AI Strategy Under Fire: What IQVIA Teaches Businesses

A six-day market selloff recently wiped out about US$830 billion in value from software and services stocks, sparked by fresh anxiety that newer AI tools could undercut established enterprise players. One of the companies pulled into that debate was IQVIA, a healthcare data and clinical research provider. After forecasting 2026 adjusted earnings of US$12.55–US$12.85 per share (below analysts’ US$12.95 estimate), IQVIA spent part of its earnings call defending a point many leadership teams in Singapore are also grappling with:

When AI is moving this fast, how do you invest with confidence—without getting displaced by the same technology you’re adopting?

This post is part of the AI Business Tools Singapore series, where we look at how real organisations adopt AI for operations, decision-making, and customer engagement. IQVIA’s situation is a useful case study because it highlights the exact tension most businesses face: AI can improve your service, but it can also compress your margins if you don’t protect what makes you unique.

“It’s really hard to disprove a generic assertion… like AI is going to displace your business. It’s exactly the opposite.” — IQVIA CEO Ari Bousbib (via Reuters)

What happened with IQVIA—and why the market cared

Answer first: IQVIA’s AI narrative collided with investor fear that general-purpose AI will replace parts of consulting, research, and other professional services—especially after new entrants raised expectations of what AI can do at lower cost.

The Reuters report (published by CNA) describes how analysts questioned whether recent AI advances could displace IQVIA’s services, even as IQVIA maintained its AI strategy is a net positive. The company’s shares fell over 8% during late morning trading after the earnings call.

The numbers that framed the debate

IQVIA’s guidance and results show a business that’s still performing operationally, but getting judged on future defensibility:

• 2026 adjusted EPS: US$12.55–US$12.85 vs US$12.95 expected (LSEG)
• Higher interest expenses: nearly US$80 million impacting profit guidance
• 2026 revenue outlook: US$17.15–US$17.35 billion vs US$17.07 billion expected
• Q4 EPS: US$3.42 (beat expectations)
• Q4 revenue: US$4.36 billion (beat expectations)

So the market wasn’t reacting because the business “broke.” It reacted because investors are asking a sharper question in 2026:

If AI gets cheaper and more capable every quarter, what stops clients from substituting your work with a tool?

The real lesson: AI doesn’t replace your moat—unless your moat is weak

Answer first: IQVIA’s defence is a classic “moat argument”: proprietary data + domain workflows + trust are harder to copy than generic AI capabilities.

IQVIA’s CEO leaned heavily on one point: the company has “the largest proprietary healthcare information assets in the world” (as quoted in the article), and those assets are not replicable by general-purpose models.

This is the stance I agree with most strongly for Singapore businesses evaluating AI tools:

AI is not your moat. Your data, processes, distribution, and compliance discipline are.

If you’re a services-heavy company, AI can absolutely commoditise parts of your delivery. But that doesn’t mean the whole business gets replaced. It means your “value unit” changes.

A simple way to think about defensibility in the AI era

Here’s a practical framework you can apply, whether you’re in healthcare, finance, logistics, education, or B2B services:

1. Commodity work (highly repeatable tasks) will trend toward automation.
2. Context work (requires your internal data, history, constraints) remains defensible.
3. Regulated work (auditable decisions, safety, privacy) rewards maturity and governance.
4. Relationship work (procurement trust, change management, stakeholder alignment) doesn’t disappear.

IQVIA is effectively saying: We sit in categories 2 and 3, and we can use AI to do more of category 1 faster.

Analyst scepticism is healthy—use it as your AI implementation checklist

Answer first: The analyst pushback mirrors what Singapore decision-makers should ask before committing budget: what’s the measurable impact, what’s the risk, and what’s the substitution threat?

During the call, analysts asked whether AI could threaten IQVIA’s core business. That’s not “anti-AI.” It’s good governance.

The 7 questions worth stealing from the IQVIA moment

When you assess AI business tools in Singapore—especially tools that touch customers, financial decisions, or regulated data—run these questions:

1. Which part of our service becomes cheaper because of AI?

   • If your pricing is time-based, you need a new value metric (outcomes, speed, risk reduction).

2. Which part becomes easier for competitors to copy?

   • If your differentiation is “we write reports,” that’s fragile.

3. What do we have that models don’t?

   • Proprietary customer interactions, workflow data, SOPs, and labelled internal datasets.

4. Where can we prove ROI in 30–60 days?

   • Don’t start with a 12-month moonshot. Start with a measurable operational bottleneck.

1. What’s our compliance posture?

   • In Singapore, many firms must think about PDPA, audit trails, vendor risk, and data residency.

