Cell C’s Cloud Turnaround: Lessons for AI Commerce

Cell C didn’t claw its way back by building more towers. It did it by changing the operating model: partnership-first networks, cloud-native platforms, and a digital-first customer experience. And in late November 2025, that turnaround became public and measurable when Cell C listed on the JSE under ticker code CCD.

For anyone building AI-powered e-commerce in South Africa (or scaling any digital service), this matters. Telecoms sit underneath your checkout, your WhatsApp support, your delivery tracking, and your loyalty app. When a mobile operator proves that cloud-native delivery and experience design can restore growth, it’s a signal: the infrastructure and playbooks that power telcos are the same ones powering modern online retail.

This post is part of our series on How AI Is Powering E-commerce and Digital Services in South Africa. I’m going to use Cell C’s 2025 moves as a practical lens—less about corporate announcements, more about what you can copy in your own business.

Cloud-native isn’t an IT project — it’s a growth strategy

Cell C’s headline technology move in 2025 was becoming the first operator in Africa to begin live VoLTE testing on a fully virtualised, cloud-native IMS platform hosted on AWS. That sounds telecom-specific, but the underlying shift is familiar to every digital leader: standardise on cloud so you can ship faster, scale elastically, and instrument everything.

For e-commerce and digital services, cloud-native matters for one blunt reason: AI is hungry. Not just for data, but for reliable pipelines, event streams, and compute capacity that doesn’t fall over during promotions.

The e-commerce parallel: “cloud-native” equals “AI-ready”

If you’re still running your store and customer data across disconnected systems, AI will disappoint you. The winning pattern is:

• Centralise key events (browse, add-to-cart, purchase, refunds, support tickets)
• Process in near real time (so recommendations and messages aren’t stale)
• Expose clean APIs (so your app, site, WhatsApp bot, and call centre see the same truth)

Cell C’s choice to prove next-gen voice on a cloud-native core is basically the telco version of modern retail architecture: build a dependable platform, then let product teams iterate quickly.

What to do next (practical checklist)

If you want “AI for e-commerce” to be more than a demo, prioritise these foundations in Q1 2026 planning:

1. Single customer profile (even if it’s imperfect at first)
2. Event tracking you trust (consistent naming, governance, QA)
3. A/B testing muscle (so AI-driven changes are measurable)
4. Cost visibility (FinOps discipline so AI spend doesn’t drift)

A lot of South African businesses skip steps 2 and 3, then wonder why personalisation doesn’t move revenue.

The Cell C App relaunch shows where AI actually pays off: the last mile of experience

Cell C redesigned and relaunched its app to make it easier for customers to manage usage, buy value, and get support in a few clicks—positioning the app as a cornerstone of its customer experience strategy. That’s not a cosmetic change. It’s a shift toward self-serve, lower-cost service, and higher conversion.

E-commerce leaders already know the pattern: the app (or mobile web) is where customers make fast decisions. The best improvements are rarely flashy. They’re usually:

• fewer steps
• clearer pricing
• predictable fulfilment updates
• support that doesn’t make you repeat yourself

Where AI fits in the app layer (without annoying customers)

The simplest, highest-ROI AI features in commerce and digital services are:

• Product discovery assistance: “show me gifts under R500 that deliver before New Year’s”
• Order support automation: instant answers on tracking, returns, and exchanges
• Personalised value bundles: recommendations based on real behaviour (not demographics)
• Churn prevention prompts: detect drop-off patterns and offer help before the customer leaves

Cell C’s app focus is a reminder: AI works best when it removes friction, not when it adds novelty.

“Try Before You Buy” is a retention strategy disguised as a sales tactic

Cell C introduced a “Try Before You Buy” option across app, online channels, and in-store touchpoints. Again: telecom phrasing, universal lesson.

For online retailers, the equivalent is reducing purchase anxiety:

• guided size tools and fit confidence
• transparent returns with pickup options
• “reserve online, collect in store” for high-consideration items
• subscriptions with easy pause/cancel

AI helps here by predicting what creates regret (wrong size, wrong spec, late delivery) and nudging customers toward safer choices.

Partnership-first models are winning — and they’re not just for telecoms

Cell C describes its strategy as capex-light and partnership-first, strengthening network partnerships instead of owning everything. For digital commerce in South Africa, this is the most transferable idea in the whole story.

