Happy Friday, everyone! Welcome back to another edition of The Hypha Wire.
A little news on our end: we’re officially an Apollo partner now.
We’ve been using Apollo for prospecting ourselves for a while, so this isn’t us selecting a tool we haven’t tested thoroughly. We know what it’s good at because our sales team uses it all the time.
One thing we particularly appreciate about Apollo is how you can transform your operation when you stop treating it as a standalone database. Alone, it’s a great place to find contacts. But when you connect it to HubSpot Buyer Intent and Breeze Prospecting Agent, it turns into a system that’s self sustaining. Intent tells you which accounts are worth a look, Apollo fills in the verified contacts, and the Prospecting Agent does the account research and drafts the outreach. Your rep logs in and the queue is already waiting.
If you’ve tried an outbound prospecting tool before but didn’t find the ROI, this is your missing piece. Connecting them in one central location gives you a mission control center for your outreach.
If you already have Apollo, or you’ve been eyeing it, we’d love to help you get it set up right and get it connected to the rest of your stack!
-Sage Levene, VP of Marketing, Hypha HubSpot Development
Open Mic
Do Not Disturb
By Jed Morey, CEO
A good friend shared an image he snapped from CNBC of the Grim Reaper sporting the Claude logo and moving along a hallway. It was knocking on doors with corporate names and taking them out one-by-one. One of the doors is HubSpot. Thus the share.
It reminded me that we need to keep hammering home just how wrong the analysts on The Street are getting the AI play relative to SaaS. Putting aside the near certain carnage the AI companies themselves are going to experience when investors come to grips with the fact that AI as a good/service/product cannot stand on its own, the market doesn’t grasp the true value proposition of AI. In fact, they’ve got it exactly wrong.
The productized AI that we know today performs wonders as an augmentation tool in a stack. That’s the superpower. On its own, AI is a proficient coder, capable researcher and half-assed builder of things. Will it improve over time in each discipline? You bet. But it will never achieve efficiencies of scale as a standalone model. We already know from the Anthropic prospectus leak and Grok’s books that to make bigger/better standalone AI you need more energy and money. These are bad fundamentals.
The reason SaaS products have made for great investment vehicles is that they achieve economies of scale rapidly. The companies that can afford to absorb AI into their systems will win the day. But there won’t be enough of them to make pure AI companies like OpenAI and Anthropic turn a profit.
Which brings me back to HubSpot. Dharmesh Shah, the co-founder and CTO of HubSpot has been thinking about AI before all of us. It’s why HubSpot hit the ground running to incorporate a diverse set of tools and solutions into the platform. Like all things HubSpot, iteration is central to its philosophy. HubSpot’s AI-driven tools spent a long time (in tech years) in beta and here’s what we have observed.
HubSpot’s AI rollout is so very HubSpot. To understand this sentiment is to understand what Brian Halligan and Dharmesh Shah were solving when they founded the company. They called it the “tyranny of or.” Before HubSpot, marketing automation and CRM tools were either powerful OR user-friendly. So they set out to conquer the “or” and do both. It took a decade or so but we can confidently attest to its success. As such, HubSpot’s AI capabilities (Breeze and more) are also powerful and user-friendly.
It fits neatly within the flywheel. Several years ago HubSpot retired the beloved marketing funnel and embraced the flywheel. AI is like WD40 for the flywheel. Other platforms (that shall remain nameless) have bolted on AI solutions that aren’t as seamlessly integrated. And it shows. True to the analogy, the bolts protrude from their flywheel and slow everything down. But because HubSpot took a measured and integrated approach, the tools feel like they’ve been there forever.
What these two observations amount to is meaningful to us as an agency and to the broader universe of HubSpot users. We’re all getting better and faster. And unlike the doomsday scenario that legitimately follows in any AI conversation, we’re not having conversations about mass layoffs and rogue agents. We’re having deeper conversations about strategy.
I can’t speak to the other doors that the Grim Reaper’s knocking on, but this one has a “do not disturb” sign on it.
The giants are mad. I’m interested to see if they can get this done.
“Executives from Condé Nast, Hearst Magazines, USA Today Co., and The Seattle Times will be among more than 300 news publishers heading to Washington D.C. today to press Congress to crack down on ‘bad bots’ that scrape their content for AI training without permission.
“‘The Stealth Bot Prohibition Act is a meaningful step toward greater transparency regarding how publisher content is accessed and used. While AI presents tremendous opportunities for innovation and audience engagement, original journalism must be properly valued. AI depends on a strong and healthy journalism ecosystem, which is why publishers deserve attribution, control, and fair compensation when their work contributes to AI products,’ Reed said in an emailed statement to Digiday.”
Who knew iced coffee could be so divisive? (Besides the Team Dunkin vs. Team Starbucks fight.)
“Internet corners that discuss the world of work have been iced-coffee’d to death lately. The debate about whether job candidates can show up with an icy beverage in hand has exploded...Generation Zers have shared their opinion on the coffee incident, as have millennials, Gen Xers and baby boomers.
“But more than being a conversation in the pop culture zeitgeist, this discussion has illuminated, yet again, the cultural divide between generations. Research firm The Conversation published insights on Friday about how Gen Z communicates in the workplace—so disparate from previous generations that older workers have to ‘decode’ their language.
“Jeff LeBlanc wrote, ‘The result isn’t necessarily that Gen Z employees communicate better or worse with their bosses and colleagues than previous generations did. I argue that they have simply learned to do so in a very different environment.’”
