Organic traffic is not homogeneous, and using a site-wide average is the most common way SEO forecasts go wrong. An informational article converts an order of magnitude worse than a comparison page or a product page; growth concentrated in the blog therefore produces far less revenue per session than the average implies. Segment by intent before modelling anything.
Track conversion by landing page and by query type, and connect it to gross profit rather than revenue. A programme that reports sessions and positions but cannot show profit is a programme nobody can defend when budgets tighten — which is usually when it gets cancelled.