Texas Tea Party

@Noicewon11 per disclosures, Meta seems to have expanded spend significantly the last few quarters. not intuitive given ScaleAI “ownership” valuation makes no sense given business quality thoughts on if interesting now?

No one talking $INOD after this + $META head of AI leaving. stock @ 32. I have FV at ~5

few $FIG / design workflow thoughts wont get into most recent quarter - but you should be asking about whether there were any one-time tailwinds from (1) starter -> pro upgrades (MCP tool call limits + daily Make credit limits on starter that don't exist on Pro/Org/Ent) and (2) seat increases from "ghost" seats as cheaper to buy an extra pro seat than buy AI credit add-on. and (3) note the risk factor changes re Make margins the big shift that will drive stock / longer term demand is how well FIG adopts to the new world = coded design components > visual components – at its core, Figma is a translation tool to bridge what designers want to build and what developers actual build. the end goal is to generate the right code – FIG was crucial in the design flow bc it was multiple times faster / easier to adjust UI using drag & drop on their canvas, vs do this in code – AI tools have made it just as fast to design with code, and make edits with a single prompt. code is native to LLMs there's a reason why design workflows start with prototyping right now. skips the tedious sketching out of screens and enables teams to get to the end "vision" quicker – teams are now going backwards into Figma to build out the design file due to org inertia and process – this process is broken - you built a working thing, then documented it in a static mockup tool, so then it could be handed to developers to rebuild it – the middle step is redundant and gets solved if either (1) org inertia breaks down and processes emerge around the prototype vs Figma file, (2) the prototype can be created with high fidelity referencing a coded design system and with code that is high quality enough to be directly pulled into a MR/PR and moved into CD dev process, or (3) better yet if the prototype is built on existing code base code is the end goal, and appears to be the new source of truth in design. as roles converge (PM, designer, engineer), this could become the natural path

@Av1dlive do the speech to speech models remove the need to include twilio?

nice article. areas of tension in the framework as i see it: (1) cost to write SW -> 0, but cost to innovate will weigh on incumbent saas margins, (2) assumes that incumbent saas must use frontier model and token tax is large. seeing OS model gap shrink and value in harness improving to improve OS model competitiveness vs frontier models. of course incumbent saas must have proprietary context to construct their harness appropriately, but i think market overreacting to margin compression longer term.

We just published our SaaS Primer: The New Industrial Order. The software industry is now locked in a brutal Red Queen’s Race. Companies are being forced to spend heavily on AI just to defend their current market share, which inevitably crushes margins and bleeds cash. To help navigate this shift, we outline three high conviction stock ideas: two in the U.S. and one in Japan. Each one offers a way to win during this transition instead of being consumed by it.

maybe. i actually don’t think it’s a layup. last year, this was a low 120s NRR business last year (excl pricing) and that was mostly seat expansions. seat expansions are collapsing. so at $1B of rev, going to take a lot of Make credits to reaccel in face of this. and customer feedback on great on Make

@Texas_Tea_Party they are going to accelerate in Q2 also w/first full quarter of Make. don't think you can short w/that as the setup.

I've been FIG hater, but Make-driven acceleration likely in Q1/Q2

people know FIG will capture some amount of it. reality is FIG has gone from a n of 1 clear market leader -> to now competing in a crowded market where they also have to compete against their suppliers. oh yeah and long term profitability looks very different too given GMs of essentially repackaging LLM inference tokens with a nice UI

It's pretty funny that there's like 5 vibecoding startups growing 10x+ YoY and the market thinks FIG is not going to capture any of that incremental spend.....by the way, NDR expanded 5% last quarter. There's a key difference between a company like ServiceNow who's core market does not benefit from infinite intelligence and Figma, who's customers would of course like to create more delightful products.

