What Singtel’s Q3 Shows About AI-Driven Growth in SG

Singtel’s latest quarter looks like a headline made for investors: Q3 net profit jumped 43.5% to S$1.89 billion, and the stock closed at an all‑time high of S$5 on Feb 12, 2026. But if you run a startup or growth team in Singapore, the more useful story isn’t “profit up.” It’s how a mature incumbent keeps finding new growth engines while staying under competitive pressure.

Because most Singapore startups face the same constraint Singtel’s CEO called out: competition doesn’t politely step aside just because you’re executing well. If anything, markets get noisier, CAC rises, and customers expect instant, personalised experiences.

This post is part of our Singapore Startup Marketing series, focused on how local teams market regionally across APAC. Using Singtel’s Q3 update as a backdrop, I’ll translate what’s happening at the telco level into practical AI business tools and marketing systems that smaller teams can implement—without a telco-sized budget.

Singtel’s Q3 numbers: the headline vs the signal

Answer first: Singtel’s profit jump is real, but the signal for operators is in the underlying performance and where the company is investing—digital infrastructure that directly benefits AI-enabled products and marketing.

A quick breakdown from Singtel’s Feb 12 business update (quarter ended December):

• Net profit: S$1.89b, up 43.5% year-on-year
• Key driver: S$1.15b net exceptional gain, mainly from a partial sale of its Airtel stake
• Operating revenue: S$3.66b, up 0.9%
• Underlying net profit: S$744m, up 9.5%, supported by strong results from associates Airtel and AIS
• CEO message: the group “remains under competitive pressure” even with potential industry consolidation
• Strategic move: acquisition of ST Telemedia Global Data Centres (STT GDC) with KKR to scale digital infrastructure and capture opportunities from digitalisation and AI growth

Here’s the stance I take: one-off gains make headlines, but durable advantage comes from compounding execution—the unglamorous work of improving reliability, scaling infrastructure, and expanding into products where AI demand is rising.

That’s the part startups can copy.

The real play: AI growth runs on infrastructure (and trust)

Answer first: AI-powered growth isn’t only about better ads or better copy. It’s about having the infrastructure and reliability to deliver fast experiences—and the trust to keep customers when things break.

Singtel called out two seemingly separate topics in the same update:

1. Optus investing to strengthen resilience after the September 2025 Triple Zero outage, and
2. pushing into data centres to scale digital infrastructure.

For startup marketers, those map cleanly to two realities:

Reliability is a growth strategy

If your product slows down, support becomes a black hole, or onboarding fails on mobile, your marketing efficiency collapses. You can’t “growth hack” your way out of churn.

AI tools help here, but not in the flashy way people sell. The highest ROI AI uses I see in Singapore teams are:

• AI support triage that tags, routes, and summarises tickets so response times drop
• Conversation intelligence that turns sales calls into searchable “why we win/lose” reasons
• Incident comms automation (drafting status updates, in-product messages, post-mortems)

The point: when your ops improves, your marketing metrics improve. Less churn = CAC payback improves = you can afford to scale.

Data and compute aren’t “backend issues” anymore

Singtel’s data centre push (via STT GDC) is about being closer to the AI demand curve. For startups, you’re not buying data centres—but you are choosing:

• where your data lives,
• how quickly you can run analytics,
• how easily you can personalise experiences,
• whether you can deploy AI safely.

In practical terms, your marketing stack is now an infrastructure decision.

If you’re expanding regionally (Singapore → Malaysia/Indonesia/Thailand), latency, data governance, and logging suddenly matter. Not because it’s fun—because it changes conversion rate and retention.

A startup-friendly AI marketing system (steal this)

Answer first: The simplest way to apply AI for growth is to build a repeatable “signal → decision → action” loop across acquisition, conversion, and retention.

Singtel’s Singtel28 plan is essentially a multi-year execution framework. Startups need the same idea, just smaller: a weekly growth loop that uses AI to reduce manual work and improve decision quality.

Step 1: Centralise your signals (without boiling the ocean)

Start with 8–12 metrics you actually act on:

• CAC by channel and by segment
• conversion rate by landing page and offer
• activation rate (your first key action)
• churn and top 5 churn reasons
• NPS / CSAT themes
• pipeline velocity (if B2B)

Then use AI to summarise what changed and why:

• Auto-generated weekly performance narrative: “Paid search up 12% due to brand query growth; onboarding completion dropped on Android after release X.”
• Theme extraction from calls and tickets: “Top objections: compliance, integration time, pricing clarity.”

This matters because humans don’t scale in dashboards. Teams scale in decisions.

