Why Only 6% of Companies See the True Impact of AI - GROWTH (Ep. 57)

Original author：FrodX d.o.o.

Original source：https://www.youtube.com/embed/C77d350biH4

As many as 90% of companies already use AI, but only 6% say it actually improves their profitability. Where is the difference? In the 57th episode of the RASTezanja podcast, Andrej Bergant and Gorazd Zakrajšek break down what happened at this year's HubSpot conference (renamed from Inbound to Unbound this year) and what it means for marketing, sales and CX teams. This is not an overview of all the new functionality. It's a conversation about why AI isn't delivering results at most companies, and what those that are are doing. ❓ What awaits you in this episode 📊 Why AI doesn't (yet) bring results Research among 6,000 customers: 90% use AI, 6% see impact on profitability Successful ones don't automate everything, but 4-5 key use cases Paradox: the more popular the use case, the lower its business impact Campaign analysis, pipeline prioritization and customer feedback analysis bring the most Before implementing use cases: which KPI will AI move and how will you measure it? 🧠 Growth context: AI + context + your data The model is the same for everyone, the context differs Analogy: a new salesperson does not first need Excel with 50,000 contacts, but introduction Three layers of context: who we are, what we know about customers, how our team works The price of bad data is potentiated with AI, as AI confidently offers a product that no longer exists The model becomes a commodity, and the value comes from the context 🤖 CRM that updates you Unstructured data from emails, transcripts and minutes automatically update contacts and deals With poor context, AI results are worse than without it Agents multiply mediocrity: instead of 12 generic emails, they send 1,200 🔍 AI search engines and the new buyer's path HubSpot lost around 80% of website visits, but got 20% more leads From the sales funnel to the hourglass: AI platforms are gatekeepers that release more warm leads Content should not only be on the blog, but also on YouTube, LinkedIn and forums 🛠️ What HubSpot presented Marketing Studio: from goal to campaign with multiple participating agents Prospecting agent with more advanced triggers, also via prompt Mobile note taker for live meeting minutes, also works in Slovenian ChatGPT ads directly from HubSpot, with conversion tracking ✅ What to do tomorrow Define the KPI you want to move with AI Check data quality and clean it, where AI can also help you Start with one agents, not five, and improve it iteratively ⏱️ Episode timeline 00:00 Intro: x-ray in a shoe store 01:50 90% use AI, only 6% see impact 04:46 Popular vs. profitable use cases 08:38 Growth context and data 11:23 Three layers of context 15:34 Self-updating CRM 18:25 Does the model even matter? 20:09 Bad context, worse results 21:18 HubSpot: fewer visits, more leads 25:15 Content for AI and the hourglass 32:04 Marketing Studio and agents 38:06 Prospecting agent and note taker 41:10 HubSpot Ads in ChatGPT via HubSpot 47:13 What to do tomorrow 🎤 Hosted by Andrej Bergant – Head of R&D, FrodX https://www.linkedin.com/in/andrejbergant/ Gorazd Zakrajšek – Innovation Consultant, FrodX https://www.linkedin.com/in/gorazdzakrajsek/ 🔗 Useful links 👉 Keynote from the HubSpot conference: https://www.youtube.com/live/2kjlwugSao8?si=PATDok0rY5_X8P_G&t=1019 👉 Spotify: https://open.spotify.com/episode/2BMGda4J1MJx7thwZpgM4j?si=t-z1mZ8MRwi5ps7IM0W5rw 👉 Ask a question for RASTezanja: https://frodx.com/vprasaj 👉 All podcast episodes: https://bit.ly/49Ct88b 👉 FrodX content: https://bit.ly/3V6Ahcr 👉 LinkedIn: https://bit.ly/3v0XcLK 👉 Instagram: https://bit.ly/3v5Pujd 👉 TikTok: https://bit.ly/3IDiQJ6 👍 If you found the episode useful, leave a like and a comment or send it to a colleague who is introducing AI in the company.