Read your first measurement

Check whether your tracked questions have answers on the selected engines before interpreting the headline figures. Coverage reports when each question was last answered, missed slots and pauses.

Before your first reading#

Until a product has its first reading, Visibility shows the product’s first days in place of its figures. The view opens on what is already happening: which engines are answering how many questions.

First check counts each engine’s answers in against the answers owed. Arriving now lists each question with the brands its answers have named.

What arrives when sets out the order: the site checks and questions first; then the first reading, with named rate, recommendation rate and share of voice brand by brand, once the answers are in; moves rated by the answers each could change, with the reading; and changes from day 2.

Getting started counts six setup items, each read from what is stored: site, competitors, competitor sites, questions, answers and site check. Progress is kept across visits.

Prompts, Competitors and Sentiment say no reading yet until then. Answers › History says no readings yet, and fills as days complete.

The first answers are asked when setup finishes, within the plan’s limits. When they cannot be asked at once, setup says when they are due: within the hour, in a number of hours, or on a date.

Readings are computed for whole UTC days, over 7-day and 30-day windows. While a day is open, its reading covers the headline figures and is rewritten as answers are read.

After a day closes, its reading becomes final once its answers are read, with a wait of up to 6 hours. Answers still unread then count as not judged.

Visibility shows its figures as soon as the first reading exists. The first-days view lasts 14 days at most from the start of tracking.

On Free, the questions are asked each day for 7 days, then tracking pauses. The first-days view names the last day as asked until, and everything measured stays readable.

Set up your first product

Start with coverage#

A reading carries the window, the coverage behind it, freshness, notes and reasons a figure is missing. Missing evidence should remain missing rather than becoming a zero.

Move from a figure to an answer#

Read named rate, share of voice and recommendation rate together. Each reading reports an estimated range and the count behind it. Open the underlying answer, citation, figure or question when you need the evidence.

Compare the brands named and the domains cited. A competitor’s mention and a citation to an independent publisher answer different questions: who got attention, and which page informed the answer.

Choose work with a check#

Work items declare what will check the change. A shipped page fix can be checked against the page; work judged by later measurements waits for the window after the ship date to be final.

The programme connects shipped work to its later verdict. Recording the change and its check helps you distinguish work completed from a result confirmed.