Step 2: Use AI where it directly reduces cycle time

If you only use AI for content generation, you’ll get volume—not advantage. Use it where it compresses time:

1. Landing page iteration

   • AI generates variant structures (not just headlines)
   • AI checks message-match between ad → page → form
   • AI proposes objections to address based on call/ticket themes

2. Sales enablement for regional expansion

   • AI builds “country-specific talk tracks” from your best calls
   • AI suggests localisation edits (currency, procurement norms, compliance phrasing)

3. Customer marketing

   • AI segments users by behaviour and predicts risk signals (e.g., feature usage drops)
   • AI drafts renewal nudges and onboarding sequences targeted to the “stuck point”

The stance: faster iteration beats “perfect strategy.”

Step 3: Personalise—but don’t creep people out

Personalisation works in Singapore and across APAC, but only when it’s earned.

A clean rule: personalise based on what customers do, not what you can infer.

Good examples:

• “You invited 3 teammates—here’s how to set permissions”
• “Your trial ends in 5 days—want us to import your data?”

Bad examples:

• “We noticed you’re hiring, so you must be scaling finance ops” (even if true)

AI makes it cheap to personalise. That doesn’t mean it’s wise to overdo it.

Competitive pressure is normal—AI is how small teams keep up

Answer first: When markets get crowded, AI isn’t a vanity project; it’s a way to run a tighter marketing operation with fewer people.

Singtel’s CEO explicitly noted competitive pressure despite potential consolidation. That’s a reminder for startups: your competitor’s budget isn’t your biggest threat—your own inefficiency is.

Here are three places where AI tools usually pay off fastest for Singapore startups marketing regionally:

1) Customer acquisition: reduce waste before you “scale spend”

Before increasing budgets, use AI to:

• classify leads by likely fit (based on firmographics + intent behaviours)
• identify which creatives are driving low-quality conversions
• detect channel cannibalisation (e.g., branded search capturing demand created elsewhere)

A quotable principle: Scale is what you do after you stop paying for mistakes.

2) Customer engagement: make messaging consistent across channels

Most teams have a “message drift” problem:

• the website says one thing,
• ads say another,
• sales decks promise a third,
• onboarding emails mention something else.

AI can audit and standardise this by:

• extracting your core value props and proof points
• checking every asset against them
• flagging contradictions and missing proof

Consistency feels boring. It also lifts conversion.

3) Retention: turn feedback into product and marketing decisions

Singtel’s underlying profit grew on strong associate performance and focused execution. For startups, the parallel is retention-led growth:

• summarise churn reasons weekly
• connect them to product events (releases, incidents, pricing changes)
• publish a “what we fixed” update to customers monthly

Customers don’t expect perfection. They do expect responsiveness.

A practical 30-day plan for AI adoption (marketing + CX)

Answer first: In 30 days, you can ship a useful AI layer on top of your existing tools—without migrating your entire stack.

Here’s a lean plan I’ve found works for Singapore teams:

Week 1: Pick one funnel and define success

Choose one:

• Lead gen → booked meetings
• Trial → activation
• Activation → paid
• Paid → renewal

Define 2 metrics and a baseline (last 4 weeks).

Week 2: Add AI summarisation to your feedback streams

Implement AI summaries for:

• sales calls (win/loss + objections)
• support tickets (themes + urgency)
• onboarding drop-offs (where users stop)

Output: one weekly “voice of customer” brief that anyone can read in 3 minutes.

Week 3: Ship 2 experiments

Examples:

• 2 landing page variants targeted to top two objections
• a 5-email onboarding sequence personalised by behaviour
• an in-app checklist that adapts to role (ops vs finance vs founder)

Week 4: Operationalise the loop

• schedule a 30-minute weekly growth review
• track experiments in a simple log (hypothesis → change → result)
• keep an AI-generated “what changed” summary so context isn’t lost

The only rule: if the AI output doesn’t change a decision, remove it.

Where Singtel’s story lands for startup marketers

Singtel’s Q3 result is boosted by a one-off gain, but the company’s direction is clear: invest in infrastructure, execute consistently, and position for AI-driven demand. That combination is what turns quarterly performance into multi-year momentum.

For Singapore startups marketing across APAC, the equivalent isn’t buying data centres or selling a stake. It’s building an operating cadence where AI improves customer engagement, speeds up experimentation, and reduces churn-causing friction.

If you’re planning your 2026 growth targets, here’s the question I’d ask your team this week: Which part of your funnel would improve the most if your cycle time dropped by 30%—and what’s the smallest AI change that gets you there?