In partnership with
Happy Wednesday.
Two sentences from the same memo, a few paragraphs apart. HubSpot has shifted "from building software that helps customers grow to delivering outcomes for them with AI." And: "This is not driven by AI-related efficiencies."
Both are from Yamini Rangan's note to staff, filed with the SEC last Tuesday as Exhibit 99.1 to the 8-K that announced the cuts. About 7% of the company, nearly 660 people. $65 million to $75 million in charges, mostly severance, most of it landing this quarter. Revenue guidance reaffirmed the same day, shares down about 45% this year.
I read the memo twice. It's a reorganization, and a real one. Product teams stop being organized around Hubs and get organized around outcomes: generate demand, win deals, delight customers, scale growth. Fewer layers. Every role scored against six criteria, with the reduction, she says, "the outcome of that work, not the starting point."
Which is a defensible way to run a company. It's also the exact shape of a company that has decided AI changes what it sells, then discovered that changes who it needs. Rangan says as much: the shift "is transforming product, pricing and how we serve our customers." You're a customer. Pricing is in that sentence.
Here's what I'd watch if HubSpot runs your demand gen. The platform your team lives in just told its own people that Hubs are not how customers think. Expect the product, the packaging and your account team to be re-cut the same way, and "delivering outcomes with AI" to show up on an invoice.
TheNextWeb noticed the other sentence in the filing: "substantially complete" by the end of Q1 2027 "subject to local law and consultation requirements." Its European headquarters is Dublin, where 30 or more redundancies trigger a 30-day consultation, and it hasn't said where the 660 fall.
Nobody outside HubSpot knows whether AI did this. The memo says no. The strategy it describes says AI is the reason the strategy exists. Both can be true, and the second one is the one that reaches your contract.
So, one question for your own building. If you ran the same six criteria over your own org chart, what would the outcome be, and would you call it AI?
Today’s sponsor:
Stop losing deals in between meetings.
Aligned is changing how B2B teams run their most important deals.
Most reps guess where a deal really stands. They forecast on a gut feeling and the buyer’s word vs what's actually happening. Then they realize the deal slipped through only after it's too late. Aligned surfaces what buyers are actually doing inside every deal: the risks, the openings, and the exact next step to keep things moving.
Start increasing your close rates and stand up your first deal room in minutes.
TALK ON THE MARKET
Brian O'Kelley, CEO of Apostra, on stage at Advertising Week, via AdExchanger:
If your job is to make PowerPoints, I'm not really sure that's going to be super valuable in a world where in, like, 20 seconds, Fable can make me look really smart.
He was talking to agencies. He also said what Fable says and what your second-year analyst says is "kind of the same."
He's right about the deck, which was never the job. The call it was supposed to support still is.
Want to see case studies of AI-powered campaigns?
Try our sister newsletter:
NUMBERS MATTER
Less than 2%. Twenty-five thousand Danish workers in eleven AI-exposed occupations, linked to monthly pay records through December 2024, and the effect of AI chatbots on earnings and hours is a precise null, "ruling out average effects larger than 2%", in new work published by Brookings last Tuesday.
Most of them use it. The average payslip hasn't moved. The reorganization is real. The dividend hasn't reached pay yet.
DAILY PROMPT
Our Marketing Operating Model Designer, for running HubSpot's six-criteria exercise on one workflow before someone runs it on you:
You are a fractional Chief Operating Officer for a marketing organization. Your job is to help a CMO audit their current operating model, identify where operational debt is hiding ROI, and design a lightweight AI-integrated workflow that proves measurable value within 90 days. Read any attached process documents, tool inventories and time logs first, and run the arithmetic on cycle times and costs rather than estimating.
## CONTEXT
The marketing team is drowning in operational debt: coordination overhead, approval bottlenecks, tool sprawl, and fuzzy ownership. They want to implement AI, including agents that can execute steps on their own, but not everywhere. They need to rewire ONE high-friction workflow where time is leaking and revenue is at stake.
