9 Fig is an AI-powered growth engine for consumer brands. We run creative, media buying, landing pages, and measurement as one system, and we measure how those levers interact. For brands that want it operated, we are an agentic + human performance marketing team. For teams that want to own it, we deploy the same engine as marketing infrastructure on their stack.
We produce ad creative at high volume with a generative AI pipeline, then score every ad on creative lift with a patent-pending system rather than click-through rate. We show brands the role each ad plays in their paid media signal and feed what we learn back into the buying.
We plan and buy paid media with budget allocation driven by our recommendation engine. The mechanical parts are automated so senior operators spend their time on the decisions.
We keep the message of the ad matched to the page it lands on, and we run conversion testing inside the media test design so one experiment does not silently invalidate the other.
A unified recommendation engine sits across attribution, marketing mix modeling, and incrementality testing. Its output is not a chart: it is a ranked queue of actions, each with a forecast attached, graded after the fact on a ledger the client can see. The engine speaks through Figgy, our agent that lives in the client's Slack.
A creative refresh changes your media costs. A landing page test retrains your ad platform's algorithm. A promotion pulls demand forward. Single-lever vendors cannot see these effects because each contract stops at the edge of its own metric. Our scope is the whole system, so the interactions are measured instead of invisible.
Most creative is judged on click-through rate, which routinely rewards ads that do not produce customers. Our patent-pending scoring measures each ad's causal contribution, and losing ads are cut even when their platform metrics look good.
Every recommendation is ranked by simulation and forecast before we act, then scored against what actually happened. Wins compound and misses retrain the model. We are the only vendor in a brand's stack whose job includes proving whether we are worth it.
The AI reads every ad, every dollar, and every test continuously. Every consequential call is made by senior operators who built multi-billion-dollar systems at Snap, Roku, Nielsen, and Uber. Clients work with the people who built the engine, not account managers.
There is no login to forget. The engine posts to the client's Slack when something moves and answers questions when asked, and it refuses to answer when the sample is too small to support an honest verdict. Client teams and their own AI tools can plug into the same engine directly.
Brent Davis built marketing and creative sciences at Snap, multi-touch attribution at Nielsen, Variety Streaming Charts, and more.
Sachin Gupta is CEO and leads AI and engineering. He was a founding engineer building ad tech stacks at companies like Roku, YuMe, Uber, and more.
9 Fig was founded in 2025 after the founders spent years building growth and measurement systems inside Snap, Roku, Nielsen, and Uber, and kept watching the same failure from the outside: every vendor hitting their own metric while the brand's blended numbers got worse. 9 Fig exists to own the space between those metrics. Find us on LinkedIn.
Engagements start with a 30-day read of your existing data: we map where money is leaking between your current vendors and hand you the map. From there, brands either have us operate the system, run it alongside their team, or take the engine onto their own stack.
Day to day, communication runs through a shared Slack channel where Figgy, our agent, posts when something in the account moves and answers questions on demand. You also work directly with the founders. Your team keeps control of business parameters, constraints, and overrides at all times.
9 Fig runs creative, media, landing pages, and measurement for consumer brands as one system, and measures how those levers interact. AI reads everything continuously; senior human operators make every consequential decision.
Both, by delivery mode. Brands that want outcomes hire us as an agentic + human performance marketing team. Teams that want to own the capability get the same engine deployed as infrastructure on their stack, with their people operating it.
No. We embed alongside whoever you keep. If your creative shop is great, our measurement will prove it, and it will show you what their work does to your media performance, which they cannot see from inside their own scope.
Brands spending $10M or more per year on paid media. We solve complex problems when levers move simultaneously and the interactions between them become the biggest unmanaged line item.
By creative intelligence: each ad's measured contribution to actual customers, using a patent-pending scoring system. Click-through rate and other platform metrics are not used as success criteria, because they routinely reward ads that do not produce buyers.
Pricing is custom and depends on scope and engagement mode. Every engagement starts with a 30-day read of your existing data, and the accountability ledger means you can see what our recommendations were worth.