ArgusTM reads the web the way an assistant does.

It is the scoring intelligence behind every number this product shows you. It has read every business below, on the same checks your own site gets, and it changes its own rules only when the evidence says so.

What it has read so far

Counted from the record, never estimated. The crawler runs continuously, so these move while you read them. Businesses found and queued is larger than businesses measured on purpose: we say what we have not got to yet rather than quietly rounding it away.

What it keeps finding missing.

Every share below is measured across 20,000 real businesses. These are the things an assistant needs and mostly does not find.

Not everything is broken, and saying so is the point: 3 checks are passed by all but a handful of businesses, so Argus scores them and moves on rather than selling you a problem you do not have.

Its rules only move when the evidence moves.

Every month Argus tests its own checks against what assistants really answered. Most months nothing changes, and that is the system working.

1. AskReal questions are put to real assistants, and the answers kept word for word, dated, forever.
2. CompareThe businesses those answers named are scored on the same checks as everybody else.
3. ReplicateA pattern that appears once is a coincidence. It has to appear again, in a fresh round, before it counts.
4. ApproveOnly then may a check earn more influence, and only with a person signing it off.

Nothing here claims a check CAUSES an assistant to recommend you. What is measured is what the recommended have in common, and that is what gets said, in those words.

By trade, measured

Point it at your own website.