Faster Pages Don't Just Sell More: They Sell Bigger
Most business owners who have heard about website speed think about it as a conversion problem: slow page, fewer buyers. Fix the speed, get more buyers. That’s the whole model.
The model is incomplete.
A 2020 Deloitte study commissioned by Google: “Milliseconds Make Millions”, analyzed 30 million mobile user sessions across 37 brands in retail, travel, and lead generation. It found two things happen simultaneously when mobile load time improves:
- More visitors convert.
- The ones who do convert spend more.
In the retail segment, a 0.1-second improvement in mobile load time correlated with an 8.4% increase in conversion rate and a 9.2% increase in average order value. Both. In the same data. At the same time.
When you run those two effects together, the revenue impact is not additive, it is multiplicative. More buyers, each spending more, compounding each other.
Why This Changes the Calculation
Here is the arithmetic for a hypothetical Tampa Bay boutique retailer with an online store.
Baseline: 500 monthly orders at $95 average order value = $47,500 per month.
Now apply the two effects from the Deloitte data conservatively:
- Conversion rate lifts 8%: 500 orders become 540 orders.
- Average order value lifts 9%: $95 becomes $103.55.
Result: 540 × $103.55 = $55,917 per month.
Difference: $8,417 more per month. Over a year: $101,004.
That arithmetic uses the Deloitte 2020 retail figures applied to a small business at plausible Tampa Bay scale. It is illustrative, not a guarantee, but the compounding mechanism is real, and it is the reason the revenue impact of speed work tends to be larger than people expect when they first run the numbers.
The explicit caveat: the Deloitte study is from 2020 and covers a specific baseline, mobile sessions at the slow end of the performance spectrum at that time. “A 0.1-second improvement” describes a marginal gain from a slow starting point, not a universal rule that applies equally at any speed. Do not scale this figure by assuming 0.2 seconds of improvement produces 16.8% more conversions. The relationship is not linear in that direction. The 2020 data establishes that the dual effect exists; it does not provide a precision formula.
What Current Data Shows
Because the Deloitte study is now five years old, it is worth asking whether more recent data supports the same direction.
In 2025, Yottaa published an interactive Web Performance Index built on 500 million shopper sessions across 700-plus ecommerce brands, daily-updated data as of September 2025. That dataset shows LCP improving from 2.5 seconds to 1.3 seconds correlates with a 50% increase in conversion rate, measured across the full session sample.
Yottaa’s data focuses on conversion rate rather than average order value, it does not replicate the AOV finding directly. But the conversion magnitude is substantially larger than what the 2020 Deloitte data showed, suggesting that current performance gaps may carry an even larger opportunity than the historical figures imply.
Contentsquare’s 2025 Digital Experience Benchmarks report, published January 28, 2025, covering 90 billion sessions across 6,000 websites globally, found that conversion rates dropped 6.1% year-over-year in 2024 while ad spend rose 13.2%. The report identifies slow-loading content as a primary driver of what it terms “frustrated sessions,” which accounted for 40% of all visits. More ad spend, fewer conversions, slower pages: the combination describes a business paying more to acquire visitors it then fails to keep.
The direction of all three data sources: Deloitte 2020 (historical), Yottaa 2025 (current), Contentsquare 2025 (current), is consistent.
Why Average Order Value Moves With Speed
The conversion-rate link is intuitive: faster page loads mean fewer visitors abandon before they can buy. But the average order value finding surprises most business owners. Why would someone spend more just because a page loaded faster?
The mechanism runs through session depth and upsell exposure. A fast, fluid site keeps users in the session longer. They browse additional products. They see the recommended add-ons. They reach the “complete the look” or “frequently bought together” section. They don’t abandon the cart during checkout because the payment form took two seconds to respond to a credit card entry.
A slow site shortens every session. Visitors do the minimum required to purchase and leave, if they purchase at all. A faster site allows the full purchase journey, including the parts that increase order size.
This is not a theoretical mechanism. The Deloitte data captured it in 30 million real sessions across 37 brands. Average order value moved alongside conversion rate, in real purchasing behavior.
