From an internal ad prototype to Gravity, a spinoff that has raised $39M
An AI tool we built to run ads for our own clients became its own company. I helped shape its strategy and positioning on the way out the door.
Flax LabsCOO & AdvisorDec 2024 – May 20263 min read- $9M
- Private equity investment
- $39M
- Total raised to date
- 4x+
- Average client ROAS from the tooling
Context
Inside Flax Labs we built a prototype: AI software that analyzed, built, managed and optimized ads across all of the major media-buying platforms. It began as a way to make our own team more efficient, and it was proven on real client accounts before anyone thought of it as a product.
The results were strong enough that it spun out as its own company, Gravity (trygravity.ai). I helped the team strategize and position it rather than build the tool itself. Gravity secured $9M in private equity investment and has since raised $39M in total. It has found its footing in advertising inside AI assistants such as Claude and ChatGPT.
The thesisThe strongest products start as tools a team cannot live without. Prove the value on your own work first, then sell it.
What the prototype did
The prototype was four separate agents, each doing a job a media buyer does every day: analyzing ads, adjusting bids, alerting the team by text message when a campaign needed attention, and creating and optimizing ads. Together they became the foundation of Gravity. People stayed in the loop to steer strategy and check the work, which is what let a small team manage so many accounts without quality slipping.
From tool to company
- 01Built for ourselvesAn internal prototype to make our media buying faster and more consistent.
- 02Proven on clientsRolled out to client accounts, starting small, averaging 4x+ ROAS.
- 03CommercializedNew sales process, team structure, account strategy and growth-aligned pricing.
- 04Spun outBecame Gravity, backed by $9M in private equity; $39M raised to date.
Each stage had to earn the next. The tool proved itself internally, then on client accounts, where it averaged 4x+ ROAS, before it was worth spinning out. Commercializing it changed more than the product: it meant rethinking the sales process, which teams were client-facing, how accounts were assigned and strategized, and a pricing model that earned less up front and much more as clients scaled.
Build the tool your own team needs, prove it on real work, and the product strategy writes itself. Positioning is what turns a useful tool into a company.