Page speed and conversion have a direct, measurable relationship: for every additional second it takes your site to load on mobile, published research shows conversions can fall by up to 20%. South Africa's conversion rate optimisation landscape makes this relationship more consequential than global averages suggest — the country's mobile-first internet culture, load-shedding-affected connectivity, and data-conscious shoppers create conditions where visitors are less patient, not more.
If your site takes three seconds to load on a mid-range Android device using mobile data, you are not competing on a level field with the sub-2-second load times that well-optimised ecommerce sites deliver. You are losing visitors before they have seen a single product, a single price, or a single trust signal — and those visitors rarely return.
Understanding exactly how much each second costs, and why the SA context amplifies the damage, is the starting point for any meaningful revenue improvement. A broader look at your ecommerce conversion rate in South Africa will show page speed is rarely the only gap, but it is almost always the fastest one to close.
Quick Answer
Page speed and conversion are directly proportional: ecommerce sites loading in one second convert at around 3.05%, while the same sites at three seconds convert at 1.12% — a 63% drop in revenue-generating visits for the same traffic spend. For South African businesses whose customers browse predominantly on mobile, on variable-speed data connections, every second above a two-second load threshold is measurably eating into conversions. Fixing page speed is typically one of the highest-ROI CRO interventions available, because it lifts every other optimisation you have already made.
Wondering what page speed is costing your store specifically? Get a CRO assessment — we'll review your real-user speed data alongside your analytics and tell you where the leak is.
How Page Speed and Conversion Rates Are Actually Linked
The clearest published dataset on conversion rates by load time comes from Portent's site speed study, which analysed B2B lead generation and ecommerce sites across thousands of sessions. The ecommerce numbers are stark:
| Load time | Ecommerce conversion rate | vs. 1-second benchmark |
|---|---|---|
| 1 second | 3.05% | Baseline |
| 2 seconds | 1.68% | −45% |
| 3 seconds | 1.12% | −63% |
| 4 seconds | 0.67% | −78% |
Source: Portent — Site Speed and Revenue. The research also found that a 1-second site achieves 2.5x the ecommerce conversion rate of a 5-second site, and for B2B lead generation, the advantage at one second is 3x over five seconds and 5x over ten seconds.
These are not theoretical losses. They represent paying traffic — from Google Ads, Meta Ads, organic search, or email campaigns — arriving at your site and leaving without converting because the page did not load fast enough to hold their attention. The money you spent acquiring those visitors is already gone. The conversion opportunity is the only part you still control.
Key Takeaway
The conversion rate at a 3-second load is 63% lower than at 1 second — based on Portent's published ecommerce data (table above). If your load time is above 3 seconds — as it is for many SA stores on shared hosting or unoptimised themes — the speed gap is already your single largest CRO priority, ahead of copy, design, and checkout optimisation.
Google's own research on mobile behaviour reinforces this at the traffic level: as page load time increases from one second to three seconds, the probability of a visitor bouncing increases by 32%. From one second to ten seconds, it rises by 123%. Separately, Think with Google data shows that 53% of people abandon a mobile page that takes longer than three seconds to load — meaning a slow site is rejecting more than half its mobile audience before they ever engage. 79% of shoppers dissatisfied with site performance say they are less likely to buy from the same site again.
Why Page Speed and Conversion Matter More in South Africa
Global speed benchmarks are measured against a global audience mix. South Africa's audience is not global. According to research published by CIPESA, 79% of South Africans access the internet primarily via mobile devices — not desktop fibre. That matters because mobile connections are variable, affected by network congestion, tower load, and terrain in ways that desktop fibre is not.
South Africa also spent 300 days of 2023 under some form of load shedding, at an estimated economic cost of R1 billion per day. During rolling blackouts, home fibre routers go offline, businesses lose their office networks, and mobile towers run on battery backup for several hours — capacity varies by tower and backup configuration — before signals degrade.
The pattern meant that a meaningful portion of South African internet sessions in 2023 happened on degraded, battery-strained mobile networks. Load shedding has eased since — approximately 83 days in 2024 versus 300 in 2023 — but the broader infrastructure reality remains: the SA audience is mobile-heavy and operating on more variable connections than Western European or North American benchmarks assume.
The Data Budget Factor
South African consumers on prepaid mobile connections are actively conscious of data usage. A slow-loading page does not just waste their time — it consumes their data before delivering value. When a product image takes four seconds to appear, the shopper has already used data for a page that failed to show them what they came for. This is a friction type that simply does not exist in the same way for a desktop-fibre user in Sandton — and it falls disproportionately on the large portion of SA shoppers browsing on prepaid data.
The commercial consequence: SA ecommerce conversion rates already run below the global average — a gap that compounds every percentage point of speed-related loss. Meanwhile, global cart abandonment research by Baymard Institute — averaging 50 studies — puts average abandonment at 70.22%. For SA stores where mobile performance is poor, the abandonment figure is likely higher.