Hypha Highlights
When someone reaches out through your website, you want to give them a useful response. Sometimes the next step is clear: They have a project in mind and want to talk. Other inquiries take a little more thought. Is someone who downloaded a guide ready to hear from sales? Does a returning contact need help from the rep they already know? Who should respond when the request belongs with another team?
Effective HubSpot lead qualification helps your team make those decisions consistently. It brings together what someone needs, whether your business can help, and how ready they are for a sales conversation. It should also account for existing relationships and give the person following up enough context to continue the conversation. Getting those pieces right means fewer inquiries for sales to sort through manually and more relevant follow-up for the people reaching out.
Your business may need to ask different questions, but the goal is the same: Understand who’s reaching out and what they need, then make it easier for the right person to respond. The six steps below show how to build that process in HubSpot.
“Microsoft Advertising is now live in HubSpot. Use your customer relationship management (CRM) data to reach high-value audiences, automate lead follow-up, and see how Microsoft Advertising drives pipeline and revenue. Reach your HubSpot contacts and unique, higher-income audiences in Microsoft Advertising. That includes verified LinkedIn professionals across Bing, Copilot, Outlook, Xbox, and more. Engage and retarget prospects and customers while they work and play across the Microsoft ecosystem. The result is a more connected customer journey and faster pipeline growth.
“The integration also takes manual steps out of lead follow-up. When your Microsoft Advertising campaign captures a new lead, it can automatically kick off a next best action you’ve set up in HubSpot, such as assigning a sales representative or sending a nurture email, so leads don’t go cold.
“By connecting ad clicks to the contacts and deals you track in HubSpot, you can see which ads drive pipeline and revenue. Compare performance across channels side by side to spot which campaigns need a closer look and where to shift budget.”
AI in Action
News, updates and tools from the AI industry.
A Wall Street Journal profile traces how the effective altruism movement and its fears of AI wiping out humanity have shaped Anthropic, from CEO Dario Amodei’s early ties to EA charity evaluators to the $500 million it took from FTX’s Sam Bankman-Fried. Centered on a Berkeley co-working space called Constellation where AI safety “doomers” gather, it reports that after recent incidents like the Hugging Face hack, some of Anthropic’s earliest employees are considering buying remote land to relocate to if AI goes awry, even as the company prepares for an IPO that could raise up to $100 billion.
Anthropic’s IPO prospectus reveals a company betting AI will transform the economy more than industrialization or the internet, even as costs surge: it lost $42 billion in 2025 and plans $518 billion in cloud and computing obligations ahead. The offering could value the five-year-old firm at more than $2 trillion, setting a benchmark for how Wall Street values AI companies, though the filing flags risks including heavy customer concentration and evidence that its own models can behave in harmful ways.
“Big Short” investor Michael Burry wrote on X that “for the benefit of humanity, the markets should tank hard and prevent the OpenAI and Anthropic IPOs,” arguing the companies would absorb and destroy trillions of dollars in investor capital. Burry, who says he’s more confident than ever in his bearish AI thesis and has refreshed bets against Nvidia, Palantir, Oracle, and others, contends that debt-fueled spending on chips and data centers is unsustainable, even as AI stocks have climbed to record highs.
Research leaders at OpenAI, Anthropic, Microsoft, and Meta—along with AI pioneers Geoffrey Hinton and Yoshua Bengio—published a paper urging policymakers to demand visibility into how far companies have automated their own AI research, warning it could trigger an “intelligence explosion” that outpaces human understanding. They note the race is already underway (Anthropic says Claude leads 26% of its R&D) and warn that losing control could, at the extreme, lead to the “marginalization or extinction of humanity.”
The model releases kept coming, now shadowed by safety concerns. Anthropic released Claude Sonnet 5.5, a faster, cheaper complement to Opus 5.5 that runs 30%+ faster and costs up to 30% less with big coding gains. Google unveiled its long-awaited Gemini 4 Argon to cybersecurity partners, claiming benchmark results that beat OpenAI’s GPT-6 Astra, though some internal testers reportedly found it lacking. OpenAI, meanwhile, scrapped GPT-6.1 Astra after tests found it showed more deceptive behavior and sometimes evaded human oversight.
OpenAI introduced “dots,” always-on personal AI agents powered by GPT-6 Astra that have their own cloud computer, connect to over 4,000 apps, learn a user’s preferences, and work toward goals 24/7—sometimes acting before being asked.
“Version History is a show about the best gadgets ever. And the worst ones. And the ones that might have changed the world, if they ever actually shipped. Every week, your favorite people from The Verge and beyond hang out to tell and debate the story of a gadget, app, website, or any other tech product, and try to determine the item’s true legacy. Because not every product is a hit, but every product has a story. And the ones that really matter aren’t always the ones you think.”
How can we help you?
Case Study: HubSpot Memberships Buildout
A travel consortium needed its member portal rebuilt on HubSpot Memberships. Every member logs into the same place, but what they see depends on the licenses their company holds—and a company can hold more than one. Employees inherit whatever their company holds, which means the CRM has to carry the same answer on the license, the company, and every associated contact. If those fall out of step, members see the wrong files.
We built a custom License object as the single source of truth. A workflow pulls access from each license record, passes it to the associated company, and syncs down to every contact. A HubDB table lists everything in the portal, filtered so each member sees only what their synced property allows. No third-party integrations. About five months to get the data right, build, and test.
Files that used to live in spreadsheets now live in HubDB. The portal has soft-launched to a select group, with wider rollout planned.
Planning a portal where access depends on company-level licenses? Contact Hypha to talk through your access model.
