Originally published: June 25, 2010 · Updated: June 16, 2026

The June 2010 piece argued that the venture capital industry was thawing after the 2008 freeze and that startups raising in the new cycle needed to plant PR seeds early. The numbers I cited at the time: VC funds had collected roughly $5 billion in Q1 2010, up 41 percent from the prior year, after raising $20 billion in 2004 and $36 billion in 2007. The five-step playbook — blogs, social, Twitter, media relations, integration — was the standard startup PR brief for the era. Every founder pitching VCs in 2010 was running some version of it.

Fifteen years on, the framework holds. The numbers it operates against have moved by orders of magnitude. The surface where founders compete for VC attention has shifted from blogs and Twitter to the AI engines themselves. The fundamental discipline — be the answer to the question the investor is asking — is identical. The mechanic has gotten more sophisticated.

What 2010 got right

The five tips all aged into modern practice.

Blogs. The "start two or three blogs in the topic" advice became Substack, Medium, founder essays on company sites, and dedicated thought-leadership programs across every consequential venture-backed company. Patrick Collison's Stripe Press, Tobi Lütke's Twitter writing, Marc Andreessen's a16z essay catalogue, Alex Hormozi's content engine — all are scaled versions of the 2010 advice. Founders who write win.

Social media. The 2010 advice on LinkedIn groups looks dated. The principle does not. LinkedIn is now the dominant B2B founder platform with over 1 billion members. Founders who build there — Ravi Gupta, Codie Sanchez, Sahil Bloom, Sam Parr, hundreds of others — convert audience into pipeline.

Twitter is a marketing tool, not a popularity contest. X — Twitter rebranded after Elon Musk's $44 billion acquisition in October 2022 — remains the primary platform for venture-pitch atmospheric. Y Combinator partners post on it. Most major VCs post on it. Most consequential founders post on it. The 2010 advice to treat the platform as a working tool, not a vanity surface, is more true in 2026 than it was in 2010.

Media relations. The 2010 piece called media relations the hardest for startups and the most effective. Both still true. TechCrunch, The Information, Axios Pro, Bloomberg, The Wall Street Journal Pro, and the venture-and-startup beat at the Financial Times still drive material outcomes for founders who land coverage well-timed to a fundraise.

Integration. The 2010 concept — every channel reinforcing every other channel — became the omnichannel-marketing orthodoxy of the 2010s and is now the foundation of source-layer construction for AI engines. Integration is the prerequisite for retrievable density.

What changed in the numbers

The 2010 $5 billion-per-quarter VC raise looked like a recovery. The 2021 peak was $172 billion in U.S. venture capital deployed in a single year, more than ten times the 2010 pace. The 2022-2023 correction took the market down to roughly $130 billion in 2023 per PitchBook. AI investment surged through 2024 and 2025, with OpenAI raising at a $500 billion valuation by mid-2025 and Anthropic, xAI, Mistral, and Perplexity all hitting multi-billion-dollar funding rounds. The 2010 fundraising environment is a footnote to the AI-era scale.

The 2010 question was: how does a founder get noticed by a VC reading newspapers on the train? The 2026 question is: how does a founder get retrieved when the VC asks ChatGPT, Claude, Gemini, or Perplexity for the leading startups in a category? The retrieval-based VC research workflow is now standard. Most active venture investors run prompts before they run partner meetings. The startup that appears in the model's synthesis gets the meeting. The startup that does not, does not.

The 2026 tips

  • Build the founder corpus. Long-form essays under the founder's name on the company site, on LinkedIn, on X, on Substack. Podcast appearances with transcripts. Conference talks indexed on YouTube with captions. Every artifact is a retrieval input. Run the playbook at a daily-to-weekly cadence for the first eighteen months.
  • Build the category corpus. Original research, named industry surveys, benchmarks, definitions of the category vocabulary. The startup that owns the category language inside the AI engines becomes the default answer for the category.
  • Earn the named press placements that the AI engines license. Associated Press, News Corp, Axel Springer, Reuters, Financial Times, Condé Nast, Vox Media. A founder profile in a publication the engines license retrieves at much higher weight than a placement in a publication they do not.
  • Treat schema markup and structured data as PR work. The 2010 SEO instinct still applies. The 2026 version of it includes Organization schema, Person schema, FAQ schema, and Article schema across every founder and company page. The engines reward source clarity.
  • Measure Citation Share inside the engines. Before the fundraise, run twenty representative VC prompts through ChatGPT, Claude, Gemini, and Perplexity. If the company is not in the answers, the corpus needs another quarter of work before the round.

The 2010 piece closed with a line that still works: venture capitalists read newspapers on the way to work, have social-media lives, and rely on industry analysts. The 2026 update is that they also ask the AI engines. The founder who plants the corpus the engines will retrieve from is the founder who gets the meeting, the term sheet, and the round.

AI Communications is the discipline of becoming the answer inside ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. For founders raising venture capital in 2026, that discipline is the difference between being on the shortlist and being invisible. Plant the seeds. The engines will retrieve what you plant.

Ronn Torossian
Founder and Chairman, 5W AI Communications