How Agentic AI is transforming customer intelligence for specialized information businesses
Real-time insights, deeper account relationships, and faster execution - without waiting on IT
If you run a subscription, membership, or events business in a specialized information market, you already know the pressure: grow average account value, deepen strategic partnerships, and do it all with a lean team. What BVR's recent experiment with agentic AI shows is that the tools to do exactly this are already available - and the barrier to entry is lower than most senior leaders think.
This blog unpacks the key lessons from a recent On the Offense session with Lucretia Lyons, Chairwoman and David Foster, Chair of BVR, where they shared how they're using agentic AI to transform customer intelligence, content acquisition, and commercial outcomes across their business.
Continue the conversation at Renewd
If these challenges resonate, sign up for the Renewd Community Digest to receive regular insights and resources for media and events leaders.
What is Agentic AI - and why should B2B information leaders care?
Agentic AI refers to AI systems that don't just respond to a single prompt - they pursue a defined mission autonomously, carrying out multi-step tasks, checking sources, summarising findings, and delivering outputs on a schedule, without human intervention at every step.
Think of it as the difference between asking a colleague a question and hiring a researcher who proactively monitors dozens of sources, flags what's relevant, and lands a briefing in your inbox every morning.
For B2B media, events, and subscription businesses, the implications are significant:
-
Your sales team gets real-time intelligence on every strategic account - without manual research
-
Your editorial and content teams can track what your most valuable customers are publishing - and act on it within 24 hours
-
Your data products can be continuously enriched - without expanding headcount
The BVR case study: From vision to working prototype
BVR's starting point was a marquee partnership programme - a strategic initiative to deepen relationships with their largest accounts, many of whom are content creators in their own right.
The insight was elegant: their customers were producing high-value thought leadership that BVR could amplify, while simultaneously generating signals that could improve sales conversations, inform content strategy, and enrich their proprietary databases.
The challenge was execution at scale. Doing this manually across dozens of accounts was impossible. Agentic AI made it achievable.
Three practical applications worth stealing
1. Real-time content monitoring for strategic accounts
Using agent mode in ChatGPT, BVR built a system that monitors the "Insights" pages of their top clients - starting with CBIZ, their largest account - and delivers a structured briefing whenever new content is published.
The agent mission, refined over many iterations, is elegantly simple. Here is the actual task summary BVR landed on:
"Every day, check the CBIZ Insights page for newly published articles or news items. For any new content since the last check, gather key details and prepare an email-ready summary. If there are no new posts, report that clearly."
Behind that simple mission statement is a seven-step prompt that specifies everything the agent needs to do its job reliably:
-
Timing - run at 9am America/Los_Angeles time each day
-
Source page - visit the exact URL (e.g. cbiz.com/insights)
-
Find new items - identify what has been published since the last check, focusing on the primary content list, not sidebars or widgets
-
Collect details - for each new item, open it and note the title, publish date, URL, and a 2–3 sentence summary of key points
-
Compose an email-ready summary - with a subject line, intro sentence, hyperlinked titles, and a snippet per item
-
No new posts - if nothing is new, output exactly: "No new CBIZ Insights posts since the last check"
-
Recency and accuracy - always fetch the live page and verify publish dates; never rely on stored knowledge
The result? Three minutes. A structured briefing with hyperlinked articles, publish dates, and editorial summaries - ready to paste into an email or drop into a shared Excel file that grows into a cumulative content log.
The same model replicates across multiple source types. For CBIZ alone, BVR runs agents across four distinct pages: the Insights hub (thought leadership and research), the Events page (webinars, podcasts, live events), the In the News section (PR, people news, acquisitions, awards), and the About page (team members, new hires, expertise). Each feeds a different part of the business - editorial databases, CRM, sales intelligence, partnership development.
They also run similar agents on Accounting Today - a trade publication they subscribe to - tracking profession news, audit developments, and M&A activity, flagging anything relevant to their market in real time.
Why it matters for subscription and membership businesses: Every piece of content your customer publishes is an opportunity - to start a conversation, personalise a renewal discussion, or identify a new use case for your product. Every award, hire, or acquisition is a reason to reach out.
2. Sales intelligence that actually gets used
The second application is feeding this intelligence directly into the sales workflow. BVR tracks not just content, but personnel moves, award wins, speaking engagements, and firm-level announcements.
The output lands in the sales team's inbox - an automated briefing that gives them a genuine reason to reach out, something specific and timely to reference, and a clearer picture of who is doing what inside each account.
For businesses where new account acquisition is limited and growth comes from expanding existing relationships, this kind of ongoing account awareness is transformational. As Lucretia put it:
The practical framework: BVR distilled their agent-building approach into four steps they call the Instruction Manual: (1) give the agent precise marching orders with exact URLs; (2) treat agents as replicants - build once, deploy infinitely; (3) iterate the instructions until the output is exactly right; (4) add output prompts for daily feeds into Excel and email for an internal "headline service."