A visitor who cannot load the checkout page quickly, especially when their data is limited, is a visitor who abandons rather than waits. The intersection of SA's below-average conversion baseline with the mobile speed problem is where the real revenue opportunity sits — and it is largely addressable through technical fixes, not creative work. See also: cart abandonment in South Africa for the full data picture on why SA shoppers leave.
Key Takeaway
South Africa's mobile-first internet access, variable network quality, and data-cost sensitivity mean that speed problems hit harder here than global benchmarks imply. The audience has less patience and more concrete reasons to abandon — which makes page speed optimisation a higher-leverage intervention for SA businesses than the global studies alone would suggest.
Core Web Vitals: The Speed Thresholds That Predict Conversion
Google's Core Web Vitals define the measurable thresholds where user experience shifts from acceptable to problematic: Largest Contentful Paint (LCP) under 2.5 seconds, Interaction to Next Paint (INP) under 200 milliseconds, and Cumulative Layout Shift (CLS) below 0.1. These are not arbitrary Google preferences — they are anchored in user behaviour research. Sites that pass all three see 24% fewer page abandonment events, which aligns directly with the conversion data above.
Only 39% of ecommerce sites globally pass all three Core Web Vitals simultaneously, according to HTTPArchive data from mid-2025. Most SA mobile sites are measured by Google against real South African user hardware and connection speeds — not the ideal lab conditions many speed tests simulate. A site that appears fast in a desktop PageSpeed Insights test may still be delivering a poor experience to a user on an entry-level Android handset in Soweto or Cape Flats.
The technical detail of what to measure, how to read field data versus lab data, and which fixes to prioritise is covered in depth in the Core Web Vitals guide for South Africa. For the purposes of this post, the key point is that CWV thresholds map directly to conversion thresholds: a site failing LCP is almost certainly losing conversions to speed, and the fix is a technical one, not a design or copy one.
What Each Lost Second Costs on a Real SA Site
The Portent data gives you a framework to calculate revenue at risk without needing a case study. Here is how to apply it:
At a 3-second load time (1.12% conversion rate from the Portent data above): approximately 224 orders per month.
At a 2-second load time (1.68%): approximately 336 orders per month.
That is 112 additional orders from a single second of improvement — before any other CRO work is done.
This is a directional calculation using published benchmarks, not a guarantee — your actual uplift depends on your product category, checkout friction, and starting load time. But the direction is consistent across every published study: faster is better, and the gains at the low end (moving from 3s to 2s) are proportionally larger than gains already in the fast zone (moving from 1.5s to 1s).
The practical implication: if you are currently at 4+ seconds, even a partial fix that gets you to 2.5 seconds represents a meaningful CR improvement before any other CRO work is done. Landing page optimisation and checkout page improvements compound on top of a fast baseline — they struggle to overcome a slow one.
Not sure what your current load time is costing you? We can pull your real-user field data and give you a revenue-at-risk estimate based on your actual traffic numbers — request a CRO assessment.
How to Find Your Speed-Conversion Gap
The measurement approach matters. Lab scores from tools like PageSpeed Insights show what a controlled test finds — they are useful for diagnosing issues but not for measuring the experience your actual visitors receive. For that, you need field data: Google's CrUX (Chrome User Experience Report) shows the real distribution of load times across your actual users on their actual devices and connections.
Inside Google Search Console, the Core Web Vitals report gives you field-data segmented by page group and device type. The first thing most SA site owners discover is that their mobile field data is significantly worse than their desktop lab score. That gap — between what the tool shows and what the visitor experiences — is exactly where the conversion loss happens.
What to check first
- Google Search Console → Core Web Vitals report: Are there "Poor" LCP URLs in the mobile view?
- GA4 → Explore → Page load time by device: Does your mobile average load time materially exceed desktop?
- GA4 → Engagement rate segmented by device: If mobile engagement is significantly lower than desktop, speed is often a contributor.
The highest-leverage fixes for most SA sites, in order of impact:
- Image compression and next-gen formats: Oversized images are the most common driver of slow LCP. Serve WebP, set explicit dimensions on every image, and lazy-load below-the-fold images to stop them blocking the main render path.
- Render-blocking JavaScript: Third-party scripts — chat widgets, tracking pixels, review apps — delay interactivity. Audit what loads on page init and defer anything not needed above the fold.
- Server response time (TTFB): A slow server adds latency before a single byte reaches the visitor. South African-hosted sites and those on shared international hosting often show high TTFB on mobile; a CDN or a local-hosted upgrade addresses this directly.
Once you can see where the gap is, you can test fixes and measure the outcome. A/B testing in South Africa is the rigorous way to confirm that a speed improvement produced a conversion lift rather than attributing it to seasonal variation or traffic mix. For most SA sites, the speed-conversion gap is visible without a test — a move from 4s to 2s is large enough that the conversion change will be statistically clear within a few weeks of normal traffic.
Why South African Businesses Choose Growth Pulse Media for CRO
Growth Pulse Media's approach to conversion rate optimisation for South African businesses starts with field data, not assumptions. Dirk built and scaled an SA ecommerce operation before founding the agency — he has seen first-hand how a slow product page kills a paid traffic campaign, how a checkout that stalls on Ozow or PayFast redirects loses orders, and how a landing page that loads in 4 seconds on a mid-range Huawei destroys a Google Ads ROAS.
