The Investment Case
TEM is interesting because its AI is connected to real diagnostic and pharmaceutical workflows. The risk is that a good business already carries a demanding price.
Why this could work
Why the market may be right
AI is entering real workflows
Hub, Lens, ECG-AI, and Tempus One reduce work for clinicians and pharmaceutical researchers.
Cash conversion remains weak
The H1 simple-FCF proxy was about negative $102M. Positive adjusted EBITDA is not sufficient.
Testing continuously expands the dataset
Each test adds genomic and clinical information that can support new research contracts and products.
Personalis is an expensive test
The $1.5B transaction enterprise value is about 17 times Personalis annualized Q2 revenue.
Growth has several engines
Oncology testing, MRD, data licensing, cardiology AI, and pharma partnerships create different revenue paths.
The price requires strong execution
The Base DCF is only about $21, so a revenue multiple explains most of the current value.
How the Business Works
Tempus is trying to combine a diagnostics company, a healthcare-data company, and an AI-software company in one system.
A tissue or blood test produces genomic information. Tempus connects that result with medical records and treatment outcomes. Clinicians use the tests and AI tools to support treatment decisions. Pharmaceutical companies use Lens, data licenses, and services to find patients, study biomarkers, and design trials. Diagnostics generated about 76% of Q2 2026 revenue, with Data & Applications contributing the remaining 24%.
The value of the AI products lies less in the chatbot interface than in their position between proprietary clinical data and actual healthcare work. Hub supports clinical workflows, Lens automates research analysis, and ECG-AI searches standard electrocardiograms for disease signals. The new Moderna and Merck agreement also puts Tempus into the testing workflow for a potential individualized cancer vaccine. The company does not disclose stand-alone AI-product revenue, however. What investors can verify today is platform-wide testing and data growth, not separate software economics.
Diagnostics earns revenue from test volume and reimbursement per test. Data & Applications earns revenue from pharmaceutical data licenses, analytical services, and software agreements. The intended flywheel is straightforward: more testing creates more data, better data attracts more research and product use, and that use creates more testing and contracts. A durable moat requires not just a large dataset, but exclusive access, clean linkage, and repeated use in clinical decisions.
Testing creates clinical data
Oncology and hereditary tests connect genomic results with medical records.
Data improves AI products
Hub, Lens, ECG-AI, and foundation models learn from a growing set of cases.
AI enters clinical and research workflows
Clinicians and drug developers use the system to make decisions and design studies.
Use drives more tests and contracts
Greater adoption creates new data and strengthens the cycle.
TEM expects roughly $1.6B of 2026 revenue and already monetizes a diagnostic network and data licenses. RXRX uses laboratory data and AI to discover drugs, leaving more value dependent on clinical outcomes and partner milestones. TEM has lower scientific binary risk, but greater valuation and integration risk. RXRX is smaller and more sensitive to pipeline outcomes. Their data agreement through 2029 suggests that the two platforms are more complementary than directly competitive.
TEM and RXRX: The Same Healthcare-AI Label, Different Investments
Both companies use healthcare data and AI, but they monetize at different points and require different evidence from investors.
TEM begins with patient testing. It links genomic results to clinical records and monetizes that system through care delivery and pharmaceutical research. Its revenue base is larger and no single drug trial determines the company’s future. The main risks are cash conversion, the Personalis purchase price, and dilution.
RXRX begins in the laboratory. It uses biological data and AI to discover drug candidates. Successful programs and partner milestones can create substantial value, but clinical failure or delay carries much greater consequences. Pipeline progress and partner decisions matter more than current revenue.
Their partnership makes the distinction clear. Tempus supplies clinical and genomic data, while RXRX supplies an RNA foundation model. The agreement now runs through November 2029 and replaces potentially discretionary fees with $42M of committed payments over three years. It is evidence of outside data monetization for TEM and a model-and-data input cost for RXRX.