The Metrics That Drive This
Three Core Web Vitals are currently measured by Google in real user field data. All three affect session depth and purchase behavior in different ways.
LCP (Largest Contentful Paint): How long it takes for the main visible content to load. A slow LCP means the page looks unfinished, and many visitors leave before they ever see the products. Google’s threshold: under 2.5 seconds at the 75th percentile of real users.
INP (Interaction to Next Paint): How fast the page responds after someone taps or clicks. A slow INP makes add-to-cart buttons feel broken and form fields feel unresponsive. Google’s threshold: under 200 milliseconds at the 75th percentile. (For a full breakdown of INP’s conversion impact specifically, see the companion article in this series.)
CLS (Cumulative Layout Shift): Whether the page jumps around as it loads, buttons that move, text that shifts, checkout fields that reposition. High CLS causes mis-taps and abandoned sessions. Google’s threshold: under 0.1.
According to the HTTP Archive Web Almanac 2024, real-user data from 16.9 million websites collected in June 2024, only 43% of mobile sites currently pass all three metrics simultaneously. The majority of ecommerce sites have at least one failing metric. When a site fails INP, every button click during checkout is slower than Google’s threshold. That is session depth that does not happen, and AOV that goes unmeasured.
”Customers Decide What to Spend Before They Arrive”
This objection comes up often. The thinking goes: if a visitor came with a $100 budget, no page speed improvement will change that budget.
The objection misses how purchase decisions actually work in an online session. Most buyers do not arrive with a fixed basket. They arrive with an intent category: “I need a dress for an event” or “I’m replacing my outdoor furniture”, and the site they land on either enables or forecloses the discovery process that fills that basket.
A fast, responsive site enables browsing. A slow one truncates it.
The difference between a $95 order and a $104 order is often not the visitor’s budget, it is whether they saw the coordinating item, whether the recommendation appeared before they left, whether the checkout page loaded fast enough that they didn’t decide the experience was too difficult for a second purchase.
The Deloitte data measured this in 30 million real sessions. The effect is real.
A Second-Order Effect: What This Means for Your Ad Spend
If you are running Google Ads or Meta campaigns and paying per click, your cost per acquisition is set by how many of those clicks convert. A 8% lift in conversion rate means your cost per acquisition drops by a proportional amount without any change to your ad spend.
The compounding works in the other direction too: if your AOV rises 9%, your revenue per acquisition rises, meaning the same ad budget produces more profit per dollar spent.
Most advertising agencies optimize bids, audiences, and copy. Very few audit the landing page performance that determines whether those clicks convert. The performance gap is often the largest untouched lever in a digital marketing budget.
The Free Assessment
The worked example in this article: 500 orders, $95 AOV, dual conversion and AOV lift, is deliberately conservative. If your business has a faster-growing order volume, a higher AOV, or a site with more severe performance problems than the Deloitte baseline, the dollar figure scales accordingly.
The only way to run the math on your specific situation is to measure what your site actually does with real users, your CrUX field data, your INP and LCP scores, your actual conversion rate, and compare it against the benchmarks.
That is exactly what iServU’s free assessment produces. No account manager, no handoffs. Gabriel Lopez-Seco runs the assessment directly and provides a specific dollar estimate based on your numbers.
Book a free assessment at iServU →
Sources
- Deloitte / Google (2020). Milliseconds Make Millions. https://cdn.meshwithus.com.au/content/uploads/2025/07/26102751/Milliseconds_Make_Millions_report-deloitte-digital.pdf
- Yottaa (2025). Yottaa Launches Interactive Web Performance Index. https://www.yottaa.com/yottaa-launches-interactive-web-performance-index/
- Yottaa (2025). 2025 Yottaa Web Performance Index. https://www.yottaa.com/resources/2025-yottaa-web-performance-index/
- Contentsquare (2025). 2025 Digital Experience Benchmarks. https://contentsquare.com/press/2025-digital-experience-benchmarks/
- HTTP Archive / Web Almanac (2024). Web Almanac 2024: Performance. https://almanac.httparchive.org/en/2024/performance
