
Helping brands thrive in the digital age 🌟💡 Proud Gunner #gunners
London
Joined June 2023
- Tweets2K
- Following217
- Followers28
- Likes17.1K
Eugenia Kuyda, founder of @Wabi, pitching "the one friend you wish you had in every conversation" 10 years ago, long before AI agents went mainstream:
Wabi 2.0 is live!
personal agents are one person, one chat. but the parts of life that matter run on more than one person
Wabi is an agent you use together. you talk to it, it does stuff, builds the apps you need, and anyone can join the thread - your partner, your best friend, the preschool moms groupchat
school changes pickup time → Wabi updates the family calendar, everyone sees it.
a show you’d both love → Wabi finds a night that works, gets the tickets.
the trip everyone keeps talking about → Wabi picks the dates, builds the itinerary, books it
go start a thread with your people 👇
Rohan OnlineExpert retweeted
Early @PalantirTech. Cofounder of @anduriltech. GP at @foundersfund.
Last time @traestephens was at @spc was back in 2022.
The world looks a little different now.
Wednesday. RSVP below.
Rohan OnlineExpert retweeted
A few months ago, I told my friend I was growing at $3k a day, but now it has dropped to $1k a day.
So he said that I have to sell Postiz because, at that growth rate, I will soon lose all the momentum.
But he was not the only person who told me that; many said, "sell," "sell," "sell."
And honestly, every time I wanted to sell, it was a bad idea.
Last day (weekend), Postiz grew $5k MRR, and today another $3k MRR.
I have the highest number of trials I have ever had. Bottom line: listen only to yourself.
Rohan OnlineExpert retweeted
🚨 $LITE CEO: NO CUSTOMER HAS ASKED TO DELAY AN ORDER SINCE RATES JUMPED. COMMITMENTS ARE "TO A LARGE DEGREE NON-CANCELABLE," AND 2027-28 GROWTH RIDES ON DATA CENTERS SWITCHING FROM ELECTRICAL TO OPTICAL
On Oct. 16, Lumentum ($LITE) chief executive Michael Hurlston said no customer has slowed, delayed or asked to renegotiate an order since interest rates started rising, and that the company has made it "a lot more difficult for customers to walk away." He spoke to Liz Claman on the first episode of Fox Business's "Claman's Power Players," in a week that closed with the 10-year Treasury yield above 5%.
- Commitments: "The purchase commitments we have are, to a large degree, non-cancelable, so they can't walk away quickly"
- Why: Lumentum runs its own fabs, "sort of the TSMC of laser chips," and invests "3 years ahead of when we might see demand," so it pushes part of that risk onto customers
- Rates: asked whether any customer had slowed, delayed or asked to renegotiate, he said, "We haven't, in fact," and expects hyperscalers to report higher capex this cycle
- Demand: "every day we build, our customers seem to want more"
- 2027-28: even in "a world where data center build outs do slow significantly," he sees "very little impact," because Lumentum's numbers "in '27 and certainly in 2028" will be governed by "that conversion from electrical to optical"
- China: about 70% of transceiver supply comes from China; Lumentum and "one or two other U.S. names" make up 30%, and a U.S. ban would be "a double-edged sword"
Rohan OnlineExpert retweeted
3 questions before i automate anything:
1. does it happen every day?
2. does it touch money or customers?
3. could a new hire do it on day one with written steps?
3 yeses = automate it this week.
2 = automate it this month.
1 or less = leave it alone.
Rohan OnlineExpert retweeted
“Family offices cost 0.5%–1% of assets”
This is a myth according to JP Morgan
The reality is more complex
Yearly cost by Size:
$1B+ AUS: ~$6.6M
$500M–$1bn AUS: ~$3.3M
$251M–$500M: $1,7M
$250M or less: $875k
Cost Distribution:
• 40% spend under $1M
• 29% spend $1M–$2.99M
• 20% spend $3M–$6.99M
• 11% spend $7M+
26% of costs are external
Costs scale with size, but the percentage spent on external providers remains relatively stable
Actual cost depends heavily on:
• Asset complexity
• Governance structure
• Internal talent model
• Generational stage
JP Morgan Global Family Office Report 2026
Rohan OnlineExpert retweeted
Meta's new business terms said it could train AI on content from your website.
After a day of pushback, that line is gone.