Most companies get this wrong: they treat partnerships as a “nice-to-have” after product-market fit. In reality, partnerships are often the shortest path to scale—especially in a market where logistics, payments, identity, and connectivity vary widely by region.

What “partnership-first” looks like in e-commerce

You don’t need to own the full stack to provide a premium experience. You need to orchestrate it:

• Payments: multiple rails (cards, EFT, wallets, pay-by-bank) with smart routing
• Delivery: dynamic carrier selection based on cost, region, and delivery promise
• Customer support: blended model (AI triage + human escalation)
• Marketing: retailer media, affiliates, creators, and performance campaigns tied to clean attribution

The point isn’t outsourcing. The point is choosing where you must differentiate (brand, assortment, pricing strategy, service design) versus where you can partner (infrastructure and reach).

Distribution partnerships = where customers already shop

Cell C expanded distribution partnerships to increase access to SIMs, devices, and offers in high-traffic retail locations. E-commerce has a similar reality: digital-only is rarely enough.

If you’re trying to grow leads and customers in 2026, consider “digital services in physical spaces”:

• pickup/return points at existing retailers
• pop-up activations for seasonal peaks (December, back-to-school, Easter)
• device trade-in or repair partners if you sell electronics

AI supports this by forecasting demand by area and optimising stock placement—so the physical footprint doesn’t become an expensive guessing game.

Why Cell C’s AWS + cloud narrative matters for AI infrastructure in South Africa

Cell C’s cloud-native move on AWS and its broader cloud cooperation signals something bigger than one telco’s roadmap: South Africa’s AI economy is becoming cloud-first by default.

If you’re building AI-powered commerce or a digital service, you’re going to face three hard constraints:

1. Latency and uptime: customers don’t care why your service is slow
2. Data protection: POPIA-aligned handling, access controls, audit trails
3. Skills: you need teams who can run data pipelines and AI systems reliably

Cloud platforms help with all three—if you implement them with discipline.

The non-negotiables: governance before “AI features”

Here’s what I’ve found separates teams that get ROI from AI from teams that ship prototypes forever:

• Define data ownership (who is responsible for customer, product, order data)
• Set model guardrails (what AI can and can’t say, especially in support)
• Measure business impact (conversion rate, AOV, repeat purchase, cost per contact)
• Plan human fallbacks (no one trusts a bot that can’t escalate)

Cell C’s emphasis on service quality and customer experience is the right priority order. AI is only helpful when the fundamentals are stable.

A practical 2026 playbook: copy the pattern, not the press release

Cell C’s story is easy to misread as “big-company stuff.” It isn’t. It’s a blueprint for any South African business selling digitally.

1) Make the experience measurable

If you can’t measure it, you can’t improve it. Set a simple scorecard:

• checkout completion rate (mobile vs desktop)
• time-to-resolution for support
• repeat purchase rate by cohort
• delivery promise accuracy (promised vs delivered)

Then use AI to target the biggest leaks first.

2) Build a self-serve layer customers actually like

Self-serve fails when it’s designed for the company, not the customer. Prioritise:

• clear order status and proactive updates
• instant refunds/credits where possible
• FAQ and chat that understand local language patterns and product terms

3) Treat partnerships as product strategy

Map your “experience chain” end-to-end. Identify where partners improve reach or reliability, then integrate properly (shared SLAs, shared metrics, escalation paths).

4) Invest in brand trust, not just performance ads

Cell C continued its brand refresh and major marketing efforts while rebuilding operations. That’s the stance I agree with: performance marketing is fragile when trust is low.

For e-commerce, trust is earned through:

• consistent delivery
• transparent pricing
• fair returns
• support that resolves issues quickly

AI can help you communicate better, but it can’t compensate for broken operations.

What this means for AI-powered e-commerce and digital services in South Africa

Cell C’s 2025 turning point shows a pattern that’s repeating across the country: cloud infrastructure + digital-first experience + smart partnerships is how you build resilience under economic pressure.

If you’re leading an online store, marketplace, fintech, or subscription service going into 2026, copy the approach: get your platform stable, make the app and self-serve experience your growth engine, and partner aggressively where it improves reach.

If you want leads from your digital channels, start by asking a tougher question than “Which AI tool should we use?” Ask: Where does customer friction cost us money every week—and what data do we need to remove it?