Calculate Exactly What Your Slow Website Wastes in Ad Spend Each Month
Here is the exact model. Run it on your numbers.
This is not a warning about web performance in the abstract. It is a self-service arithmetic tool with five steps, a worked example at realistic Tampa Bay SMB spending levels, and a specific dollar figure at the end. The number you get is a floor estimate, the real waste is almost certainly higher, but it is the one you can calculate today, before you call anyone.
You need two things: your Google Ads dashboard and a free mobile speed test. You need about ten minutes.
What This Model Measures, and What It Does Not
Before the math, one honest framing note: this is an illustrative model, not an empirical study. It applies documented population-level statistics to your specific account using stated assumptions. Your actual waste may be higher or lower depending on your industry, your traffic mix, and how badly your mobile page is underperforming. The model gives you an order-of-magnitude estimate and a direction. That is enough to decide whether the conversation is worth having.
The model does not capture Quality Score degradation, Meta ad feedback loops, or brand trust erosion from repeat slow experiences. Those are real costs, you will see a note about them at the end, but this model measures only the one cost that is directly arithmetic: clicks you paid for that bounced before converting.
Step 1: How Many Clicks Are You Buying Each Month?
Open Google Ads. Pull your total monthly spend and your average cost-per-click (CPC) for the past 90 days. Divide spend by CPC.
Your formula:
Monthly clicks = Monthly spend ÷ Average CPC
If you do not have a Google Ads account or want an industry benchmark to pressure-test your own number: WordStream’s 2025 Google Ads Benchmarks report (16,446 US-based campaigns, April 2024 through March 2025) put the median CPC at $5.26. Their 2026 report (13,474 US-based campaigns, April 2025 through March 2026) put it at $5.42, confirming that click costs have more than doubled since 2016, when the median was $2.32.
Worked example: Tampa Bay home-services company:
Monthly budget: $4,000 Average CPC: $5.26 (using 2025 WordStream median as fallback) Monthly clicks purchased: 760
Step 2: How Slow Is Your Mobile Landing Page?
Go to PageSpeed Insights. Enter the URL of the page your Google Ads traffic lands on, not your homepage, the actual landing page. Run it on mobile. Record the number in seconds under “Largest Contentful Paint,” which is the metric that best represents how long a real visitor waits for your page to visually appear.
You are looking for one number: your current mobile LCP in seconds.
LCP: Largest Contentful Paint, is how long it takes for the biggest visible element on your landing page (typically a headline, hero image, or offer block) to fully appear on screen. It is the moment a visitor decides whether the page is loading or broken.
Worked example: LCP on mobile: 3.5 seconds
Step 3: Apply the Bounce Multiplier
Google published data from its internal mobile analytics in 2017 showing that bounce probability increases 32% when page load time goes from 1 second to 3 seconds. This figure was released by Google and has been cited extensively in Search Engine Land and Google’s own research publications. It is a 2017 dataset, noted here as a labeled historical anchor, and it establishes the directional relationship this model uses.
The interpretation: a visitor landing on a 3-second page bounces at a rate approximately 32% higher than the same visitor on a 1-second page. For a 3.5-second page, the bounce uplift is at minimum in that range, and likely higher.
How to apply it to your click count:
Incremental bounce rate from speed = 32% (as a multiplier on baseline bounce probability) Additional bounced clicks = Total monthly clicks × 0.32
Worked example: 760 clicks × 0.32 = 243 additional bounced clicks per month
These are clicks that would not have bounced if the page had loaded in under a second. They are sunk spend the moment they land.
Step 4: Apply the Mobile Conversion Impact
Google Ads Help: Google’s own primary documentation, states explicitly: “A 1-second delay in mobile load time can impact mobile conversions by up to 20%.”
The conversion rate (CVR) baseline for context: WordStream’s 2024 report (17,998 US-based campaigns) measured a median CVR of 6.96% across Google Ads accounts. The 2025 report (16,446 campaigns) measured 7.52%. These are what you expect to convert from a properly-performing page.
When your page loads slowly, you are applying a 20% reduction to an already-thin CVR. At 7.52% CVR, a 20% impact drops your effective conversion rate to roughly 6%. The difference: 1.52 percentage points, may sound small. Multiplied across hundreds of clicks and months of spend, it is not.
Worked example using bounce figure only (conservative model):
The arithmetic here uses the bounce multiplier from Step 3, because bounced clicks are the clearest and most direct form of pure waste: the visitor arrived, the page took too long, they left. No conversion opportunity existed at all.
243 bounced clicks × $5.26 CPC = $1,278 per month in sunk spend
Step 5: Calculate Annual Waste
Monthly waste × 12 = annual figure.
