Our forecast algorithm considers all the possible variables that could influence non-brand organic traffic, to generate precise and reliable estimates. It generates three key metrics: additional sessions, conversions, and revenue. We calculate conversions by simply multiplying the estimated additional sessions with the conversion rate of the targeted keywords.

Custom conversion rate

We have a 3-level precision approach in place to find the most probable rate - what we call the best-match conversion rate. But if you expect a different conversion rate for the forecast timeframe, you can select Custom to input a custom conversion rate. This will be applied to all targeted keywords in the forecast.

Best-match conversion rate

We try to pull the conversion rate for each keyword from all the data available in the following order:

  1. Keyword-level conversion rates: Organic Traffic processes historical performance data from Google Search Console and Analytics, including keyword conversion rates. This is the most precise, as it is based on actual traffic, and is used if available. In some cases, this historical data might not be available - for instance, a new keyword that is being tracked only for the past few days or weeks will not have historical data.

  2. Landing page conversion rates: When the keyword-level conversion rates are not available in Organic Traffic, we get the conversion rate for the landing page that appears in the SERP for that keyword.

  3. Overall conversion rate for all non-branded keywords: If neither of the above is available, we use the conversion rate of the entire non-brand organic traffic recorded for the last 30 days. This is less precise but still valuable, as it excludes the branded keyword conversion rates, which would significantly distort the estimations.

The most probable conversion rate for every keyword is pulled from one of the three sources above, and this, with the other variables we consider, ensures the most precise calculation of the number of conversions, giving you more reliable forecast estimates.

Did this answer your question?