Maja Voje

Recently, I hit 70K LinkedIn followers.
But I didn’t feel like celebrating much.

Because in 2025… LinkedIn got tough.

Organic reach is down.
Posts stay in feeds for weeks.
Automations? Punishable offenses. 👮‍♀️

Some creators saw engagement drops of 90%+.

And honestly, I began to feel like I was wasting my time here.

Somewhere between the “50 meetings booked” posts and AI-generated comment loops, it all started to feel… tedious and disheartening.

But instead of rage-quitting LinkedIn,
I decided to make it fun again.

How?

By vibe-coding 4 AI agents that now help me research, engage, prospect, and learn - all without breaking a single rule or spending a single euro on tools.

→ 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 𝐀𝐠𝐞𝐧𝐭: Builds moodboards of top-performing posts and updates my Miro boards automatically.

→ 𝐌𝐚𝐬𝐭𝐞𝐫 𝐨𝐟 𝐄𝐧𝐠𝐚𝐠𝐞𝐦𝐞𝐧𝐭: Curates posts I should comment on and delivers them straight to Slack.

→ 𝐈𝐂𝐏 𝐅𝐢𝐧𝐝𝐞𝐫: Finds ICP-qualified leads from my post reactions and sends them to me, pre-classified.

→ 𝐋𝐢𝐧𝐤𝐞𝐝𝐈𝐧 𝐂𝐨𝐚𝐜𝐡: Analyzes my and others’ performance weekly, gives personalized feedback, and even chats on Slack.

These agents give me faster feedback loops, sharper intelligence, and help me focus my time here where it actually matters — building relationships and ROI instead of screen-sucking the feed.

Want to see how they work and how to build them yourself?

Read the full breakdown 👇
https://lnkd.in/dSNECzNE

🔴 Live from NYC: Miro Canvas 25

I’m attending Miro’s flagship event in New York, where the company unveiled a new wave of AI-powered collaboration features and innovations.

Highlights from today’s announcements (coming soon to Miro):

→ Miro Sidekicks – context-aware AI partners that work directly on your canvas

→ Flows – automate entire workflows and connect ideas seamlessly from start to finish

→ Prototypes – transform early sketches into testable, interactive concepts in seconds

→ Insights & Portfolios – bring product and GTM alignment into one shared, data-driven view

It’s inspiring to see how Miro, AWS, Intercom, GitHub, Red Hat, and NFL use AI to empower collaboration and innovation.

📹 Watch the livestream here: https://lnkd.in/dZtZkriR

P.S. Don’t miss “Breaking Through the AI Scaling Gap” at 5PM EDT with Eduardo Ordax, Principal GenAI, Go-to-Market at AWS (184K LinkedIn), as he shares best practices to scale AI across the enterprise.

#miropartner

Miro just crossed 100M users and launched the AI-first collaboration workspace.

I attended Canvas 25 in New York.

In addition to the ground-breaking product innovations (Flows, Sidekicks, MCPs, and seamless interoperability with tools like AWS, Lovable, Cursor), I fell in love with their brand marketing.

Here are 9 ideas from Miro that can spark more brand love in your GTM strategy for 2026:

→ Go all in on branding.
The venue, visuals, and atmosphere were beautiful. Every detail felt intentional and brought joy.

→ Go all in on visibility.
From Times Square billboards to a mural in Brooklyn, Miro made sure the city knew they were there.

→ Go all in on creators.
15 of us attended live, and 80+ influencers covered the launch. So many unique voices, so much love.

→ Go all in on team ambassadorship.
Practically every Miro employee I’m connected with posted about the event or shared announcements. That’s culture in action.

→ Go all in on customer success.
Real client stories and genuine appreciation took center stage. It all felt truly authentic.

→ Go all in on partners.
The NFL team shared their Miro projects and templates on stage with pride and affection - collaboration is in their DNA.

→ Go all in on community.
Users weren’t spectators; they were part of the story. Community members got shoutouts, contributed ideas, and shaped the roadmap.

→ Go all in on experience. Custom shirts, stickers, creative swag, even coffee holders to encourage conversations - thoughtful details made everyone feel special.

→ Go all in on relationships.
The entire team was approachable - from marketing to leadership - taking great care of everyone with smiles and genuine kindness.

Dear Miro,
Your culture shows.
Your marketing inspires.
And your execution is world-class. 💛

Claude Code runs my entire GTM system. Not a chatbot. Not a copilot. The actual infrastructure.

My GTM work is organized into folders that Claude reads. ICP research, competitive intel, campaign data, call transcripts, voice guides. A CLAUDE md file tells it how I work.

