SEO forecasting with real math.

What SEO forecasting is

SEO forecasting is the math of projecting how much organic search traffic, leads, and revenue a specific SEO investment will produce over a defined time period. Done correctly, it's a tool for budget decisions and stakeholder communication. Done wrong (which is most of the time), it produces precise-looking numbers that turn out to be confidently wrong.

The honest version starts from a position of acknowledging uncertainty. SEO forecasting isn't the same as forecasting paid ad performance, where the data feedback loops are tight and the relationship between spend and outcome is measurable within days. SEO forecasting deals with ranking positions that move based on factors outside the agency's control, CTR curves that vary by industry and SERP layout, and conversion rates that depend on offer quality, brand strength, and macro conditions.

That said, "uncertain" doesn't mean "useless." A directionally accurate forecast within plus or minus 40% of actual results is significantly better than no forecast at all. It's also significantly less than the false precision most agency proposals project. This guide builds the honest version with three worked examples using real keyword data.

Why most agencies skip SEO forecasting

Three reasons.

It's harder than it looks. Sourcing reliable inputs for each variable (search volume, CTR curve, conversion rate, close rate, customer value) takes work most agencies don't want to do for free during the sales process. Easier to promise outcomes without showing the math.

It exposes assumptions. A forecast forces explicit assumptions about ranking achievability. "We'll get you to page one" is a vague promise. "We project position 3 for this specific keyword in 9 months based on competitive analysis" is testable. Agencies that don't want to be tested don't forecast specifically.

The numbers are sometimes disappointing. When the math gets done honestly on a low-volume keyword with strong competition, the realistic forecast might not justify the proposed retainer. Agencies who do the math have to walk away from deals or restructure proposals. Easier to skip the math and let the client figure out the ROI question on their own.

None of this means forecasting can't be done. It means honest forecasting is a competitive moat for anyone willing to do it. Sophisticated buyers notice the difference.

The basic SEO forecasting formula

The full formula, with each variable that has to be sourced:

Monthly Search Volume × Expected CTR at Target Position × Website Conversion Rate × Close Rate × Customer Lifetime Value = Monthly Revenue at Full Ramp

Multiply that monthly number by 12 for an annual revenue forecast at full ramp. Then apply a time-to-rank discount for the first year because rankings don't appear day one (more on this below).

Each variable needs explicit sourcing:

• Monthly search volume: from a keyword research tool, ideally one that triangulates multiple data sources rather than relying on a single platform.
• Expected CTR at target position: from Google Search Console data on similar queries the site already ranks on, OR from industry CTR curves if no historical data exists. Adjusted for SERP features (AI Overviews, featured snippets, local packs) that reduce organic CTR.
• Website conversion rate: from Google Analytics or the site's existing conversion data. If the site is new, use 1-3% as a starting range for service businesses and 0.5-2% for ecommerce.
• Close rate: from the sales team's actual data. For most service businesses, lead-to-customer close rate runs 15-35% depending on industry and lead quality.
• Customer lifetime value: first-year revenue is the floor. Multi-year LTV is the ceiling. Use first-year revenue for conservative forecasts and full LTV for the upside case.

Now let's walk it through three actual examples using real keyword data.

The CTR curves that matter

Click-through rate by position varies more than most forecasting guides acknowledge. The numbers below are reasonable starting points for queries without major SERP features, derived from publicly available CTR studies and adjusted for what shows up in 2026 SERPs:

Adjustments to apply

AI Overview present: reduce CTR by 30-50% across all positions. AI Overviews answer the question in the SERP and reduce click-through to source sites.

Featured snippet present: reduce CTR by 20-30% on positions 2-10 if the snippet isn't being captured. Capture position 0 with the snippet and CTR can be similar to position 1.

Local pack present: reduce CTR on traditional organic positions by 30-50% on local queries because the 3-pack absorbs most clicks for "near me" and city-based queries.

Brand or commercial intent on the query: CTR for commercial queries tends to run higher than for informational queries, especially at position 1. Adjust up 10-20% for clearly commercial queries.

For the worked examples below, the assumed CTRs use the table above without major SERP feature adjustments unless noted.

Worked example: SEO for dentists

Let's run the math on a real keyword Whitewater targets: the keyword "SEO for dentists" with 2,900 monthly searches. This represents the agency-side of dental SEO, but the exact same framework applies in reverse for a dental practice forecasting their own SEO returns.