@envoy_capital this was a nice one - congrats

$MEDP (Medpace) - earlier short I put out last week, thesis was growth decel from "GLP-1" downturn Stock down 20% today on Q1 print from weak book-to-bill numbers, likely reflecting continued slowdown and cancellations in metabolic given "GLP-1" exposure. Wouldn't be surprised if management points to that in tomorrow's call Top-line guidance numbers unchanged but should be increasingly harder to reach based on these booking numbers Surprised to see Jesse Geiger (President) leaving as well, him and Troendle were noted to be very well connected in the metabolic space so timing doesn't look great Full write-up here: envoycap.substack.com/p/short-medpac…

@MetacriticCap customer spend inflecting big time as more focused on software supply chain security top of mind and AI coding tools increasing number of artifacts FROG manages (usage-based pricing) and set up very nice with guide sandbagged

Why $FROG trades at 7x Sales??

@MRatable agree. ESTC consumption but looks increasingly troubled across search and observability. funny how obsessed the market was with RAG

I am increasingly convinced that the seats vs. consumption dichotomy is meaningless

this will age like deli meat in the sun you do realize that Figma Make is a wrapper around Claude, right? they have gone from being a clear market leader to competing against a new vibecode start up launched weekly.. and now they’re competing against their supplier. how’s that going to work out

Unpopular opinion: Claude Design is *good* for Figma. This isn't a bundled loss-leader (Teams/Zoom); it is an extremely token-expensive UI layer on top of Claude. Imo it creates a Figma price umbrella by mentally preparing CFOs to spend far, far more on design software.

@eshita then you haven’t spoken to any (former) customers. pick your head and ask around

Figma is staying as intentional as possible as these AI design tools come out. In the beginning, it seemed obvious that Figma should emulate parts of Photoshop in the browser, but the team stayed intentional and built toward how designers actually work and that was the right choice The selloff right now is fear-based. I don’t think Figma is in any trouble.

there’s data gravity in FIG for existing projects (eg adding a feature to an existing app), but none for new projects bulls betting on Figma Make to save the day is delusional. go ask WIX how that worked out with Base44. worthless revenue and competitive set now way worse than when Figma was clear market leader

@Texas_Tea_Party Agreed - that is the risk. Workflow/existing data gravity of fig vs creation (esp at low end) at LLM layer

If $FIG moat is multiplayer/network effect of designers working together on a project, is there argument that explosion Claude/OAI/Gemini content = more Figma subs and sign ups including casual users?

@bengold if Figma Make isn’t successful, the business is screwed. seats are coming down in design orgs / fewer needed as AI prototyping gets incorporated in the process if they don’t get the usage revenue = no bueno

Figma will be fine, but they should really stop wasting time on Figma Sites and Figma Make. Those products will never be successful.

@FiscalDaddy @TwannsWorld truly a joke. and they are ruining tegus content quality with all the AI-led interviews… API or die in this new world, they need to figure it out bc closed-ecosystem won’t cut it. people will build their platform in-house and it’s already happening

@TwannsWorld AlphaSense one of the worst products I’ve ever used. Somehow they own BamSEC but AlphaSense can’t load files at the same speed. Daloopa running circles around Canalyst after they bought that. AI offering is shit & won’t open it up to Claude.

Bloomberg is so silly there’s a 25% uplift at signing via AlphaSense + Quartr + tokens in perpetuity via everything else if they just made it open w/ a better UI

@based16z @ContrarianCurse yep -> $0

@ContrarianCurse x.com/zoink/status/2…

I can FEEEELLLLL the skittish in software $FIG commons for now

fast followed by Anthropic CPO resigning from FIG board effective immediately… 💩

@MetacriticCap @lefttailguy no innovation, Make sucks, leaning on price increases for growth, large companies are moving away in real time… FIG in a real tough spot

@lefttailguy You can just ask Claude to write the prototypes for you. Ask it to write 20 options. Doesn't matter. $FIG is a zero.

This is the real risk to Microsoft. I thought for a while that in an economy dominated by digital products Figma would be the new Excel/PPT, but turns out its just Claude Code.

@lfg_cap $FIG short

Starting a regular thread. What is your favorite 6 months single stock trade right now? I’ll start👇