About RankCause
Every SEO tool shows you that rankings moved. We test whether your change caused it.
That's the one question this category has never answered honestly — and it's the entire reason RankCause exists.
SEO tooling is a mature, crowded, well-funded category. Ahrefs, Semrush, Moz, Screaming Frog, Surfer and dozens of others already do most of what a rank tracker does, and we have no interest in out-spending any of them. We didn't build another dashboard. We built the one thing almost nobody in the category does at all: a real test for whether the change you logged actually caused the ranking move you're looking at.
Domain Authority and Domain Rating are proprietary scores with no published methodology — Google doesn't use them, and most SEOs know it, but the numbers still get quoted in client reports every day. Most "SEO Score" widgets are the same story: a weighted average nobody can defend, shown to two decimal places as if that made it truer. We'd rather show you three data points and say so than show you a confident number built on three data points.
“Your SEO tool has never once told you ‘I don't know.’ That should worry you.”
— Why we built RankCause
How it actually works
When you log a content change on a tracked page, RankCause runs a real before/after test on that keyword's ranking — not a chart you're left to interpret, an actual hypothesis test. Testing ten changes at p<0.05 produces false positives by construction, so every family of tests gets a Benjamini–Hochberg correction before we call anything significant. When nothing clears the bar, the page says so instead of guessing.
Where we forecast a ranking trend, the number carries a confidence interval from a moving-block bootstrap, which accounts for the autocorrelation in daily rank data — a plain confidence interval would be too narrow and simply wrong. Where we compare your click-through rate against what a position predicts, the verdict is gated on a Wilson interval, so two clicks on seven impressions is never reported as "69× better than expected."
We also track whether AI assistants cite your pages — Google's AI Overviews, Perplexity, ChatGPT among them — but it's a supporting signal inside the same causal framework, not the headline. It's gated the same honest way: a page type that no AI assistant has ever cited gets no score at all, rather than a number that looks precise and means nothing.
Three principles
Causal, not correlational
Every dashboard tells you that rankings changed. RankCause tests whether the specific edit you logged is why — with a real hypothesis test and a correction across every test you've run, not a chart you're left to interpret alone.
Numbers with uncertainty attached
Forecasts carry a confidence interval, not a false-precision point estimate. Click-through comparisons are gated on a Wilson interval, so a handful of clicks never gets reported as a multiple of what was expected.
Honest nulls
When a signal isn't measurable yet, we say so and tell you why — never a zero, never a fabricated default.
Who this is for, and who it isn't
RankCause is built for the in-house SEO who has to justify a change to whoever controls their budget, and for the small agency that has to prove what a retainer bought at renewal time. It requires Google Search Console to be useful, and most of what it does gets better once you're actually logging changes to test.
It is not built for enterprise SEO teams right now — we don't have case studies, and we're not chasing that market yet. It's also not a backlink index or a live SERP scraper; we don't compete with Ahrefs or Semrush on either of those, and we won't pretend otherwise.
Where we are today
RankCause is a small, self-funded team. We don't have a sales team, we don't run paid ads, and we have one real production customer today — our own site, which we run the product against every day. If the causal-testing framework holds up against our own numbers, we think it'll hold up against yours.
We're not chasing the SEO market as it exists today. We're building the measurement layer we think it needs. If that resonates with how you think about your own work, we'd love to show you what we've built.
Some specifics, since you asked
RankCause is a small, independently-run team with no sales org and no formal support desk — founder-led support is a feature at this stage, not a placeholder for one. We're not chasing enterprise procurement checklists right now; if that's what you need today, we're probably not the right fit yet. What we do have: a live status page at /status that reports the same health checks our own infrastructure watches, and every customer's data isolated per account at the database level.
You can reach a real person building the product at hello@rankcause.com.
See what RankCause sees
Start a 3-day trial. No credit card. Connect Google Search Console, log your first change, and watch the test run.