What investors should compare
| Test | TEM | RXRX | Investment reading |
|---|---|---|---|
| Starting point | Patient testing and clinical data | Laboratory data and drug discovery | TEM is closer to healthcare services; RXRX is closer to biotechnology. |
| Current monetization | Testing, data licenses, analytics, and software | Partner agreements, milestones, and limited product revenue | TEM has a larger and relatively more recurring revenue base. |
| Proof of AI value | Testing volume, Data & Applications growth, and cash conversion | Candidate progress, clinical results, and partner expansion | The correct KPI is different even though both use AI. |
| Primary risk | Valuation, Personalis integration, SBC, and cash burn | Clinical failure, development delays, and partner dependence | TEM has more execution and price risk; RXRX has more scientific binary risk. |
| Valuation method | Revenue multiples cross-checked with long-term FCF | Cash plus pipeline option value | Applying TEM-style revenue multiples to RXRX would create a misleading comparison. |
TEM offers greater business certainty but little current margin of safety. RXRX may look cheaper, but its clinical probabilities require separate analysis. A more established business is not automatically the more attractive stock.
The KPI That Matters
The first numbers to watch are not product announcements. They are testing volume and Data & Applications growth.
Testing-volume growth × Data & Applications growth
More tests → better data → more clinical and pharma use → more tests and contracts
Q2 oncology volume grew 31%, Diagnostics revenue increased 20%, and Data & Applications revenue rose 28%. Insights revenue grew 36%. MRD volume increased from 6,500 tests in Q1 to 9,000 in Q2. If testing grows while data revenue slows, the flywheel is weaker than advertised. If both expand together, the data is reaching real contracts and product use.
Growth across testing and data
Both oncology volume and Data & Applications revenue grow above 25%.
Either metric grows between 15% and 25%.
Either metric falls below 15% while cash burn continues.
Financial Profile
Growth is strong and gross margin has improved, but profit has not yet become durable cash flow.
FY 2025 revenue rose 83% to $1.27B and gross margin reached 62.7%. FY 2026 company guidance calls for $1.595-1.605B of revenue and about $65M of adjusted EBITDA, excluding Personalis. The Base model includes some acquisition contribution in 2027E revenue of $2.10B, so that growth rate should not be read as purely organic.
Revenue and gross margin
Segment revenue mix and gross margin
Quality of Earnings & Capital Allocation
Positive adjusted EBITDA is progress, but cash flow and equity compensation determine the outcome for shareholders.
H1 operating cash flow was negative $80.8M. After $21.1M of capex and capitalized software, the simple-FCF proxy was about negative $101.9M. Q2 adjusted EBITDA was $8M, while stock compensation and related payroll tax reached $55.6M. The company expects roughly $200M of 2026 stock compensation and had about $1.21B of convertible-note principal outstanding.
The distance between adjusted EBITDA and cash flow
Personalis buys growth and dilution at the same time
The transaction enterprise value is about $1.5B. Tempus may elect up to 50% cash consideration. Depending on the final mix, Personalis holders could own roughly 8.0% to 14.8% of post-close Tempus Class A shares. The Base model uses 207M diluted shares, but a different mix changes per-share value.
The Model
The Base model includes Personalis growth and integration costs. The 2027 growth rate therefore includes acquisition effects.
What anchors the estimates
Recent results and company guidance set the starting point. Revenue, margins, and cash flow after 2026 are Lazy Valuation estimates.
| Evidence | Value | Type | Source | What it supports |
|---|---|---|---|---|
| Q2 revenueQ2 2026 | $382.5M / +22% | Actual | S1 | Sets the starting pace for the near-term model. |
| Segment growthQ2 2026 | Diagnostics +20% / Data & Apps +28% | Actual | S1 | Shows that testing and data revenue are expanding together. |
| FY 2026 guidanceFY 2026 | $1.595-1.605B revenue / ~$65M adjusted EBITDA | Company guidance | S1 | Anchors the near-term revenue and adjusted-EBITDA estimates. |
| Personalis acquisitionAnnounced 2026 | $1.5B transaction EV / 8.0-14.8% potential ownership | Actual | S3 | Adds MRD capability and growth, but also purchase cost and dilution. |
| H1 cash flowH1 2026 | -$101.9M H1 simple FCF proxy | LV estimate | LV model | Explains why positive adjusted EBITDA is not enough. |
| 2027E owner-FCF proxy2027E | $50M owner-FCF proxy | LV estimate | LV model | Provides the common starting point for the DCF tests. |
The Personalis revenue contribution and final consideration mix remain uncertain. Organic growth and acquisition effects must be separated after closing.