If you advertise on Facebook or Instagram, this is worth 5 minutes 🧵
Rohan OnlineExpert retweeted
5 Years and Tesla continues to sell more BEVs than the entire 🇩🇪 German auto industry combined.
Rohan OnlineExpert retweeted
How to diagnose any failing ad:
1. No spend → wrong/boring desire
2. Low hook rate → weak first 3 secs
3. High hook, low hold → weak body
4. High hold, low CTR → no reason to click
5. Clicks, no sales → ad and page don't match
6. Dies in a week → looks like your last 10 ads
7. Good CPA, tiny spend → desire too small
Fix the leak, not the whole ad.
Rohan OnlineExpert retweeted
A masterclass on how to rank #1 on Google in less than 4 months using this 5-step SEO process: xitter.cf/SEOKeval/status/211109…
Rohan OnlineExpert retweeted
the tech stack i use to run my agency:
- outreach ➜ Smartlead
- leads ➜ Arc AI
- booking ➜ Cal. com
- calls & notes ➜ Circleback
- inbox ➜ Superhuman
- team chat ➜ Slack
- client updates ➜ Loom
- projects ➜ Jira
- forms ➜ Tally
- research ➜ Claude
- coding ➜ Codex + Cursor/ Claude
- code ➜ GitHub
- database ➜ Supabase
- hosting ➜ Cloudflare
- automations ➜ Activepieces
- tool connections ➜ Composio
- scraping ➜ Firecrawl
- LLM routing ➜ LiteLLM
- voice & SMS ➜ Twilio
- voice agents ➜ ElevenLabs + Retell AI
- agents ➜ Grok Bot
- virtual assitant ➜ Victor AI
none of these are here because they're trendy.
every one saves me time every single week.
Rohan OnlineExpert retweeted
The dispute with Jensen Huang centres on who sets the limits of artificial intelligence: the companies that create it, or the society that lives with its consequences.
Huang wants safety to be decided by those who make the technology, and Britain will soon discover the value of a government assessment that relies on the same company’s equipment.
The real test will come with the first disagreement: when British researchers conclude that a system is dangerous, but Nvidia or its customers judge existing protections sufficient.
tomorrowsaffairs.com/who-dec…
Rohan OnlineExpert retweeted
10 email flows every store needs (and when to send each one):
set them up once and they work for you every single day.
but please optimize them ongoing.
I decided to build a single section landing page for my project, Foldrr.
So instead of explaining to users about what Foldrr is, i had the idea to create an interactive mockup demo in the hero section like this.
Anyways, still on development. You can get the app on hivinz.com :)
Rohan OnlineExpert retweeted
Ahrefs found only 38% of pages cited in AI Overviews also rank in the top 10 for the same query, down from roughly 76% seven months earlier.
Ranking alone no longer gets you cited.
nealschaffer.com/organic-seo…
Wispr Flow has a different angle for every buyer
Deepstash does the exact opposite
9,662 ads live and almost all of them sell one idea: "the TikTok alternative for smart people"
The only thing they change is the first couple of seconds:
- "My boss who works at Google installed this on my phone"
- "I could KISS the Wall Street guy who told me about this..."
- "me: idk how bro knows everything / also what he scrolls:"
Then the same 13 seconds of the app
Their oldest ads have been running for 170 days
your PM, engineer, and CS teammate all vibe-coded the same feature
designer opens the three versions
same idea
three different products
this is the part AI didn’t fix
@lumenridgehq just put out a free playbook for it
shared components, context, and rules so humans and agents stop inventing their own version of the product every time
designers still decide what actually ships
🤝 Paid partnership
STOP LETTING YOUR AI AGENTS BREAK YOUR PRODUCT.
For the last 2 years, I’ve been building and shipping alongside founders and large enterprise teams at @lumenridgehq, seeing Product Design change forever.
Today, we’re sharing how we work.
Build Your Product Design Harness. ↓
When you fork our AEO/SEO Bot, you get the playbook we run and none of our access.
A Grok Bot template copies four things:
1. The bot's name and job description
2. Its skills, the playbooks it follows step by step
3. Its routines, the jobs that run on a schedule without anyone asking
4. Its identity, so it behaves the way ours does
It leaves behind our computer access, our logins, our chat history and anything the original bot remembers. You connect your own site and sources, so it learns your content instead of ours.
That split is why I'm comfortable open sourcing it. The process and its guardrails travel, including the rule that no page goes live until an editor approves it, and the private data stays with us.
If you lead a marketing team, forking a tested process is a faster start than building a bot from a blank prompt.
I'm covering how we run SEO and AEO together live Wed Oct 21, 11:30 AM PT. The template and registration link are in the reply.






































