Adobe beats earnings but price drop—can AI strategy sustain growth?

Moomoo Insights

joined discussion ·

Tempus AI Earnings Deep Dive: AI Healthcare's "Data Flywheel" Is Starting to Translate Into Profits

After the close on Feb. 24, 2025, $Tempus AI (TEM.US)$ reported its 4Q and full-year results. The key takeaway: it shows that two of the hardest things to achieve in AI healthcare can happen at the same time—real scale in clinical volume (Diagnostics) and accelerating monetization of productized data (Data & Applications). More importantly, the company guided to ~$65 million in adjusted EBITDA for 2026, signaling a shift from a “growth-at-all-costs” narrative to a new phase defined by growth plus a clear profitability inflection.

Financial Breakdown: Two Growth Curves Rising in Tandem

4Q highlights

Total revenue: $367 million, +83.0% YoY

– Organic growth (excluding the Ambry acquisition): +33.5% — a useful way to separate M&A-driven growth from true underlying momentum

Diagnostics revenue: $267 million, +121.6% YoY

– Oncology testing volume: +29%

– Genetic testing volume: +23%

MRD testing: ~4,700 tests, +56% QoQ — one of the most important acceleration signals this quarter, suggesting ramp in what could become the next multi-billion-dollar market entry point

Data & Applications revenue: $100 million, +25.1% YoY

– Insights (data licensing): +69.5% underlying growth, excluding the impact of $AstraZeneca (AZN.US)$ warrants in 4Q24 — the data segment’s “acceleration” is back at the center of the story

Contract visibility: Remaining total contract value (TCV) > $1.1 billion; Net revenue retention (NRR) of 126% — existing customers aren’t just staying; they’re expanding

Liquidity: $760 million in cash and marketable securities — ample dry powder

Profitability: Operating leverage is starting to contain losses

– 4Q gross profit: $238 million, +94.7% YoY

– Adjusted EBITDA turned positive for the first time: $12.9 million (vs. -$7.8 million a year ago)

Full-year:

– Revenue: $1.3 billion, +83.4%

– Adjusted EBITDA improved by $97.3 million to -$7.4 million — notable given multiple acquisitions (including Paige AI and OneOme), implying meaningful operating leverage is emerging

2026 guidance: Slower growth, but higher-quality focus

– Revenue: ~$1.59 billion (about +25% YoY)

– Adjusted EBITDA: ~$65 million

This likely shifts investor attention from “how fast can it grow?” to “how durable is margin expansion—and can the business continue to shift toward higher-value data revenue?”

Tempus in the Bigger AI-Healthcare Trend: From “Model Wars” to “Data + Workflow” Execution

AI healthcare is undergoing a clear paradigm shift: the industry focus is moving from “who has the flashiest model” to “who can embed into real clinical workflows—and get paid by hospitals and pharma.”

The macro signals are increasingly explicit:

– OpenAI launched HIPAA/compliance-oriented offerings for healthcare organizations (e.g., ChatGPT Health) to bridge patient data and EMR ecosystems

– $NVIDIA (NVDA.US)$ and pharma giant $Eli Lilly and Co (LLY.US)$ announced a $1B AI drug R&D initiative

– The FDA introduced new frameworks to support continuous iteration for AI medical devices

Together, these point to one conclusion: AI healthcare is moving out of the lab and into scaled deployment, where the value chain shifts from algorithms to durable data assets, workflow integration, and reimbursement/payment loops.

Against that backdrop, Tempus stands out.

Over recent years, Tempus has aggressively built out an AI product portfolio spanning key stakeholders—patients, providers/care teams, and pharma—reinforcing that its moat isn’t a single algorithm, but a high-quality, multimodal data ecosystem tightly coupled to clinical use cases—a barrier that’s exceptionally difficult to replicate.

Based on aggregated 13F data, 500+ institutional investors now hold Tempus AI, with total position value exceeding $5.5 billion. Institutions appear to be making a concentrated bet on a data-driven future of AI-enabled healthcare with real-world impact.

Network Effects: How Tempus’s “Data Flywheel” Could Compound

Tempus repeatedly emphasized “network effects” on the call. In practice, this is a self-reinforcing loop powered by three core levers:

1. Diagnostics is the data intake engine

With oncology testing volume up 29% and genetic testing up 23%, every test expands Tempus’s multimodal dataset and feeds the AI models. The +56% QoQ acceleration in MRD suggests this intake channel is widening rapidly.

2. Data & Applications is the monetization engine

Insights grew 69.5% on an underlying basis, with 126% NRR. 19 of the world’s top 20 pharma companies are customers, and remaining TCV still exceeds $1.1B. That implies customers aren’t merely piloting—they’re scaling partnerships. As the data asset appreciates, this becomes a business model that pure testing companies struggle to match.

3. AI products are being embedded into hospital systems

Examples like a longitudinal molecular tracking collaboration with NYU Langone and expanded oncology testing access at Northwestern Medicine illustrate real workflow integration—not just lab-side research partnerships, but deployment inside clinical processes.

Worth highlighting: Immune Profile Score (IPS) Tempus claims this AI-derived score outperforms traditional biomarkers in predicting immunotherapy response and can identify patients who might otherwise be missed (e.g., 13% of colorectal cancer patients and 17% of rare cancer patients). If validated in larger clinical studies, IPS could become a key bridge for Tempus to evolve from a “testing provider” into a true clinical decision-enablement platform.

What to Watch Next

Tempus AI has largely validated Chapter 1 of its growth story (high growth) and is opening Chapter 2 (profitability improvement). 2025 results came in strong—especially the positive adjusted EBITDA in 4Q—helping address the market’s biggest concern: “burning cash without a path to profits.”

Key focus areas going forward:

1. TCV-to-revenue conversion: Remaining TCV of >$1.1B is a revenue “reservoir.” The pace of recognition will directly shape near-term performance.

2. Ramp of new tests and AI products: Adoption curves for MRD, HRD-RNA, and clinical validation timelines for AI products like IPS.

3. Durability of profitability improvement: Can gross margin stay elevated (64.7% in 4Q) and can S&M and G&A ratios continue to decline with scale—ultimately supporting sustainable net profitability?

4. Mix shift toward data revenue: This is central to valuation expansion. Whether Insights can sustain ~70% underlying growth will influence whether the market prices Tempus more like a technology/data company rather than a testing company.

131K Views