The inputs

• Monthly search volume: 2,900
• Target position: 3 (realistic for a year-one forecast against established competition)
• CTR at position 3: 11%
• Website conversion rate to consultation booking: 1.5% (conservative for a service-business landing page)
• Close rate from consultation to client: 25%
• Average client value (first year): $54,000 (based on Multi-Dentist tier at $4,500/mo × 12 months)

The math

The honest commentary on this number

That looks impressive on paper. Now the caveats:

Position 3 takes 9 to 18 months to achieve from a standing start, not month one. Time-to-rank discount applies to year one (covered below). Real year-one revenue will be 30 to 50% of the full-ramp number, not 100%.

The 1.5% conversion rate assumes a well-designed page. If the landing page is generic, conversion will be closer to 0.5-1%, cutting the forecast by half or more.

The 25% close rate assumes leads from this keyword are well-qualified. Some inbound leads from a "SEO for dentists" search are tire-kickers comparing five agencies. Real close rate on this lead source might be 15-20% rather than 25%.

So a realistic year-one revenue from this single keyword, after all discounts, lands somewhere in the $150,000 to $300,000 range, not the $777,600 ceiling. Still a strong ROI on any reasonable SEO investment, but materially different from the headline number.

Worked example: SEO for attorneys

Next keyword: "SEO for attorneys" with 1,900 monthly searches. Lower volume than the dental keyword but higher average client value because law firm engagements tend to run larger and longer.

The inputs

• Monthly search volume: 1,900
• Target position: 3
• CTR at position 3: 11%
• Website conversion rate to consultation: 1.5%
• Close rate: 25%
• Average client value (first year): $60,000 (based on Multi-Attorney tier at $5,000/mo × 12 months)

The math

Lower keyword volume than the dental example produces a lower top-line forecast, but the higher client value partially compensates. Same caveats apply: real year-one revenue lands in the $110,000 to $220,000 range after time-to-rank discount and realistic conversion adjustments.

Worked example: SEO for electricians

Third keyword: "SEO for electricians" with 2,900 monthly searches. Same volume as the dental keyword but different customer economics on the agency side because electrical contractor SEO retainers tend to run slightly smaller than dental retainers.

The inputs

• Monthly search volume: 2,900
• Target position: 3
• CTR at position 3: 11%
• Website conversion rate to consultation: 1.5%
• Close rate: 25%
• Average client value (first year): $42,000 (based on Multi-Truck tier at $3,500/mo × 12 months)

The math

Same traffic numbers as the dental keyword (because the search volume matches at 2,900), lower revenue because the customer value is smaller. The interesting question this forecast surfaces: if all three keywords get equal priority in the SEO roadmap, the dental and electrician keywords produce roughly comparable traffic but the dental keyword produces more revenue. Stack-ranking keywords by traffic alone misses this.

Position sensitivity analysis

The forecasts above all assumed position 3. Position assumptions drive the forecast more than almost any other variable. Here's what changes if the same keyword ("SEO for dentists" at 2,900/mo with $54,000 client value) lands at different positions:

The forecast ranges from $212,000 (position 8) to $1.97M (position 1) for the same keyword. That's the position sensitivity problem in plain math. A forecast that doesn't model multiple positions hides this entire range.

The honest way to present this to stakeholders: show three positions (conservative, expected, aggressive) and let leadership see the range, rather than presenting a single point estimate that creates false precision.

Adjusting for competition

Position assumptions have to factor in what it takes to get to that position. Three competitive levels to think about:

Low-competition keywords (KD under 20). Position 3 is realistically achievable in 6 to 9 months with focused work. Position 1 is achievable in 9 to 18 months. Most niche or specific commercial keywords fall in this range. The forecast inputs above are reasonable for low-competition keywords with focused effort.

Medium-competition keywords (KD 20-50). Position 3 takes 12 to 18 months. Position 1 may take 18 to 36 months or may not be achievable at all without exceptional content and link work. Discount the forecast for the first 12 months because the ramp is slower.

High-competition keywords (KD 50+). Position 3 takes 18 to 36 months if it happens at all. Forecasts for these keywords should be heavily discounted and treated as multi-year ambitions, not year-one targets. Many high-competition keywords are won by sites with multi-million-dollar SEO budgets and decade-long content investments. Realistic forecasts acknowledge what the site is actually competing against.

For Whitewater specifically, the three example keywords (SEO for dentists, attorneys, electricians) fall in the low-to-medium competition range with KDs in the teens to twenties. The forecasts above are reasonable for a 12 to 18 month timeline.