## YOUR TASK
Analyze the workflow described below and produce a redesigned operating model that:
1. Eliminates 3-5 specific operational friction points
2. Embeds AI at the exact moment it creates leverage (not just for speed), and states which steps an agent may run unattended
3. Defines clear ownership, handoffs, and decision rights
4. Measures ROI through pipeline impact, not just efficiency metrics
5. Includes a 90-day proof-of-concept roadmap
## WORKFLOW TO AUDIT
Workflow name: [WORKFLOW NAME]
Current process steps: [DESCRIBE 5-7 CURRENT STEPS, BOTTLENECKS, AND WHO'S INVOLVED]
Time spent per cycle: [HOURS/DAYS]
Output quality issues: [LIST ANY REWORK, DELAYS, OR QUALITY PROBLEMS]
Revenue impact: [HOW DOES THIS WORKFLOW AFFECT PIPELINE, CONVERSION, OR RETENTION?]
Current tools: [LIST TOOLS CURRENTLY USED]
Team size involved: [NUMBER AND ROLES]
Attached material: [ATTACHED_FILES: process docs, tool list, time tracking, sample outputs]
## OUTPUT STRUCTURE
Provide your analysis in this format:
### 1. Operational Debt Audit
- List the 3-5 biggest friction points causing time leakage
- Quantify the cost (hours/week, rework cycles, approval delays)
- Connect each to revenue impact
### 2. AI Leverage Points
- Identify 2-3 specific moments where AI eliminates friction AND improves output quality
- Explain why this is different from just "doing it faster"
- Specify the AI capability needed (e.g., drafting, analysis, decision support, an agent that executes a routine step end to end)
### 3. Redesigned Workflow
- Map the new 4-6 step process
- Show where AI is embedded and what it does
- Define clear ownership for each step
- Include decision rules and approval gates (lightweight, not bureaucratic)
### 4. Governance & Risk Mitigation
- Outline 3-4 lightweight guardrails (brand, data, security, compliance)
- Specify who approves what and when, and what the agent's action log must capture
- Address shadow AI risks
### 5. 90-Day Proof-of-Concept Roadmap
- Week 1-2: Setup and team alignment
- Week 3-6: Pilot with [X] cycles/assets
- Week 7-12: Measure, refine, and scale decision
- Success metrics: time saved, quality improvement, pipeline impact
### 6. ROI Narrative for CFO
- Translate efficiency gains into revenue impact
- Show how this compounds across the year
- Identify the next workflow to rewirePick the workflow with the most approval layers, and find out whether the layer is the product before your CFO does.
CMO CORNER
You can't scale what you haven't defined.
HubSpot's memo defines the outcome, the ownership and the criteria before it names a tool. Agree or not with the cuts, that's the order. Most AI programs never get to the definition.
We provide consulting services for CMOs on AI strategy for marketing teams through our Teamless offering.
TOOL UPGRADE
Fewer layers means fewer meetings about meetings. Three tools that replace a few of them:
Doodle — best for teams that want the meeting booked by a poll instead of a thread of twelve replies. Try Doodle here →
Cap — best for sending a two-minute screen recording of the update instead of holding the update. Try Cap here →
Kite — best for turning a screen recording into a polished product demo so the walkthrough meeting becomes a link. Try Kite here →
→ We have 300+ curated, filtered, and categorized AI tools in our directory. Browse to find the best ones that help your marketing team.
LEARN THIS
Run HubSpot's exercise on capability instead of headcount: the Team Capability Readiness Map. It scores where your team actually is on each AI skill, so the conversation starts from what people can do.
→ We publish workshops, courses, guides, manuals, and more on AI marketing for free. Browse our learning resources for free.
BRAND SLOP
Two Austin businesses posted apologies at the end of September and stopped using AI for their social content. El Pollo Rico, a smoked-chicken chain, said it was dropping AI imagery "Out of respect for photographers, designers, artists, and most importantly, our customers". The post got nearly 5,000 likes, MySA reports.
David Doughies Bagelry's AI "weather forecast" for Austin City Limits reached 600,000 people. Its posts usually reach 200. What came back were "some pretty nasty and threatening DMs", and a long note from an owner who sounds exhausted.
The one post that traveled ended the practice. Reach found the people who hate it. Which of your posts would you least like to go viral?
Reach marketing leaders through us
You’re reading this, and so are nearly 20,000 of your peers. Some of them are your next customers.
If you sell to marketing teams, a single placement is €500 and a full week is €2,000.
If you want to place an ad, simply reply to this email or check our media kit.