Forecasting Methodology

Step 1: We Pull Your Data Connect your Google Search Console or Bing Webmaster Tools. We fetch up to 16 months of historical traffic data - clicks, impressions, and patterns. Step 2: Pattern Detection Our engine analyzes your data to find trends (going up, down, or flat), weekly patterns (which days perform best), and seasonal variations (holiday peaks, summer dips). Step 3: Forward Projection Using these patterns, we project your traffic forward for your chosen horizon (1-12 months) with confidence ranges showing best and worst case scenarios. Data Freshness: Google Search Console has a 3-day delay before data is complete. We automatically exclude the most recent 3 days to ensure accurate forecasts.

The more historical data available, the more reliable your forecast. Minimum requirements vary by how far ahead you want to predict:

90% Confidence: There's a 90% chance actual results fall within the shaded range. Reading the Range: • Upper Bound - Optimistic scenario. Useful for setting stretch goals. • Center Line - Most likely outcome based on current trends. • Lower Bound - Conservative scenario. Use this for budget planning. Why Ranges Vary: • Further into the future = wider ranges (more uncertainty) • Volatile traffic history = wider ranges • Stable, predictable patterns = narrower ranges Pro Tip: When reporting to stakeholders, present the range rather than a single number. It builds trust and sets realistic expectations.

Each forecast includes a Confidence Score so you know how reliable it is. We convert the technical MAPE (Mean Absolute Percentage Error) into a simple 0-100% score where higher is better. A score of 85% means forecasts are typically within 15% of actual values:

Choose how far ahead to forecast based on your planning needs:

Forecasts can be viewed at three levels: Weekly: 7-day totals (Monday to Sunday). Best for short-term monitoring and detecting patterns. Monthly: Calendar month totals. Ideal for monthly reporting and comparing month-over-month performance. Quarterly: Three-month totals (Q1: Jan-Mar, Q2: Apr-Jun, etc.). Best for strategic planning and year-over-year comparisons. Partial Periods: When a forecast doesn't cover a complete period, we pro-rate the values so you can fairly compare them. Partial periods are clearly marked.

Run a new forecast: • Monthly, to track how projections evolve • After major Google algorithm updates • After significant site changes (migrations, redesigns) • When launching major content campaigns Compare before/after: Run a forecast before making changes, implement your strategy, wait 2-4 weeks for data, then compare actual results to what was projected. The difference shows your impact.