Testing volume
Oncology and MRD adoption support Diagnostics growth.
Data revenue
Long-term pharma agreements and Lens usage expand higher-margin revenue.
Personalis
The model includes acquired revenue and integration cost beginning in 2027.
Cash conversion
SBC and integration costs must fall before adjusted EBITDA becomes owner FCF.
2027E cash-flow bridge
The Base model starts with $180M of adjusted EBITDA and deducts $130M for interest, taxes, investment, working capital, and Personalis integration, leaving a $50M owner-FCF proxy.
View full Base model
| Year | Revenue | Growth | Adjusted-EBITDA margin | FCF | FCF margin | SBC / revenue |
|---|---|---|---|---|---|---|
| 2026E | $1.60B | 25.8% | 4.1% | -$100M | -6.3% | 12.5% |
| 2027E | $2.10B | 31.3% | 8.6% | $50M | 2.4% | 10% |
| 2028E | $2.62B | 24.8% | 11.8% | $150M | 5.7% | 8.5% |
| 2029E | $3.28B | 25.2% | 15.9% | $300M | 9.1% | 7% |
| 2030E | $3.85B | 17.4% | 19% | $500M | 13% | 6% |
| 2031E | $4.35B | 13% | 22.1% | $750M | 17.2% | 5% |
LV estimate Only FY 2026 revenue and adjusted EBITDA use company guidance. All later rows are Lazy Valuation estimates.
What’s Priced In?
The reverse DCF asks how fast $50M of 2027E owner FCF would need to grow to support the current value plus the Personalis purchase.
Cash-flow growth implied by the price
Starting-FCF sensitivity
$50M → 163.4% · $80M → 133.3% · $100M → 120.2%
Even doubling 2027E FCF from $50M to $100M leaves the implied growth rate above 100%. The price depends more on revenue growth and future margins than on near-term cash flow.
Reverse-DCF disclaimer
The market-implied rate is a Lazy Valuation model output, not an external forecast. It is unusually sensitive because the starting FCF is small. The reverse DCF is therefore a stress test of expectations, not a precise prediction.
The Base path grows from $50M in 2027 to $750M in 2031, a 96.8% four-year CAGR. The reverse DCF requires about 163.4%.
Valuation
The report uses owner-FCF DCF and 2027E pro forma revenue multiples. The gap between the methods is itself the main risk.
Owner-FCF DCF produces $4.5 in the Bear case, $21.4 in Base, and $46.4 in Bull. Applying 4x, 6x, and 8x to 2027E pro forma revenue, after acquisition cost and dilution, produces $32.6, $58.2, and $86.5. Because near-term cash flow is small, the market is likely to trade the stock on revenue growth for now. The composite therefore weights DCF at 25% and the revenue method at 75%.
| Method | Period / timing | Bear | Base | Bull |
|---|---|---|---|---|
| Owner-FCF DCF | 2027E-2031E Owner FCF discounted to Oct. 7, 2026 | $4.5 | $21.4 | $46.4 |
| 2027E pro forma EV / revenue | 2027E Pro forma revenue with Personalis cost and dilution | $32.6 | $58.2 | $86.5 |
| 25% DCF / 75% revenue multiple | Current value 25% DCF and 75% revenue multiple | $25.6 | $49 | $76.5 |
LV estimate The Base composite is about $49. The same price is the return ceiling when a 12-month Base value of $58 is discounted at 18%. A strict 25% margin of safety gives about $37.
Scenario value ranges
Scenario Map & Thesis Breakers
At $71.41, TEM is above the Base range and already inside the Bull range. Cash conversion now matters more than another attractive announcement.