Time-to-rank discounts

Year-one forecasts have to account for the ramp period where rankings haven't materialized yet. The math:

For a keyword projected to hit position 3 by month 12, a reasonable ramp curve looks like:

• Months 1-3: 0% of full-ramp traffic (site is being built/improved, no ranking gains yet)
• Months 4-6: 15% of full-ramp traffic (initial rankings starting to appear)
• Months 7-9: 40% of full-ramp traffic (rankings establishing, traffic growing)
• Months 10-12: 75% of full-ramp traffic (approaching target position)
• Months 13+: 100% of full-ramp traffic (full ramp achieved)

Average year-one realization: roughly 32% of the full-ramp annual number. So the dental forecast of $777,600 at full ramp becomes $249,000 realized in year one. The math becomes:

This is where SEO forecasting separates from paid acquisition forecasting. Paid ads produce close to full-ramp results from day one (with media efficiency improving over time). SEO produces compounding results that look weak in year one and strong in years two and three. Forecasts that don't account for this make SEO look like a worse investment than it is.

The confidence ranges nobody publishes

Honest SEO forecasting includes explicit confidence ranges, not just point estimates. A reasonable structure for a single-keyword forecast:

Three scenarios. Each with a range, not a point estimate. The honest version tells stakeholders the planning number is the expected case, the conservative case is the floor for budget protection, and the aggressive case is the upside ceiling. Building budget around the aggressive case is how SEO programs get unwound when reality lands in the middle.

Stakeholders actually trust this format more than confidence-faking precise numbers. Showing the math, showing the assumptions, and being explicit about ranges signals that the agency knows what it doesn't know. That's more reassuring than confident-sounding projections that turn out to be wrong.

Common SEO forecasting mistakes

After watching many forecasts get built and watched the results land 12 months later, the same mistakes show up repeatedly.

1. Forecasting only traffic, not revenue. The hardest part is the conversion funnel. Skipping it produces a forecast that can be technically correct on traffic and completely useless for budget decisions.
2. Using single-source keyword volume data. Different keyword tools produce different volume numbers for the same query (sometimes 3x apart). Triangulate across at least two sources.
3. Assuming top-of-funnel content converts like commercial pages. A blog post ranking for an informational query converts at a fraction of the rate of a service page ranking for a commercial query. Forecasting them at the same conversion rate produces wildly optimistic numbers.
4. Ignoring SERP features. A keyword with an AI Overview, a featured snippet, and a local pack might have CTR 50-70% lower than the bare position-based CTR. Forecasts without SERP feature adjustments overstate likely results.
5. Skipping the time-to-rank discount. Presenting full-ramp annual revenue as year-one expected revenue is the single most misleading practice in SEO forecasting. Real year-one results typically land at 25-40% of full-ramp ceiling.
6. Using LTV instead of first-year revenue for conservative numbers. First-year revenue is the floor. LTV is the ceiling. Mixing them up produces overstated forecasts that take 18-24 months to fully materialize even when everything goes right.
7. Forecasting at the keyword level when the strategy works at the topic level. Most modern SEO produces traffic across hundreds or thousands of related queries, not just the head terms. Keyword-only forecasts understate the deeper query traffic that compounds in months 9-18.
8. Presenting point estimates instead of ranges. "We project $850,000 in year one" sounds more confident than "we project $200K-$300K year one with a $600K-$900K year-two range." It also turns out to be wrong more often.

Putting it together for stakeholders

The full workflow for building an SEO forecast that actually holds up:

1. Build the keyword universe. Not just the head terms. The 50-200 queries the SEO program will likely produce traffic on at maturity, with realistic volume estimates from triangulated sources.

2. Group keywords by funnel stage. Commercial-intent queries (where service pages will rank) get higher conversion rates. Informational queries (where blog content will rank) get much lower conversion rates. Forecasting them as a single bucket understates this distinction.

3. Build three scenarios. Conservative (lower positions, lower conversion), expected (target positions, baseline conversion), aggressive (top positions, optimized conversion). Each with a range, not a point estimate.

4. Apply time-to-rank discount. Year one realistic. Year two full ramp. Year three steady-state with continued growth from compounding content. Most SEO programs hit their highest growth rate in year two, not year one.

5. Present revenue, not traffic, as the headline number. Traffic is the input. Revenue is what stakeholders care about. Lead with revenue ranges and back into the traffic numbers that produce them.

6. State the assumptions explicitly. Position assumptions, CTR assumptions, conversion rate assumptions, close rate assumptions, customer value assumptions. Every variable visible. Every number testable.

7. Schedule the forecast review. Compare actual results to forecast quarterly. Update the forecast as new data comes in. Forecasts that get built once and never revisited stop being useful within 6 months.