Methodology · Documentation

Lost-Leads Calculator — Methodology

The Lost-Leads Calculator on our homepage is an illustrative sizing tool, not a forecast. The dollar figures it produces are modeled from publicly available research on mobile page speed and from the inputs you provide. We make no representation that fixing page speed alone will recover the figure shown — bounce rate is multi-causal, and so is revenue. The calculator's purpose is to start a conversation, not to predict an outcome.

The bounce-rate curve

The calculator's core assumption is the well-documented relationship between mobile page load time and the probability that a visitor leaves before engaging. Google's published research on mobile site speed reports that bounce-rate probability rises sharply with each additional second of load time — roughly a 32% increase from one to three seconds, and roughly 90% from one to five seconds, with the curve continuing to steepen at higher load times.

Source: Google / Think with Google, "Find out how you stack up to new industry benchmarks for mobile page speed." Full article and underlying data: thinkwithgoogle.com/marketing-strategies/app-and-mobile/mobile-page-speed-new-industry-benchmarks . We paraphrase the published findings; we do not reproduce Google's charts or extended text.

We translate Google's load-time curve into a function of PageSpeed Insights score (0–100), because that's the input most small-business owners can actually look up for their own site. A high score (90+) corresponds to a fast page and a small bounce-rate multiplier; a low score (under 50) corresponds to a slow page and a large multiplier.

Formula in plain English

The calculator runs five steps. Inputs you provide are in bold.

V = your monthly Google visitors.
S = your PageSpeed Insights score (0–100). We compute a smooth bounce-rate multiplier B(S) from this score — a continuous curve, not a step function. B(90) is small (a fast site loses few visitors to speed); B(40) is large (a slow site loses many). The curve is calibrated against Google's published load-time-to-bounce-rate relationship cited above.
Lost visitors = V × B(S). This is the share of your traffic the model attributes to speed-driven bounce.
Lost quote-requests = Lost visitors × Q, where Q is your quote-to-visitor conversion rate (i.e. the fraction of visitors who would have requested a quote if they had stayed).
Lost monthly revenue = Lost quote-requests × R × J, where R is your average job revenue and J is your quote-to-job close rate.

This is a deliberately simplified single-variable model. Real bounce rate is multi-causal: offer clarity, trust signals, design quality, copy, brand recognition, traffic source, device mix, and dozens of other factors all push the number around. The calculator isolates one variable — page speed — to make the impact visible. It is not a regression on your actual revenue.

Worked example

Suppose a typical Indianapolis contractor enters the following inputs:

Inputs

Monthly Google visitors (V): 800
PageSpeed Insights score (S): 38
Quote-to-visitor rate (Q): 5% (= 0.05)
Average job revenue (R): $4,000
Quote-to-job close rate (J): 25% (= 0.25)

Computation

Step 1 — Bounce multiplier B(38): ≈ 0.55 (a score of 38 sits in the slow tier of Google's curve)
Step 2 — Lost visitors: 800 × 0.55 = 440
Step 3 — Lost quote-requests: 440 × 0.05 = 22
Step 4 — Lost monthly revenue: 22 × $4,000 × 0.25 = $22,000

Modeled lost revenue: ≈ $22,000 / month

Read this as: "a model based on Google's published curve estimates that a contractor with these inputs is leaving roughly this much on the table each month due to speed-driven bounce." It is not a guarantee that fixing the site will recover $22,000.

Limitations and caveats

Traffic-source mix. Google's data describes mobile traffic broadly. A visitor who searched specifically for your brand will tolerate a slow site far better than a visitor who clicked a generic search result. The calculator assumes generic search-driven traffic.
Vertical-specific bounce baselines. Bounce rates vary by industry — a roofer's site and a SaaS landing page do not share the same baseline. The calculator uses a single curve.
Brand familiarity. Established brands suffer less bounce from slowness because visitors arrive with intent. The model does not adjust for brand recognition.
Offer quality, design, copy, and trust signals. A fast site with a weak offer will still convert poorly. Page speed is necessary but never sufficient.
Mobile vs desktop split. Google's curve is mobile-specific. If a large share of your traffic is desktop, the real bounce impact is lower than the model implies.
Geographic and seasonal effects. Indianapolis traffic in January looks different from Indianapolis traffic in July, especially for seasonal trades. The calculator uses a single monthly number.
Measurement noise at low traffic. At fewer than ~300 monthly visitors, normal week-to-week variation will swamp the calculator's precision. Treat the output as an order-of-magnitude estimate.
No causal guarantee. Improving page speed does not by itself guarantee any recovered revenue. Speed is one of many levers; the calculator quantifies the upper bound of that single lever, not the outcome of changing it.

What we use this for

We use the calculator as a conversation-starter for our free site audit. If the modeled number looks meaningfully large, it usually means the site has measurable speed and conversion issues worth diagnosing in detail — and the audit is where we actually look at your analytics, your funnel, your offer, and your competitive context before saying anything about expected results.

It is not a financial forecast, not a sales promise, and not the basis for any guaranteed-return claim. Any engagement we enter into is governed by a written Master Services Agreement with its own scope, deliverables, and (where applicable) performance language.

Back to the calculator

Last updated 2026-05-23.