Custom skills handle the repeatable work, and agents run deeper tasks.

Here are the 5 workflows I run every week:

1. 𝐏𝐥𝐚𝐧 𝐌𝐨𝐝𝐞
I give Claude a task: "Plan an outbound campaign targeting companies showing hiring signals." Before building anything, it reads my client's ICP, competitive landscape, and past campaigns. Then it asks me what it still needs. What segment? What triggers? What angle? Once it has that, the plan comes back sharp. No five rounds of corrections.

2. 𝐎𝐩𝐮𝐬 𝟒.𝟔 + 𝐂𝐥𝐚𝐲 𝐒𝐤𝐢𝐥𝐥
A custom skill I built that knows waterfall enrichments and Claygent prompts. "Find VP-level buyers at e-commerce companies using job posting signals, score at contact level." It picks the right data sources, chains the enrichments, and asks clarifying questions instead of guessing. Opus 4.6 just gets what I'm trying to do.

3. 𝐂𝐡𝐞𝐜𝐤𝐩𝐨𝐢𝐧𝐭𝐬 
Version control for AI work sessions. I draft an outbound angle → Save a checkpoint. Try a different approach → Save another. Then a third. I can jump between all three, compare side-by-side, and keep building from whichever landed best. Without losing any work. I use this constantly for messaging experiments.

4. 𝐄𝐱𝐭𝐞𝐧𝐝𝐞𝐝 𝐓𝐡𝐢𝐧𝐤𝐢𝐧𝐠
For ICP analysis, positioning, and battle cards. I'd rather give it more time to think and get higher quality output than something instant but shallow. It works through buyer psychology and competitive dynamics using full client context. The output is client-ready.

5. 𝐌𝐚𝐫𝐤𝐞𝐭 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 (𝐂𝐨𝐧𝐭𝐞𝐱𝐭 + 𝐖𝐞𝐛)
Claude already has my client context and past research. I point it at industry reports, market trends, or a new vertical a client wants to enter. It connects what it finds online with what it already knows. (Also runs in the browser when I'm away from my desk).

Honest caveat: It takes about a week to organize your context and build your first skills. But that's one week vs. doing it the old way.

Get started with Claude Code on the Max plan: http://clau.de/majavoje

#ClaudePartner

McKinsey says 78% of enterprises have adopted AI.
Momentum.io found only 7.6% actually use it operationally.

There is a 10× gap between what companies say and what they do.

It might be the biggest disconnect in enterprise AI today.

Instead of asking leaders how “AI-transformed” they are,
Momentum’s new Voice of the Market & AI Maturity Report does something far more powerful:

It listens.

The team analyzed 1,069 real B2B sales conversations across 150 industries.

They did it using Momentum’s 𝐃𝐞𝐞𝐩 𝐑𝐞𝐬𝐞𝐚𝐫𝐜𝐡 and 𝐀𝐮𝐭𝐨𝐩𝐢𝐥𝐨𝐭 𝐁𝐚𝐭𝐜𝐡 features to extract, clean, and validate insights directly from customer calls.

The result: 56 slides and 42+ proprietary visuals revealing what revenue teams actually struggle with:

→ 74.4% admit they’re behind in digital transformation
→ 31.1% still rely on manual processes for core work
→ 41.9% cite CRM pain as their biggest blocker
→ 69.9% suffer from data visibility gaps
→ Only 5.6% are actively transforming with AI - but those deals show 2.4× more exec involvement and 89% higher close rates

If you lead RevOps, Sales, or GTM, this study is your mirror into what’s actually happening in the market - not what’s written in press releases.

📘 Download the Voice of the Market & AI Maturity Momentum Report -
and see true state of AI in organizations and where we are heading in 2026: https://lnkd.in/dsveCHbQ

And just between you and I 🤫
Do you have AI FOMO sometimes? How do you manage it?

👑 Reddit takes the crown!

Ahrefs analyzed its database of 9.6 million ChatGPT queries to identify which domains are most often cited in ChatGPT responses.

These are the top cited ones globally:
👾 Reddit
🌍 Wikipedia
🛒 Amazon

ChatGPT prefers authoritative and well-known publications, including news sites like TechRadar, The Sun, Forbes, and Wired.

What does it mean for positioning your brand on GEO (Generative Engine Optimization)?

