The verdict in three sentences
In West Africa, most buyers are on entry-level phones and intermittent 3G: your checkout must be light before it is beautiful. Each second of load time costs about 7% of conversion, and a tunnel under 150 KB makes the difference between a sale and an abandonment. Performance is not a technical detail, it is revenue directly.
The real cost of slowness
On 3G, milliseconds turn into abandonment. Here is the estimated impact of checkout load time in 2026.
| Load time (3G) | Estimated abandonment | Relative conversion |
|---|---|---|
| Under 1.5s | 24% | Baseline |
| 2.5s | 32% | -7% |
| 4s | 45% | -18% |
| 6s | 58% | -30% |
| 8s and above | 70%+ | -45% |
The 2026 target: LCP under 2.5s on 3G and interaction possible in under 3s. Beyond that, every lost second erases a direct share of orders.
Techniques for a light checkout
A fast 3G checkout relies on restraint. Each optimization below has measurable impact.
| Technique | Estimated gain | Priority |
|---|---|---|
| Checkout under 150 KB total | -10% abandonment | High |
| Minify / drop unused JS | LCP -1 to 2s | High |
| WebP/AVIF images + lazy loading | -40% image weight | High |
| Server-side render (SSR) payment page | Lower TTFB | High |
| Offline-first (resume after drop) | +5% carts saved | Medium |
| USSD fallback if web fails | +4% conversion | Medium |
| System fonts (no heavy webfont) | -100 to 200 KB | Low |
| Aggressive static asset caching | Faster reloads | Medium |
An offline-first checkout lets a customer resume an order after a network drop, common on the move: the cart is not lost when the connection returns.
Mini case study
Moussa sells groceries online in Bamako. His checkout weighed 480 KB and loaded in 5.2s on 3G, with 52% abandonment at the payment step. On 700 weekly paying sessions and an average cart of 14,000 FCFA (about 21 EUR), he lost a lot at the last moment.
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
After shrinking to 140 KB, moving to SSR and AVIF images, his checkout loads in 2.3s and abandonment falls to 38%. He now converts 14% more sessions, roughly +840,000 FCFA per week, without changing a line of his catalog or his advertising.
FAQ
Why aim for 150 KB rather than just good design?
Because on 3G, weight is paid in seconds. A 150 KB checkout loads in about 2s where 500 KB takes 6 or more. Design is still possible, but optimized: system fonts, compressed images, minimal JS.
Is SSR really useful for a checkout?
Yes. Server rendering shows the payment page immediately, without waiting to download and run large JavaScript bundles. On entry-level phones, this cuts time-to-interaction by several seconds.
What is a USSD fallback?
If web payment fails (unstable network), the customer can complete via a USSD code on their phone, without a browser. It is a valuable safety net in low-data-coverage areas, recovering about 4% of conversion.
How do I measure my LCP in real conditions?
Use Core Web Vitals on real traffic, filtering 3G and mobile sessions. Lab tests on fiber hide what your customers actually experience. Aim for LCP under 2.5s at the 75th percentile.
Let's talk about your project. We audit your checkout performance under real 3G conditions and make it light. WhatsApp +221 77 596 93 33.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.
