AI-Ready Telecom: What Cell C’s 2025 Shift Means

Cell C didn’t just “turn a corner” in 2025 — it proved a point that South African digital businesses can’t ignore: cloud-first operations plus a serious digital customer experience can restore growth even in a brutal market. That’s why its return to profitability and its JSE listing in late November matter beyond telecoms.

If you’re building an e-commerce brand, a marketplace, or a digital service in South Africa, you’re playing the same game. Customers expect fast, reliable service. They want self-service options that work. They punish friction instantly. And increasingly, AI is the layer that makes those expectations affordable at scale.

Cell C’s 2025 story gives us a useful case study for this series, How AI Is Powering E-commerce and Digital Services in South Africa: not because it’s “about AI” on the surface, but because the building blocks it doubled down on — cloud-native infrastructure, app-led service, and partner ecosystems — are the same foundations AI depends on.

Cloud-native infrastructure is the precondition for AI at scale

AI doesn’t fix a brittle stack. It amplifies a modern one. Cell C’s move to live VoLTE testing on a fully virtualised, cloud-native IMS platform hosted on AWS is a telecom-specific example of a broader pattern: when core systems become software-defined and cloud-hosted, it becomes easier to test, iterate, and scale.

For e-commerce and digital services, the parallel is direct:

• If your product catalogue lives in three disconnected tools, AI personalisation will be shallow.
• If your customer support data is scattered across inboxes and spreadsheets, AI agents will hallucinate or fail.
• If your website performance is inconsistent, AI-driven acquisition won’t convert.

What “cloud-native” really buys you (even if you’re not a telco)

The value isn’t the cloud logo. It’s the operating model:

1. Faster experimentation: You can run A/B tests, roll out features, and measure impact without waiting for heavyweight infrastructure changes.
2. Elastic capacity: December retail peaks in South Africa are real. Cloud-ready platforms handle demand spikes without turning the checkout into a loading screen.
3. Cleaner data flows: AI tools are only as good as the data pipelines feeding them.

Here’s the stance I’ll defend: most “AI projects” fail because the plumbing is ignored. Cell C’s emphasis on cloud-native core network capability is exactly the sort of unglamorous, high-leverage investment that makes AI useful later.

A digital-first customer experience is where AI becomes visible

Cell C’s redesigned and relaunched app is the part customers actually feel. It’s also where AI-driven engagement typically lands first, because apps generate intent signals: top-ups, usage checks, device browsing, help searches, cancellations, upgrades.

Even without claiming specific AI features, an app-led strategy naturally sets you up for AI outcomes:

• Personalised offers based on behaviour (not guesswork)
• Next-best action prompts (e.g., “switch to this bundle” at the right time)
• Smarter support journeys that reduce call centre volume

“Try Before You Buy” is a conversion tactic e-commerce should copy

Cell C introduced a “Try Before You Buy” option across app, online, and in-store touchpoints. That’s not just a product tweak — it’s a trust and risk-reversal mechanism, and it maps beautifully to online retail.

If you sell online in South Africa, risk is the silent conversion killer:

• “Will it arrive?”
• “Will it work on my device?”
• “What if I need to return it?”

Borrow the idea in practical ways:

• Device or accessory trial windows (especially for higher-ticket items)
• Subscription sampling (7–14 days) with in-app reminders
• Guided setup + proactive support during the first week

AI fits here because it can automate the follow-through: onboarding messages, anomaly detection (e.g., a customer struggling), and targeted support prompts.

One-liner worth remembering: Great digital experiences aren’t “pretty”; they’re predictable.

Partnerships are the fastest route to modern capability

Cell C’s capex-light, partnership-first approach is a reminder that you don’t need to own everything to compete. It leaned on network partnerships, cloud partnerships (including a memorandum of understanding with Huawei Cloud around cloud services and industry solutions), and expanded distribution relationships.

For e-commerce and digital services, the equivalent is building a strong partner mesh:

• Payments (PSPs, alternative payment methods)
• Logistics (couriers, pickup points, returns partners)
• Cloud + data tooling
• Customer support platforms
• Marketing tech (CDPs, email/SMS, attribution)

Why partner ecosystems matter for AI adoption

AI is increasingly delivered as capabilities inside platforms: customer service suites, analytics tools, marketing automation, fraud detection, search, merchandising.

