How To Get Into A16z’s Super-competitive Speedrun Startup Accelerator Program

Without a doubt, 1 of nan hottest caller startup accelerators successful tech correct now is Andreessen Horowitz’s Speedrun program. Launched successful 2023, nan accelerator has an acceptance complaint of little than 1%. In a January blog post, nan programme said that complete 19,000 startups sounded and less than 0.4% were accepted into nan latest cohort. 

The programme utilized to attraction connected gaming startups, past expanded into intermezo and media, and is now a “horizontal program,” Joshua Lu, nan program’s wide head and a partner astatine a16z, told TechCrunch. Today, founders of immoderate type of startup tin apply, and nan programme runs for astir 12 weeks successful San Francisco. It erstwhile had a programme successful Los Angeles, but Lu said nan attraction will beryllium connected SF from now on. 

There are 2 cohorts a year, and astir 50 to 70 startups are accepted into each. The programme invests up to $1 million into each company, though nan downside is that it’s a spot pricey. It typically invests $500,000 up beforehand successful speech for 10% of nan startup’s institution via a SAFE note, and different $500,000 if nan adjacent information is raised wrong 18 months, astatine immoderate position agreed to by nan different investors. In comparison, Y Combinator typically takes 7% of nan institution for $500,000.

Speedrun said its programme is much “equity expensive” because of what it offers founders. It provides them pinch entree to a16z’s advisory and business networks that assistance pinch tasks for illustration go-to-market, marque development, media strategy, and talent sourcing. Plus it offers nan startups perks for illustration $5 cardinal successful credits to vendors specified arsenic AWS, OpenAI, Nvidia, and Deel.

Given nan precocious interest, and debased acceptance rate, TechCrunch said to Lu for immoderate tips connected really startups tin champion guidelines out. The latest cohort began successful January and will extremity successful April pinch a Demo Day. Applications for the adjacent cohort unfastened successful April, though it looks astatine off-season applications year-round, Lu said. 

Focus connected nan founding team  

Speedrun focuses connected early-stage startups. Because of this, they really analyse who is connected nan founding squad and whether their skills complement each other, Lu said.

“That doesn’t mean 1 has to beryllium method and 1 has to beryllium commercialized and 1 has to beryllium marketing,” Lu said. It intends that “we for illustration not to spot immoderate glaring holes successful capabilities aliases interests. We want nan founding squad to beryllium self-aware and for that to beryllium portion of nan hiring plan.”  

They besides for illustration to spot teams that person worked together earlier aliases person a shared history. 

“There are tons of things that a founding squad has to navigate successful their startup travel and having a spot of shape recognition, being capable to activity pinch each other, knowing really to disagree and really to travel retired nan different broadside of a disagreement, those are each things group connected founding teams pinch shared histories person an easier clip with, connected average,” he continued. 

Even though AI has lowered nan barriers to building software, it’s still incredibly adjuvant for a founding squad to beryllium technical, Lu said. At nan aforesaid time, because AI has made it overmuch faster to build and validate hypotheses and get a merchandise retired there, Lu said nan Speedrun squad likes to spot erstwhile a startup already has a small spot of marketplace validation aliases traction for their product.

“Speedrun arsenic a programme is really awesome astatine helping teams move gasoline connected a very mini spark aliases fire,” he said. “We look for teams that person endeavored to build and effort to show america that there’s a small spark we tin instrumentality nan flames on.” 

Limit nan marketplace “theory”

Lu said 1 communal correction founders often make successful nan exertion process is spending excessively overmuch power talking astir nan marketplace mentation aliases why location is simply a defined problem and why their solution is nan correct one. “All of that whitethorn beryllium true,” he said.

At nan aforesaid time, he added, moreover nan biggest, astir successful tech companies faced unexpected blockades erstwhile they were young, sometimes moreover pivoting completely. What a institution thinks it’s going to build astatine nan opening isn’t needfully what will make it successful astatine nan end.

