You ship a page, a schema fix, a comparison guide. Then what? A campaign pins a baseline the moment you start, then measures whether AI visibility moved against it, and tells you plainly when a change is real and when it is just sampling noise.
Tag the prompts a campaign should track and llemmy locks a starting point. Everything after is measured against where you actually began.
Every impact figure carries its sample size and a 95% confidence interval, with a plain "significant" or "within noise" tag from whether the intervals overlap.
Visibility, share of voice, sentiment and citation rate, each rebased to day 0, so you see the full picture of what your content moved.
A single "you're up 9 points" number is easy to fool yourself with. A campaign compares current to your day-0 baseline for each metric, shows the sample size behind both, and only calls a change significant when the two 95% confidence intervals do not overlap.
The trend is rebased so day 0 is your baseline, and every point is a daily bucket since the start. Drop markers on the timeline for when you published, when a page was first cited, or any event, and they snap onto the chart so cause and effect line up.
A campaign is not just a chart. It is the plan behind it. Add the tasks a campaign depends on, assign them to teammates, and move them from to-do to in-progress to done, right beside the impact they are meant to produce.
Spin one up from a content opportunity in your data, a GEO audit finding, an onboarding step, or from scratch. Describe a topic and llemmy drafts campaign ideas: a name, a description, and prompt ideas you can add in one click.
Pick a visibility or topic gap surfaced in your AI data and llemmy drafts campaign ideas from it, ready to track.
Turn a fix from your GEO audit into a campaign that tracks whether shipping it actually moved visibility.
Frame a getting-started step as a campaign, so your first weeks of GEO work are measured from the start.
Describe a topic or gap and llemmy drafts a name, description and prompt ideas to track. You keep, edit or discard them.
Campaigns measure content effectiveness over time with confidence intervals, not vanity numbers. Presence is a rate with sample size (n) and a 95% confidence interval over recent answers, measured over the official model APIs and Google AI Overviews read off the SERP. A confidence interval tames sampling noise; it does not erase the biases we name plainly in our methodology. When a change stays within noise, we say so.
The mention rate, with confidence intervals, that a campaign moves.
Audit findings become campaigns that track whether shipping them worked.
One campaign on the free plan, up to unlimited on Enterprise. Baseline captured the moment you tag prompts. No credit card.