Norrman, Elin Larsson, Måns Liljenlov, Elise Vreeman, Larissa Hällefors, Julia Gullbrandsson, Cathy Andreasson, Alexander Tonelli, Mattias Andersson, Torgil Lenning, Anna Holmgren

Barry O'BrienSep 1, 2026

🤖 AI in marketing has mostly meant "help me create." The more interesting shift happening right now is AI that can see what actually worked - and act on it. HubSpot just expanded its connector for Claude. MCP (Model Context Protocol) is the open standard behind the connection - it lets Claude securely access HubSpot data and, where permissions allow, take actions in HubSpot. As a HubSpot partner, we're watching this closely at Marketinghouse International Inbound Marketing Agency. It works like this - instead of just asking Claude to draft a campaign, you can ask which of your campaigns generated the most deals last quarter, have it analyse what they had in common, and use those insights to help build the next one - inside the same conversation. Worth noting: this sits alongside HubSpot's own internal AI, Breeze - native agents built into the HubSpot CRM for things like content and prospecting. The MCP connector is a different bet. Rather than building every AI capability in-house, HubSpot is also opening its platform so external models like Claude can work with CRM data and tools directly - two paths towards the same goal: AI that's grounded in your actual CRM data, rather than guessing. The latest connector update adds two capabilities I think marketers should pay attention to: ✅ Landing pages and marketing emails can now be created and edited from Claude, with performance data pulled back in for analysis - including landing-page conversion rates and email clicks and conversions. It starts to close the loop between "make the thing" and "see if it worked." ✅ Campaign and revenue attribution, plus lead and customer journey context, can now be queried conversationally. So a marketer can ask questions like "Which campaign actually drove pipeline?" without having to jump between tools and reports. The way this is heading - CRM data + AI + natural-language workflows, whether native or connected - feels less like a feature update and more like a preview of how marketing teams will run their CRM day to day. This is still early, as HubSpot says it's working towards full API parity for apps and agents built on its MCP server. Is your team using HubSpot's own AI agents, connecting external tools like Claude, or both? Let me know. If you have Hubspot and need support - reach out. See the Hubspot announcement link in the comments. (...and, yes, the image is created using AI)

Barry O'BrienAug 25, 2026

🤖 AI search isn’t replacing Google overnight - but user behaviour is clearly shifting. These five stats show what’s changing, and why visibility in AI answers now matters alongside traditional SEO. Connect with me if you need help figuring this all out for your business...