“What we really want to perceive astir is why this founding squad is really bully together,” he continued, “why they’re a awesome founding team, nan champion imaginable founding squad to lick this peculiar problem.” And past connected apical of that, immoderate validation connected nan thought itself. 

It’s okay to usage AI for nan application, but…

Lu said nan programme encourages each laminitis to usage AI to “clean up” their application. He said location is now nary excuse for grammar errors aliases misspellings fixed nan rising sophistication of AI tools. He besides said AI tin thief founders benignant retired their thoughts, making them clearer, much concise, and much coherent. 

But if AI did each nan activity successful explaining nan startup, that whitethorn backfire. If a laminitis makes it to nan adjacent round, it will beryllium a unrecorded video-call interview. “At that point, their unrecorded narration mentation skills are going to beryllium put to nan test,” he said. So founders should beryllium prepared to talk cogently astir their startup without nan thief of AI.

Only astir 10% of founders make it to nan video-call stage. There are typically 2 to 3 investors connected nan judging sheet astatine a time.

After nan unrecorded interview, nan squad typically conducts a fewer much screening calls pinch nan founders, and past a last determination connected nan cohort is made.

Be greedy to network

There are, of course, different accelerator programs for startups to take from. Lu said Speedrun itself was inspired by immoderate of these different programs. 

Still, he said, this accelerator prides itself connected giving founders entree to a large, specialized operating team. In fact, he said nan champion teams that get nan astir retired of nan programme are nan ones astir “greedy astir getting vulnerability to nan astonishing group and programs” Speedrun has to offer. 

Lu listed disconnected conscionable a fewer points: a16z has astir 600 people, and 10% of that unit is connected nan finance team, he said; everyone other is an usability who supports nan companies nan patient useful with. As a result, founders successful Speedrun will person entree to experts who tin thief pinch marketing, banking, finance, management, and galore different functions. So it helps to cognize who nan startup wants to link pinch and why. 

“We show founders that travel done nan program, what you get retired of Speedrun is what you put into it,” he said. “We deliberation founders who want to return advantage of world experts successful galore different domains early successful their startup travel would beryllium really smart to take us.”

Advice from a laminitis successful nan program

Founder Mohamed Mohamed, who is successful nan caller cohort, just announced a $5 cardinal raise for his proptech startup Smart Bricks led by a16z’s Speedrun. He was attracted to nan programme because he said it stood retired arsenic 1 of nan fewer “explicitly designed for co-founders moving connected frontier AI applications,” and he picked it because he wanted a programme that would let him to “stress-test an eager method vision.”

Mohamed said he treated nan exertion for illustration an soul strategy memo alternatively than a pitch. “Instead of polishing buzzwords, we focused connected clarity — nan existent problem, why it’s structurally hard, and why our squad is unusually well-positioned to lick it,” he said. “We were definitive astir what was working, what wasn’t, and wherever we needed help. I deliberation that honesty and clear articulation of why this problem matters” is what helped nan institution successful nan exertion process.

He called nan full process “rigorous but refreshingly thoughtful,” and said it was designed to understand really founders think, not conscionable what they person built truthful far. “The conversations went heavy into merchandise architecture, information strategy, and semipermanent ambition. It felt person to a partner-level chat than a emblematic accelerator interview, which was a beardown awesome for us,” he said.

His wide proposal is to beryllium “intellectually honorable and precise.” For example, he said successful his exertion he avoided “over-optimizing” for nan liking of hyping up his company. “If you’re vague, derivative, aliases overly protect astir your idea, it shows quickly. Don’t effort to sound bigger than you are; clarity astir wherever you really are is acold much compelling than inflated narratives,” he said.

In nan end, “Speedrun isn’t looking for cleanable companies; they’re looking for founders who tin logic intelligibly astir analyzable problems and build pinch conviction,” he said. “Articulate nan difficult parts of what you’re doing and why they’re worthy tackling. Depth thumps polish each time.”