Ten blue links used to be the whole game. Now a model reads the web, forms a view, and hands your buyer a single answer. Sometimes it names three brands. Often one. Either way the shortlist is settled before anyone visits a website.
That one shift breaks a lot of assumptions. You can hold every ranking you have and still never reach the buyer who asked, because the question got answered somewhere you weren't.
Slate exists for that gap.
We kept running into the same problem from inside marketing teams.
The standard playbooks — content, technical SEO, keywords, paid — worked, then worked less, and not for any reason the dashboards could explain. Rankings held. Discovery didn't.
What became clear is that being findable and being cited are different problems. Search engines index you. Answer engines have to trust you enough to say your name out loud — and most of that trust gets built off your own site, in places marketing teams rarely look. AEO and GEO aren't tactics bolted onto SEO. They're a separate discipline with their own inputs.
Almost nobody is set up for it. Most brands can't answer the simplest version of the question: when a buyer asks AI about our category, do we come up?
We started Slate in 2026 to answer that question honestly, and then to change the answer.
Two halves, one idea — work natively with AI instead of around it.
The core of our work is AI visibility: the SEO, AEO and GEO that gets a brand cited across ChatGPT, Gemini, Perplexity, Claude and AI Overviews. We map the questions you should own, fix what sits on your site, and build the third-party signals models actually read. For brands that want to move faster, we also run paid campaigns on AI platforms.
The same fluency runs in the other direction. We design and build custom AI agents — support, lead qualification, research, operations — and larger AI systems for teams that need something shaped around their stack rather than bought off a shelf.
Visibility is where most of our work sits. The build side exists because being AI-native shouldn't stop at marketing.
This field is full of confident advice nobody has verified. We run experiments across the major AI platforms and publish what comes back — including results that argue against our own position.
You get the prompts, the raw answers, and the method behind every number. Reports should be checkable, not impressive.
We'd rather run one discipline properly than list six on a website. Depth is what earns citations.
Slate is founded and run by Chirag Aggarwal.
Before starting Slate, Chirag spent five years in marketing and growth roles at early and growth-stage companies — two of them founder's office positions, where the brief is usually to work out why growth stalled and then fix it. Running those playbooks while the ground moved underneath them is what led to Slate.
Every audit is run, read and signed personally. You work directly with the person doing the work — whoever reads your data is the one who writes your plan.
Slate is new, founded in 2026. The advantage of that timing is that the field is still young: models lean on sources they've already learned to trust, so authority built now tends to hold. In most categories there are still questions nobody owns. That won't stay true.