A partner ecosystem helps you:

• Adopt AI faster (use proven modules instead of building from scratch)
• Control cost (pay for usage, avoid heavy internal build-outs)
• Reduce implementation risk (battle-tested integrations)

Cell C’s approach signals a broader South African reality: speed matters more than pride of ownership. If your competitor integrates an AI-enabled support workflow in six weeks and you’re still scoping a custom build, you’ll feel it in churn and CAC.

Distribution still wins — but it has to be data-driven

Cell C highlighted expanded distribution partnerships that put SIMs, devices, and offers in high-traffic retail locations nationwide. That’s an old-school truth: availability beats awareness.

But the modern version of distribution isn’t just shelf space; it’s measurable, optimisable distribution.

What e-commerce teams can take from this

If you operate across online and physical touchpoints (or even just online with partners), use AI and analytics to answer these questions weekly:

• Which channels drive first purchase vs repeat purchase?
• Where do customers drop off — product page, cart, payment, delivery options?
• Which promotions create margin-destroying behaviour (discount-only buyers)?
• Which locations/partners produce high return rates or high support loads?

A practical starting point I’ve found works: build a single “channel truth” dashboard combining acquisition source, conversion rate, delivery SLA, return rate, and support tickets per order. Once you can see it, you can automate decisions.

The JSE listing matters because transparency forces operational discipline

Cell C’s JSE Main Board listing (ticker code CCD) is more than a headline. Public markets punish fuzzy execution. They demand repeatability, reporting cadence, and credible strategy.

That matters for AI because AI initiatives have a reputation problem: too many businesses treat them as innovation theatre.

If you want AI to drive growth in e-commerce or digital services, run it like Cell C describes running its turnaround: disciplined, measurable, and customer-outcome led.

A simple AI adoption scorecard (steal this)

Use this four-part scorecard to decide if an AI project is real:

1. Customer outcome: What metric moves? (conversion, AOV, churn, NPS, resolution time)
2. Data readiness: What data feeds it, and who owns quality?
3. Workflow integration: Where does it sit inside day-to-day operations?
4. Unit economics: What does it cost per incremental order or per ticket resolved?

If you can’t answer all four, it’s not ready.

People also ask: What does a telecom’s tech shift have to do with e-commerce?

Direct answer: Telecom upgrades shape the reliability, speed, and affordability of the connectivity your customers use to shop, pay, and get support.

More importantly, telecoms are facing the same customer-experience pressures as online retailers: instant service expectations, low tolerance for friction, and constant price comparison. When a telco rebuilds around cloud-native systems and app-led service, it’s a preview of what “table stakes” looks like for every digital service in South Africa.

Practical next steps for South African e-commerce and digital service leaders

Cell C’s 2025 playbook translates into a clear action list for teams chasing growth (and leads) in 2026.

1) Modernise one core system before you “add AI”

Pick the system that causes the most customer pain (checkout, fulfilment visibility, returns, support). Upgrade that. Then add AI.

2) Turn your app (or logged-in web) into the control centre

If customers can’t self-serve quickly, support costs balloon and churn rises. Make these flows effortless:

• Order tracking and delivery changes
• Returns and exchanges
• Subscription management
• Saved payment and address options

3) Treat risk reversal as a growth lever

Adapt “Try Before You Buy” into your category:

• Sample packs
• Flexible returns
• Pay-on-delivery options where feasible
• Warranty clarity and instant support

4) Build a partner roadmap

Choose partners that reduce time-to-value: cloud, payments, customer service, analytics, fraud. Integrate deeply, not superficially.

Where this goes in 2026: AI will reward the prepared

Cell C’s CEO framed 2026 as a year for disciplined execution and expanding digital inclusion. I agree with the direction: connectivity is becoming the substrate for the whole digital economy — shopping, banking, education, work, entertainment.

For this series, the bigger lesson is straightforward: AI in South African e-commerce won’t be won by the flashiest model. It’ll be won by the companies that fix their foundations, instrument their customer journeys, and operationalise what they learn.

If your team wants to use AI to improve customer engagement, reduce support load, and increase conversion, start where Cell C started: build the infrastructure and experiences that can carry the weight. What part of your customer journey would feel dramatically better if it simply became more predictable next quarter?