Worked example: $1,278 × 12 = $15,336 per year
That is the floor estimate for one Tampa Bay home-services company running $4,000/month in Google Ads with a 3.5-second mobile landing page. The model uses a single bounce multiplier and a documented CPC. It does not include the additional CVR suppression from Google’s 20% mobile conversion impact figure. It does not include Quality Score degradation. It does not include Meta.
To run this on your own numbers:
| Input | Your number |
|---|---|
| Monthly Google Ads spend | $ _____ |
| Average CPC from dashboard | $ _____ |
| Monthly clicks (spend ÷ CPC) | _____ |
| Mobile LCP in seconds (from PageSpeed Insights) | _____ s |
| Incremental bounced clicks (clicks × 0.32) | _____ |
| Monthly sunk spend (bounced clicks × CPC) | $ _____ |
| Annual sunk spend (× 12) | $ _____ |
“Our CVR Is Above Average: Does This Still Apply?”
Yes. Here is the math.
If your CVR is double the WordStream median, say 15%, which puts you in the top-performing tier of Google Ads accounts, you have 760 clicks and a realistic conversion count of 114 per month.
The 243 bounced clicks from Step 3 are gone before they ever had a conversion chance. Those 243 bounced clicks would have produced, at 15% CVR, approximately 36 additional conversions per month if the page had loaded in under a second.
At $5.26 CPC, that is still $1,278/month in lost clicks. Above-average CVR makes each recovered click more valuable, not less. The model holds harder, not softer, as your CVR improves, because each click is worth more.
What This Model Is Not Capturing
Three categories of additional cost sit outside this arithmetic.
Quality Score degradation. Google’s Ad Rank, the formula that determines your ad position and your actual CPC, factors in landing page experience. Poor landing page experience, which Google measures partly through load speed and user behavior signals, degrades your Ad Rank. A degraded Ad Rank means you either pay more per click to maintain your position or lose position entirely. Google Ads Help and Google’s Quality Score documentation both confirm that landing page experience is one of the three components shaping the quality factor in Ad Rank. This model does not quantify that cost, it is real but harder to isolate arithmetically for a specific account.
Meta feedback loops. Meta’s ad delivery system tracks post-click behavior. Visitors who bounce fast from slow landing pages send negative quality signals back to Meta’s system. Meta’s diagnostic system includes engagement rate ranking and conversion rate ranking as components that affect how competitively your ads are delivered. This model covers only Google Ads arithmetic.
Cumulative re-acquisition cost. Visitors who bounce and never return have to be re-acquired as paid traffic. The cost of that re-acquisition cycle compounds monthly. That math is in a separate article in this series.
The Actual Floor
The $15,336 annual figure in the worked example is a floor. It accounts for one measurable mechanism, incremental bounced clicks, and ignores at least three others. The realistic total for a $4,000/month Google Ads account with a 3.5-second landing page is almost certainly higher.
The model is designed to show you that the conversation is worth having, not to give you a final audit number. The final number requires looking at your actual CrUX field data, your Quality Score components, and your landing page analytics together.
Run Your Audit Now
Gabriel Lopez-Seco runs a free assessment at iServU that pulls your actual CrUX field data, your landing page speed measurements, and your ad account quality indicators together into a single picture. You leave knowing exactly what your slow page is costing you in paid spend, not a model, your actual numbers.
No account manager. No handoffs. One call.
Schedule your free assessment at iServU.
Sources
- WordStream / LocalIQ (2025). Google Ads Benchmarks 2025: 16,446 US-based campaigns, April 2024–March 2025. https://www.wordstream.com/blog/2025-google-ads-benchmarks
- WordStream / LocalIQ (2026). Google Ads Benchmarks 2026: 13,474 US-based campaigns, April 2025–March 2026. https://www.wordstream.com/blog/2026-google-ads-benchmarks
- WordStream / LocalIQ (2024). Google Ads Benchmarks 2024: 17,998 US-based campaigns. https://www.wordstream.com/blog/2024-google-ads-benchmarks
- Google Ads Help (evergreen, confirmed 2025). Evaluate the performance of your landing pages: “A 1-second delay in mobile load time can impact mobile conversions by up to 20%.” https://support.google.com/google-ads/answer/7543502
- Google Ads Help (evergreen, confirmed 2025). About Quality Score, landing page experience as a component of Ad Rank. https://support.google.com/google-ads/answer/6167118
- Google / SOASTA (2017, HISTORICAL). Mobile page speed benchmarks, bounce probability rises 32% when load time goes from 1 second to 3 seconds. Cited via Search Engine Land: https://searchengineland.com/people-wait-website-load-2024-stat-445223