→ Try to get valuable mentions on the domains from the list - ChatGPT finds them credible sources

→ If the top domains are hard to reach or not relevant for you, start at the bottom

→ Aim for a variety of mentions - mimic questions that customers are typing in chatbots

→ Regularly monitor your brand presence in AI chatbots - positions can change much faster than in traditional SEO

Check the full list of 100 top domains on the link in the comments.

Use Ahrefs Brand Radar to track your brand’s presence in ChatGPT and other AI chatbots: https://lnkd.in/dEhNF8hJ

How often do you check your GEO presence?

3.5 million users with no marketing budget.
Only a 2-person team.
How is that possible?

The Glasp story is a masterclass in what organic growth means in the age of AI:
→ Building for curiosity, not just clicks.
→ Letting small experiments go viral.
→ Turning human knowledge into something that lasts.

From building one of the first ChatGPT-based Chrome extensions back in November 2022...

...to figuring out AI search (AEO).

Everything is based on a sound business strategy and one key insight:

AI didn’t just change how we create.

It’s changing how we learn.

Kei Watanabe and Kazuki Nakayashiki, founders of Glasp, shared their story with me for this week's GTM Strategist newsletter.

Find the link in the comments 🔗

Most marketers trust their Google Analytics numbers.
Until they realize… 20% of their traffic is invisible 🙈

Why?
Every time someone clicks “reject all cookies,” your data disappears.

To capture all the traffic, you need a privacy-focused alternative.

Here’s where Ahrefs Web Analytics comes in.

It doesn’t collect personal data - so it doesn’t require cookies.

Here’s what I loved:
✅ Real-time traffic (no 24-hour delay as with GA4)
✅ Instant insights into campaign & UTM performance
✅ Intuitive interface (just what I was missing from the old GA)
✅ Shows you traffic from ChatGPT, Perplexity, and other AI tools

And it’s totally free to use!

(for up to 1 million events per month)

If you’re serious about growth, you can’t afford to fly half-blind.

Get started with Ahrefs Web Analytics for free: https://lnkd.in/d4PA_sBr

💡 𝐏𝐫𝐨 𝐭𝐢𝐩: install it side by side with GA4 and compare the difference in the quality of your data.

#ahrefspartner

The GTM tool stack debate is a distraction.
Pouring $45K/month into tooling and GTM services is a complete waste if your team can't answer one question:

Who exactly is in pain, and why does your solution fix it better than anything else?

Tools amplify the strategy. They don't create it.

Here are the 7 layers to actually engineer product-market fit:

1️⃣ Strategy
Find your North Star Metric. Nail OKRs.

2️⃣ Market
Find your minimum viable segment. Map where your solution is a painkiller, not a vitamin.

3️⃣ Early Customer Profile
You're too big a risk for the ICP you eventually want. Start with early adopters already saying "shut up and take my money" - not the ones who might in 6 months.

4️⃣ Product
Build the minimum viable version of your UVP. Validate before you scale. Build in iteration, not in isolation.

5️⃣ Pricing
Base it on willingness to pay and your value metric. Segment by use case - different users extract different value. One-size-fits-all pricing leaves money on the table.

6️⃣ Positioning
Turn your vision into a strategic narrative. In a sea of AI sameness, differentiation isn't optional. Nail the "what is this and why should I care" pitch - then iterate. This one is nearly impossible to get right on the first try.

7️⃣ GTM System
Funnels died in 2024. Move to growth loops, revenue systems, compounding engines.

If any of these layers are shaky, no tool will save you.

The compounding revenue system everyone wants starts here.

Not in your tech stack.

Strategy before tactics. Always.

GPT-5 launch in August divided the AI world.
Everyone was talking about the new AI model like it was the new iPhone.

Early previews, speculations, pre-launch hype - it was all there.

While it was announced as the PhD-level model, users quickly noted that it behaved … 𝘥𝘪𝘧𝘧𝘦𝘳𝘦𝘯𝘵𝘭𝘺.

I went to analyze feedback and online discussions about GPT-5 with Talkwalker’s market intelligence data (ad) to get some learnings for you.

It includes conversations on X, Reddit, in news articles, forums, and other online communities.

Here is what stood out to me:
→ Users described GPT-5 as “colder” compared to previous models like GPT-4o.
→ They were discussing how to write effective prompts to unlock GPT-5's full potential.
→ Only 28.3% of conversations had positive sentiment.
→ OpenAI was criticized for removing the ability to select a specific model.

This is what you can learn as a GTM strategist or PM:

𝐋𝐢𝐬𝐭𝐞𝐧 𝐭𝐨 𝐲𝐨𝐮𝐫 𝐮𝐬𝐞𝐫𝐬.
OpenAI quickly brought the model picker back while the users were still getting used to GPT-5.