Testing growth slows, data contracts become less predictable, and Personalis integration costs persist.
Owner FCF remains near zero in 2027 and reaches only $280M in 2031. The revenue multiple falls to 4x.
- Composite value
- $25.6
- Confirmation
- Oncology volume growth, Data & Applications growth, and cash conversion all weaken together.
Testing and data revenue sustain growth in the 20s, while Personalis closes without material customer losses.
Owner FCF rises from $50M in 2027 to $750M in 2031. The revenue method applies 6x to 2027E sales.
- Composite value
- $49.0
- Confirmation
- Q3 guidance is met and exit-quarter cash flow turns positive.
MRD, data licensing, and AI products grow together while Personalis integration improves margins faster than expected.
Owner FCF reaches $1.2B in 2031 and the revenue method applies 8x to 2027E sales.
- Composite value
- $76.5
- Confirmation
- Testing and Data & Applications both grow above 25%, with diluted shares near 204M.
Thesis breakers
Two or more persistent failures would require a lower Base model.
- Oncology testing growth remains below 15%.
- Data & Applications revenue growth falls below 15%.
- Annual FCF is still negative in 2027.
- The deal is materially delayed, or customer losses and integration costs exceed the model.
- SBC remains above 10% of revenue in 2027.
- Convertible debt and acquisition funding increase the probability of another capital raise.
What I’m Watching Next
The next quarter is a test of whether growth is beginning to convert into cash.
| KPI | Healthy | Watch | Warning |
|---|---|---|---|
| Oncology-volume growth | >25% | 15-25% | <15% |
| Data & Applications growth | >25% | 15-25% | <15% |
| Quarterly adjusted EBITDA | >$20M quarterly | $10-20M | <$10M |
| Quarterly cash flow | Positive exit-quarter FCF | Near break-even | Cash burn persists |
| Personalis integration | On-time close / stable retention | Minor delay or cost | Material delay or customer loss |
| Diluted shares | <205M diluted shares | 205-215M | >215M |
A strong quarter must do more than exceed $410M of revenue. Testing and data need to grow together, adjusted EBITDA must begin to reach cash flow, and the final Personalis cost must remain within the modeled range.
Bottom Line
Tempus has interesting AI products and a credible data moat. The current price already recognizes much of that potential.
TEM has a larger revenue base and less scientific binary risk than RXRX. Its valuation, however, requires a high revenue multiple, rapid cash conversion, and successful Personalis integration at the same time. The Moderna and Merck collaboration and new ECG-AI clearances improve the business, but they do not close the cash-flow gap today.
The current conclusion is Watch. A price below the top of the Base range, or clear evidence of positive exit-quarter cash flow and acceptable Personalis terms, would justify recalculating the entry price.
Confirming a good AI product and buying its stock at a good price are separate decisions.
Sources & Disclosure
- S1Tempus reports second-quarter 2026 resultsCompany earnings release · 2026-07-30
- S2Tempus Form 10-Q for the quarter ended June 30, 2026SEC filing · 2026-07-31
- S3Tempus agreement to acquire PersonalisCompany release · 2026-08-14
- S4Tempus and Personalis merger registration statementSEC filing · 2026-09-18
- S5Tempus Q2 2026 corporate deckCompany earnings materials · 2026-07-30
- S6Tempus and Recursion extend data partnershipCompany release · 2026-09-21
- S7Tempus announces collaboration with Moderna and MerckCompany release · 2026-10-07
- S8TEM historical price and market dataMarket data · 2026-10-07
- S9Tempus investor relations company highlightsCompany website · 2026-10-08
Position disclosure As of the publication date (October 8, 2026), the author did not hold a position in TEM.
General disclaimer Lazy Valuation provides general information, not personalized investment advice or a recommendation to buy or sell any security. Estimates and valuation ranges are uncertain and may be wrong. Readers should verify the source materials and make decisions appropriate to their circumstances.
Discussion
Leave a question or a point of view. Comments appear immediately.