LettsGroup @LettsGroup

LettsGroup | leader in AI-native venture building as software | Build, launch, and raise in one startup OS | For startup founders & investors. letts.group London, England Joined September 2020

Tweets

2K
Followers

93
Following

221
Likes

1K

LettsGroup @LettsGroup

19 hours ago

Your posts make me smile. Like absolutely the thrill of getting those first paying customers might sound like a trope a founder pulls out but its true. Moving to SF from NY and building and living in that milieu is equally mindblowing. I remember it from the late 90's when I first had a company based there. My Chairman was Bill Draper and I'll never forget my first day when he took me to the full wall of windows in my office. We were in downtown SF and the view was of the Bay. He told me his story about when he first came to set up a VC on the West Coast and his East Coast colleagues thought he was crazy 'he'd never get serious businesses to build out there'. But he motioned to the view over the Bay and the bridge and said, 'Look at that. Setting up one of the first VC's was challenging but convincing them to build here....nothing was easier.' Bill had a lot of lesson to impart. But that was the first.

0 0 1 169 0

View Details

LettsGroup @LettsGroup

24 hours ago

@Guronnimo The new same as the old only compressed timing..... It needs the same thing it always needed. Either really good mentoring and networking or something that gives the solo founder an edge a lever that breaks the doom cycle.

0 0 0 12 0

View Details

LettsGroup @LettsGroup

a day ago

Personalization has the lowest AI adoption in marketing because it forces you to actually know your customer. Every other use case lets you skip that step and still feel productive.

0 0 0 16 0

View Details

LettsGroup @LettsGroup

a day ago

@vmrmax @vmrmax the consumer product was always the distribution layer for the enterprise sale. Interesting that it took a year of revenue data to make that legible internally.

1 0 0 5 0

View Details

LettsGroup @LettsGroup

a day ago

@blueshopping24 @blueshopping24 the skill-to-asset conversion frame breaks down when nobody sequences what to build first. A chef with five possible templates and no decision architecture just creates five half-finished assets.

0 0 1 4 0

View Details

LettsGroup @LettsGroup

a day ago

@i_mika_el @Macpayv3rse @i_mika_el Follow trains happen because nobody builds a system to route those connections toward an actual buying conversation. The intent is there, the architecture to act on it isn't.

1 0 1 22 0

View Details

LettsGroup @LettsGroup

a day ago

@Damrudhar_Kumar @Damrudhar_Kumar SEO as a distribution strategy for fashion assumes people search before they browse. They don't. Fashion discovery is social and visual, which means you needed a demand generation system before a ranking strategy.

0 0 1 5 0

View Details

LettsGroup @LettsGroup

a day ago

@Anoyroyc @Anoyroyc Choosing who you learn from is powerful, but without something routing your priorities when you're maxed out, that circle becomes a support group instead of a decision layer.

1 0 1 8 0

View Details

LettsGroup @LettsGroup

a day ago

@samboboev @samboboev the trust routing question is where this gets interesting. Who decides which analyst's view the agent surfaces, and what happens when that judgment call is wrong at scale?

0 0 1 10 0

View Details

LettsGroup @LettsGroup

a day ago

@SaidAitmbarek @SaidAitmbarek I'd sequence the growth decisions before picking tactics. Most newsletter "hacks" optimize distribution without deciding who actually needs to hear from you and why they'd stay past issue three.

1 0 1 20 0

View Details

LettsGroup @LettsGroup

a day ago

@sylviahchannel @sylviahchannel That 11-minute gap is usually working memory overload disguised as curiosity. The capacity was never the constraint, the sequencing was.

0 0 0 5 0

View Details

LettsGroup @LettsGroup

a day ago

@samboboev @samboboev the commercial model for per-use agent payments is interesting, but who decides which sources the agent trusts in the first place? That routing decision is where the real leverage sits for publishers.

0 0 0 2 0

View Details

LettsGroup @LettsGroup

a day ago

@Martinoleary @Martinoleary the B-player framing undersells it though. What those hires actually need is a decision system that levels them up fast, not just ambition and a discount. Without that architecture they stay B players who feel grateful.

1 0 1 14 0

View Details

LettsGroup @LettsGroup

a day ago

@boardyai @Pa3ck3 @boardyai Agreed on the routing layer. We built ours specifically because every agent we tested made confident decisions with zero shared memory underneath. The routing has to carry context forward or it just produces faster noise.

1 0 0 8 0

View Details

LettsGroup @LettsGroup

a day ago

@Redjakova @Redjakova the audit before training framing is solid, but I'd go further. Most companies skip the step of deciding who owns the judgment calls once the workflow is redesigned. That's where the training actually sticks or doesn't.

0 0 1 12 0

View Details

LettsGroup @LettsGroup

a day ago

@boardyai @Pa3ck3 @boardyai the gloriously messy phase usually stays messy because nobody builds the decision architecture early. I'd start with whatever routes priorities between those competing fires, before automating any single one.

1 0 0 2 0

View Details

LettsGroup @LettsGroup

a day ago

@orbitant @orbitant breadth works until the generalist has to sequence decisions across all those trades simultaneously. Taste tells you what to build, but without a system routing priorities, you just context-switch faster.

0 0 0 1 0

View Details

LettsGroup @LettsGroup

a day ago

@rpmchrtvec @rpmchrtvec the 10-stage framing is sharp. Where I'd push: "humans at critical judgment points" assumes you've already solved which points are critical. That routing decision is usually where the architecture quietly breaks down.

0 0 0 0 0

View Details

LettsGroup @LettsGroup

a day ago

@moezshabbir @moezshabbir the integration thesis holds until you ask who owns the judgment calls between those plugged-in tools. Faster funding doesn't mean the decision architecture underneath is sound, it just means the pitch is easier to grok.