That operator background shapes how we approach speed and conversion work: we audit your actual CrUX data across your traffic mix, identify the specific pages where load time is suppressing conversion, and prioritise technical fixes by revenue impact rather than by score. We work with Shopify, WooCommerce, and custom-built SA stores, and we handle the full stack — image optimisation, server configuration, JavaScript trimming, and render-blocking resource sequencing — rather than handing you a report and leaving the implementation to chance.
We deliberately limit our active client load so that every engagement gets senior attention, not a junior team running a template. If your site is already fast (sub-2-second LCP in field data), we will tell you that speed is not the priority and redirect the budget toward the conversion gaps that remain. Honest scope is part of the service.
Who This Is NOT For
If your site receives fewer than a few thousand monthly visitors, the speed-conversion correlation is real but the absolute revenue impact will be small until you have built a more substantial traffic base. Fix speed as part of your technical foundation, but do not deprioritise traffic acquisition to do it.
If your website's primary job is to generate a phone call, a showroom visit, or an in-person consultation — and the contact form or phone number is visible above the fold — load time below the fold matters less. Speed still affects bounce rate, but the conversion calculation is different.
A fast site that sends the wrong message to the wrong audience still converts poorly. If your traffic is unqualified, your value proposition is unclear, or your pricing is misaligned with your market, speed optimisation will not fix it. The conversion problem is upstream of the technical one.
If your field LCP is already under 2.5 seconds across your key pages, you are already in the conversion-friendly zone on speed. The marginal gain from moving from 2.2s to 1.8s exists but is small compared to optimising checkout, trust signals, or offer structure. Speed is not your bottleneck — focus elsewhere.
Not sure which category you're in? A CRO audit will tell you exactly where your highest-leverage gap is. See what a CRO audit covers before deciding.
Frequently Asked Questions
How much does page speed affect conversion rate?
The relationship is significant and well-documented. Portent's site speed study found ecommerce conversion rates of 3.05% at a 1-second load time, dropping to 1.68% at 2 seconds and 1.12% at 3 seconds. That is a 63% drop in conversion rate between a 1-second and a 3-second load time. For B2B lead generation, a 1-second site achieves three times the conversion rate of a 5-second site. Think with Google's mobile research found that a one-second delay can reduce mobile conversions by up to 20%. Across all published studies, the direction is consistent: every additional second of load time costs conversions.
What is a good page load time for a South African website?
Google's Core Web Vitals threshold for Largest Contentful Paint — the metric most directly tied to perceived load speed — is under 2.5 seconds, measured in field data from real users. For South African mobile sites, achieving this is harder than the global benchmark implies, because Google measures performance on the devices and connections your actual visitors use. A useful working target for SA ecommerce and service sites is an LCP of under 2.5 seconds in the CrUX mobile field data for your most-visited pages. Lab scores from desktop tests are useful for diagnosis but do not reflect real SA visitor experience.
How does load shedding affect website conversions in South Africa?
Load shedding affects conversions in two ways. During outages, some visitors lose fibre or office connectivity and switch to mobile data — or drop off entirely. Cell towers run on battery backup for several hours — capacity varying by tower and backup configuration — before signal quality degrades, meaning sessions during long outages happen on progressively worse connections. Separately, the infrastructure diversion caused by years of load shedding (2023 saw 300 days) has reinforced mobile-data dependence as the primary internet access mode for most South Africans — which raises the stakes for mobile page speed even in periods of stable power supply.
Does page speed affect Google rankings as well as conversions?
Yes, via two mechanisms. Core Web Vitals are a confirmed Google ranking signal — sites failing LCP, INP, or CLS thresholds in field data can be disadvantaged in mobile search rankings relative to faster competitors. The second mechanism is indirect: a slow site raises bounce rate, which reduces the engagement signals Google measures. For South African sites competing in mobile-first search results, poor Core Web Vitals are both a ranking problem and a conversion problem. Fixing them addresses both simultaneously. The Core Web Vitals guide for South Africa covers the technical details of each metric and what Google's ranking system actually measures.
Can I improve conversions without improving page speed?
Yes, but with a ceiling. Copy improvements, trust signals, checkout simplification, and offer changes can all lift conversions — and they should be part of any CRO programme. However, they operate on the visitors who stay long enough to be influenced. If 53% of your mobile traffic is bouncing because the page takes longer than three seconds to load, copy and design changes only affect those who stayed to see them. Fixing speed expands the audience that your other CRO work can reach. For most SA sites with load times above 3 seconds, speed is the multiplier that makes every other conversion improvement more effective.
Ready to Find Out What Speed Is Costing Your Site?
We pull your real-user CrUX field data, map it against your GA4 conversion funnel, and identify exactly which pages are losing conversions to load time — and which technical fixes will recover the most revenue. We work across Shopify, WooCommerce, and custom SA-built stores, with full integration for PayFast, Peach Payments, and Ozow checkout flows.
No obligation — we'll get back to you within 24 hours.
