He built for 2 years before raising a dollar—then hit $13M ARR and a $40M Series B. | Alex Halliday, Co-Founder & CEO of AirOps

AirOps bets information gain wins the answer engine era

AirOps is building the SEO of answer engines, helping brands get cited when buyers ask ChatGPT or Claude a question. It began as a spreadsheet workflow tool and moved to marketing teams once early OpenAI models arrived and a Product Hunt extension found users building 50-step content workflows. The bet is that answer engines cite pages which add genuinely new information to the web rather than pages which add more volume.

Takeaways

• Marketers, not product teams, drove the pivot. AirOps began as a workflow tool for product teams stuck in spreadsheets. After early OpenAI models arrived in 2022, a Chrome extension on Product Hunt led the founders to marketing teams building 50- to 80-step content workflows.
• Information gain is what earns AI citations. Halliday argues models win on context rather than capability, so adding materially new information is the theme. Zola publishes wedding pricing data and Klaviyo publishes email data, which is why answer engines reach for them.
• A Deepgram workflow multiplied traffic 35x. A Deepgram CMO built pages for every startup launching on Product Hunt, multiplying the site's traffic 35x. It became AirOps's first cluster of programmatic SEO and the proof point that carried outbound.
• Hiring the first sellers was the inflection point. Halliday and his co-founder sold to about $1.3M ARR themselves with eight or nine people, then handed the process to hired sellers. Enterprise contracts run $60k to $250k, and AirOps reached $13M ARR by Q4 2025.
• Positioning and proof points beat product claims. Coming from consumer, Halliday underestimated positioning. Saying clearly what the company does so it resonates with the buyer takes continuous iteration, and a customer case study outweighs every other detail in outbound.
• They built for two years before raising. AirOps was incorporated during the pandemic and Halliday worked nights and weekends with his co-founder for almost two years, raising money and forming a team at the end of 2022.

The agent-driven content playbook

1. Load every past transcript into AirOps and identify the highest-priority questions on a regular cadence.
2. Ask the host to approve those questions and add his own perspective.
3. Publish content rich with references and links to where people can watch or listen.
4. Audit the off-domain pages that shape ChatGPT's answers, producing a heat map of influence across places like Reddit or YouTube.
5. Work those places through outreach and participation instead of publishing more of your own pages.