3. Enriching proprietary data products
BVR's third use case is the most strategically significant. The content surfaced by these agents doesn't just inform sales conversations - it feeds into their AI-powered research databases, BV Research Pro and Deal Stats.
The key principle is what they call "BVR Intelligence" - a second layer of human curation that sits between the agent output and the database. Nothing enters the data product without being vetted against clear standards: Is the author a subject matter expert? Is it citable? Has an internal expert selected and contextualised it? Is it auditable and scoreable?
This curation layer is the moat. It's what separates a proprietary database from a generic AI data dump - and it's what makes the data defensible, valuable, and licensable. The goal BVR has set themselves is 100% SME-authored and auditable content in their databases - with agentic AI accelerating discovery, and human expertise determining what makes the cut.
The Claude Connector Programme: A distribution channel worth considering
One of the most forward-looking parts of the discussion was around Claude's connector programme - a framework that allows information providers to authenticate their paid content within Claude's interface.
The model works like this: when a professional uses Claude for research, the connector programme allows Claude to draw on authenticated, paywalled sources - including yours - while respecting your access controls and licensing structure. Users with paid access get your content surfaced alongside Claude's broader capabilities. Users without access hit the paywall.
Major players including Morningstar, PitchBook, and S&P Global's CapIQ are already participating. For specialized information businesses, this represents a potential new distribution and revenue channel - one that reaches professionals exactly at the moment they're doing research, without requiring them to navigate to your website first.
As David Foster put it:
The zero-click future and what it means for your business model
One of the sharpest observations in the session was about changing user behaviour. Younger professionals don't start their research day by visiting your website. They expect information to come to them - surfaced by AI, delivered to their workflow, with no browsing required.
This is the "zero-click" phenomenon already visible in Google search: 72% of searches now result in no click through to a website, because the AI provides the answer directly.
For subscription and membership businesses, the implication is clear: being findable on your own website is no longer sufficient. Your content and data need to be present in the tools and interfaces your customers are already using - whether that's Claude, an enterprise AI assistant, or a sector-specific workflow tool.
The connector programme approach is one answer. But the broader principle is to think about distribution first, destination second.
How to get started: The practical path
The most encouraging message from both BVR and the broader Renew community is how accessible this actually is. You do not need a large IT project or a six-figure technology investment to get started.
BVR distilled this into four steps they call the Instruction Manual - the clearest practical guide for getting started:
Step 1: Give the agent precise marching orders. Specify the exact URL you want it to monitor, the time you want it to run, what counts as "new," and what to do if there's nothing to report. Vague instructions produce vague output. The more precise your prompt, the more reliable the agent.
Step 2: Treat agents as replicants. Build the agent once for one account and one source. Once it works, the underlying structure - timing, find new items, collect details, compose summary - is almost identical for every other account. You change the URL and the account name. That's the work of replication.
Step 3: Iterate until the instructions are perfected. ChatGPT's agent mode will push back and ask clarifying questions as you build - what should I do if there are no new posts? How many items should I include? That back-and-forth is the refinement process. Expect to go through it many times before the output is exactly right. That iteration is the work.
Step 4: Send it forth and choose your output format. BVR started with a growing Excel file - a cumulative log of all content captured - and layered an email summary on top. The key is getting intelligence into the hands of the people who need it, in a format they'll actually use.
The bottom line for specialized information leaders
Agentic AI is not primarily a cost-cutting tool for information businesses. It is an account growth tool, a content intelligence tool, and - potentially - a new distribution channel.
The businesses that will benefit most are those that use it to do what they already do well, faster and at greater scale: know their customers deeply, surface what's relevant, and deliver it in ways that create genuine value.
Your moat - curated, expert-authored, auditable content and data - doesn't disappear in an AI world. If anything, it becomes more valuable, precisely because AI makes it so easy to generate undifferentiated content at volume.
The question is whether you're building the infrastructure to defend and distribute that moat effectively. The tools to do so are already here.
This post is based on the Renew On the Offense webinar "Agentic AI: A Case Study - Turbocharge Content & Data Acquisition, Deepen Customer Partnerships & Increase Average Account Value," featuring Lucretia Lyons and David Foster of BVR.
Continue the conversation at Renewd
If these challenges resonate, sign up for the Renewd Community Digest to receive regular insights and resources for media and events leaders or check out our free and Member-Exclusive events here.
Or join Renewd as a Premium Member and have these conversations in person, with peers who know exactly what it's like to sit in your seat.
Premium Membership North America Premium Membership UK/Europe