𝐂𝐨𝐥𝐥𝐞𝐜𝐭 𝐟𝐞𝐞𝐝𝐛𝐚𝐜𝐤 𝐚𝐧𝐝 𝐢𝐧𝐯𝐨𝐥𝐯𝐞 𝐭𝐡𝐞 𝐮𝐬𝐞𝐫𝐬.
When you let any product “into the wild”, you will run into a number of scenarios you couldn’t envision even with the most rigorous testing.

𝐄𝐱𝐩𝐥𝐚𝐢𝐧 𝐭𝐡𝐞 𝐭𝐞𝐜𝐡 𝐛𝐞𝐡𝐢𝐧𝐝 𝐭𝐡𝐞 𝐟𝐞𝐚𝐭𝐮𝐫𝐞𝐬.
While the core messaging must be simple, your power users will want to dig deeper. Be transparent, explain the technology, and share best practices.

Before the launch: manage expectations. We all love a good hype, but it needs strong foundations.

After the launch: move fast and communicate proactively.

Your fans are interested in making it work - co-create with them and help them.

Explore all the Talkwalker-powered insights from this launch here: https://ow.ly/4c2G50Xf6aI

These unfiltered findings are invaluable for PMs analyzing performance of their product launches.

Without the gauge on the actual market feedback, you might as well be shooting in the dark.

In 2025, we learned how to build AI agents.
In 2026, we’ll connect them into teams - to run truly agentic workflows.

Join me at Agent Swarms: From Multiplicity to Market - Driving the GTM Loop, organized by Hard Skill Exchange.

We’ll explore how coordinated agent swarms are reshaping go-to-market.

They are replacing static playbooks with adaptive, signal-driven systems.

These systems align sales, marketing, and success in real time.

In my session - "Agent Swarms for LinkedIn GTM" - I’ll show how my team built a self-hosted multi-agent system that:

→ Researches top-performing content and builds live moodboards in Miro

→ Curates engagement opportunities and delivers them to Slack

→ Identifies ICP-qualified leads from post reactions

→ Analyzes performance weekly and delivers personalized feedback

This framework transforms LinkedIn into a continuous GTM loop ♾️ - where agents drive content, engagement, and pipeline together.

Also on stage:
📍 Fuel Your Agent Swarms with Orchestration and First Party Data by Jonathan M K., Momentum.io
📍 The Architecture of Compounding Intelligence by Amos Bar-Joseph, Swan AI

Apply here (it's free!): https://lnkd.in/dN2BEkZV

2026 will belong to teams - human and agentic alike.

90% of lead scoring models in 2026 are failing for one reason: they are built on generic templates, not your actual data.

10 points for a “big company.”

5 points for a webinar registration.

But if your best customers are actually mid-market tech startups with high growth intent, those "generic" points are actually noise.

The fix? Reverse engineer from your top customers.

Attio lets you build custom scoring models using workflows, auto-enriched data, and integrations.

You can pull in product usage and use an AI agent for web research.

There are 4 layers of lead scoring that predict buying behavior in modern B2B GTM:

1/ 𝐂𝐨𝐧𝐭𝐞𝐱𝐭𝐮𝐚𝐥 𝐟𝐢𝐭.
Who they are before they ever touch your product. Revenue band. Industry. Tech stack. Hiring signals.

2/ 𝐁𝐞𝐡𝐚𝐯𝐢𝐨𝐫𝐚𝐥 𝐦𝐨𝐦𝐞𝐧𝐭𝐮𝐦.
What they do once inside. Frequency of usage. Depth. Activation of sticky features. Time to first value. This is where interest turns into conviction.

3/ 𝐈𝐧𝐭𝐞𝐧𝐭 𝐚𝐧𝐝 𝐭𝐢𝐦𝐢𝐧𝐠.
When buying moves from theoretical to urgent. Pricing revisits. Comparison content. Event attendance. Tight return windows.

4/ 𝐑𝐞𝐯𝐞𝐧𝐮𝐞 𝐏𝐨𝐭𝐞𝐧𝐭𝐢𝐚𝐥.
Estimated annual value. Recent funding. Hiring velocity. Multiple office locations.

With Attio, you can:

✅ Use auto-enriched data from your leads

✅ Bring in product data via integrations

✅ Automate workflows based on your winning patterns

✅ Pull in external signals with web research AI agents

✅ Give sales and marketing one system of truth they both trust

What criteria matter most in your lead scoring model?

#AttioPartner