The Sizing Problem
GEO and AI visibility investment is a new line in most 2026 marketing budgets. The discipline of MMM-based budget allocation gives the right framework for sizing the investment, but most brands do not yet have an MMM with AI visibility included. The interim question is how much to budget before the MMM produces a response curve to optimize against.
Step 1: Benchmark Against Category Norms
Industry data from 2026: GEO investment is typically 3 to 8 percent of total organic and content marketing budget for brands actively investing, with leaders at 10 to 15 percent. Brands not actively investing are at 0 to 1 percent (often unintentionally, as the work is happening inside SEO and PR teams without being labeled).
For total marketing budget, GEO is typically 0.5 to 3 percent at active brands, with leaders reaching 4 to 6 percent. The smaller percentage versus organic-and-content reflects that paid channels still dominate total marketing spend; GEO competes for a share of the upper-funnel and brand budget.
Step 2: Use MMM-Derived Response Curve Where Available
If the MMM includes AI visibility, the response curve tells you the marginal return on each next dollar of AI investment. Optimal budget is where the marginal return equals the marginal return across all other channels under the total budget constraint. This is the rigorous answer; the budget allocation algorithm computes it.
Step 3: Use the Heuristic Approach Before MMM Maturity
For brands without MMM-derived response curves yet, the heuristic is to size GEO investment proportional to AI search's estimated share of buyer research in the category. If 30 percent of category research happens through AI assistants and your total marketing budget is $10M, sizing GEO at 30 percent of upper-funnel and brand budget is a reasonable starting allocation.
This is rougher than MMM-derived sizing but better than zero, which is the alternative for brands that wait for MMM perfection before starting.
Step 4: Allocate Across GEO Sub-Activities
The GEO budget needs to allocate across: PR and earned media (40 to 60 percent for most brands, because PR drives AI training data inclusion), structured content production (20 to 30 percent), technical infrastructure (10 to 20 percent, including llms.txt, MCP, schema), measurement and tooling (10 to 15 percent, including Presenc AI subscription), and Wikipedia and authoritative source work (5 to 10 percent).
The proportions shift by maturity. Early-stage GEO programs over-weight technical and measurement (foundational work); mature programs over-weight earned media (compounding work).
Step 5: Reallocate From Saturated Channels
Most GEO budget should come from reallocating spend out of saturated channels, especially branded search, retargeting, and bottom-funnel paid social. These channels show high attributed ROAS but low incremental ROAS in lift tests; the reallocation is funded by closing the gap between attributed and incremental ROAS.
Step 6: Stage the Investment
Phase one (quarters one and two): foundational technical infrastructure plus measurement. Phase two (quarters three and four): structured content production at scale. Phase three (quarters five and onward): earned media and Wikipedia work. The staging compounds because measurement enables prioritization and infrastructure enables content production to be operationally efficient.
How Presenc AI Helps
Presenc AI is the measurement layer for the GEO budget. The platform tracks whether the spend is moving the AI signal it is supposed to move and provides the input data for MMM to value the channel against alternatives. Without measurement, the GEO budget is a leap of faith; with Presenc, it is a measurable investment with quarterly accountability.