Read the results

The dashboard

Visibility, share of voice, citations, position and sentiment: what each number means and how it is counted.

Period and filters

The dashboard shows one project over a period of up to 92 days. Every number on it can be narrowed by engine, country, language, topic, tag and branded or non-branded prompts. All tiles, charts and tables use the same filter, so they always agree with each other. The filter is part of the page address, so a copied link opens the same view.

Visibility

Visibility is the share of answers that mention your brand: answers that name you, divided by answers that were received and read. An answer that names you three times counts once. Answers that did not arrive, failed, or have not been read yet are left out, so they never pull the number down.

Next to it you see the change against the previous period of the same length. Botlum shows the change only when the two periods can be compared: the same prompts were asked every day and the answers were read the same way. Otherwise it says why there is no comparison.

Is the change real?

Answers vary from one ask to the next, so a small change can be chance. Botlum labels each change:

Too few answers: fewer than 30 read answers in one of the periods. Wait for more data.
Likely real: the two periods differ by more than chance would explain.
Could be noise: asking the same questions again could easily give a difference this size.

Cited as a source

The share of read answers that cite a page on one of your domains. Below it, named first is the share of answers in which your brand is the first brand named.

Your mentions divided by all mentions of the brands you track: your brand and your competitors. Here every mention counts, so an answer that names you twice adds two. The shares of all tracked brands add up to 100%. The dashboard also shows the counts behind the share, because 100% of three mentions and 100% of three hundred are different news. Adding a competitor changes everyone's share.

Sentiment and context

A language model reads each answer that mentions a brand and judges how it speaks of it: positive, neutral, negative or mixed. It also notes the role the brand plays: recommended, listed among options, compared, or mentioned negatively. The dashboard shows the share of positive answers and the share of answers that recommend you. Sentiment is the model's judgment, so treat it as a reading, not a fact.

Google position

For the Google engines Botlum also keeps the regular results page. Google position is your average position among the top 10 results when you are there. Next to it is how often the AI Overview cites your site.

Engines and trend

The engine breakdown shows each number per engine. The trend chart shows visibility per day, for you and for any competitor, on all engines together or one by one.

The matrix

The matrix lays out visibility by prompt, country or topic against each engine. Each cell has a small daily trend. Click a cell to read the answers behind it: see Answers and sources.

Partial days

Answers are read shortly after they arrive. If some answers of a period are not read yet, the dashboard marks the numbers as partial. They settle within minutes or hours, depending on the engine.