2. What’s our human-in-the-loop design?

   • If AI drafts, who approves? If AI recommends, who signs off? If AI flags risk, who owns it?

3. What’s the downside scenario?

   • Wrong outputs, data leakage, model drift, staff misuse, reputational risk.

These questions don’t slow adoption. They prevent expensive “AI theatre.”

AI in healthcare is the clearest example of why data quality beats model hype

Answer first: In healthcare use cases, the limiting factor is rarely the model—it’s data access, provenance, governance, and workflow integration.

IQVIA operates at the intersection of clinical research, real-world evidence, and healthcare operations—areas where bad data and weak governance aren’t just costly, they’re dangerous.

That’s why the CEO’s emphasis on proprietary information assets matters. General AI models can be impressive, but they typically don’t come with:

• Verified provenance and consistent coding across datasets
• Longitudinal patient journey context (where legally obtained and properly de-identified)
• Embedded compliance and auditability
• Deep integration into trial operations and regulatory expectations

What Singapore businesses can learn (even outside healthcare)

You don’t need to be in pharma to benefit from this lesson.

If you’re running a Singapore SME or mid-market firm adopting AI for marketing, operations, or customer engagement, your advantage often comes from building an internal “data spine”:

• Clean customer master data (one customer, one profile)
• Consistent tagging for enquiries, tickets, and sales stages
• Standard operating procedures captured in a searchable knowledge base
• Clear permissions: what data AI tools can access, and what they can’t

Most companies buy an AI tool and hope for magic. The ones who win treat AI as a force multiplier on well-managed systems.

A practical adoption plan: how to invest in AI when the payoff feels uncertain

Answer first: The safest approach is a portfolio: one “productivity win,” one “customer win,” and one “data moat” project running in parallel.

If you’re feeling the same tension investors expressed—AI is exciting, but will it pay off?—this three-track plan works well for Singapore businesses because it balances quick results with long-term advantage.

Track 1: Productivity (prove value fast)

Pick one high-volume process that’s currently slow:

• Sales: meeting notes → CRM updates → next-step emails
• Customer service: ticket triage + draft replies + knowledge retrieval
• Ops/finance: invoice matching, document classification, variance explanations

Success metric examples:

• Reduce handling time by 20–40%
• Cut first response time by 30–50%
• Increase throughput per employee without hiring

Track 2: Customer experience (protect revenue)

Choose one customer-facing workflow where AI increases speed and accuracy:

• Faster, more consistent proposals
• Personalised follow-ups based on interaction history
• Better lead qualification using structured scoring + human review

This is where AI business tools Singapore searches often concentrate: teams want tools that help them respond faster and close deals with fewer errors.

Track 3: Data moat (make yourself harder to replace)

This is the IQVIA-style bet.

Invest in assets that make your AI results uniquely useful:

• A proprietary dataset (e.g., anonymised service outcomes, process benchmarks)
• A domain knowledge base that’s kept current (policies, product rules, edge cases)
• A governance layer (audit logs, approval flows, access control)

A good internal line to use is:

“Models are rented. Moats are built.”

People also ask: “Will AI displace my services business?”

Answer first: AI will displace tasks, not whole businesses, unless your value is purely task-based and you don’t adapt your pricing and positioning.

If your offering is “we do X manually,” you’re exposed. If your offering is “we achieve outcome Y with documented controls, domain expertise, and accountability,” AI usually strengthens you.

Here’s a concrete repositioning example I’ve seen work:

• Before: “We provide monthly reporting and analysis.”
• After: “We run an always-on performance monitoring system with weekly actions, tracked outcomes, and board-ready reporting.”

Same domain. Different defensibility.

What to do next if you’re evaluating AI business tools in Singapore

IQVIA’s earnings call wasn’t really about IQVIA. It was a public version of the discussion happening in boardrooms everywhere: AI is accelerating, and the market is punishing vague stories.

If you’re building your 2026 plan now, take a stance:

• Use AI to automate commodity work aggressively.
• Invest in data quality and governance so AI outputs are reliable.
• Protect your moat by anchoring AI to proprietary workflows and owned datasets.

If you want help choosing the right AI stack (and avoiding tools that look good in demos but fail in production), start with an audit: where the data lives, which workflows matter, and what “good” looks like in measurable terms.

The question worth ending on is simple: If a competitor bought the same AI tools tomorrow, what would still make your business hard to copy?

Source article (landing page): https://www.channelnewsasia.com/business/iqvia-backs-ai-strategy-analysts-question-impact-business-5910261