Alarum (ALAR) Q4 2025 earnings review
Huge AI Revenue Growth Comes at the Cost of Margins
Alarum delivered a mixed Q4. On the surface, revenue grew 60% YoY to $11.8 million, driven by expanding artificial intelligence (AI) data workloads. However, the underlying financials reveal a painful transition: net profit halved to $0.2 million, and Adjusted EBITDA fell 33% YoY to $1.0 million. The company is actively sacrificing near-term profitability to capture enterprise AI market share, leading to a severe gross margin compression. Furthermore, while YoY metrics look spectacular, revenue is actually decelerating sequentially—dropping from $13.0 million in Q3 to $11.8 million in Q4, with Q1 2026 guidance pointing to a further slide to $11.0 million.
🐂 Bull Case
Alarum is successfully embedding itself into the foundational layer of the global AI boom. Full-year revenue jumped 28% to $40.7 million, proving their data collection tools are essential for large language model training.
The company is successfully graduating from self-service proxy customers to deep enterprise engagements, resulting in a significant increase in average revenue per customer (ARPU).
🐻 Bear Case
Gross margins plunged from 72.4% in 24Q4 to 53.8% in 25Q4. Winning massive enterprise AI workloads requires heavy upfront infrastructure investments and third-party costs that are crushing unit economics.
Net Retention Rate (NRR) has steadily collapsed from 1.27 a year ago to just 0.83 today. While AI wins are masking the damage on the top line, the core legacy business is leaking customers rapidly.
⚖️ Verdict: ⚪
Neutral. Alarum is successfully capturing a massive macro tailwind in AI data collection, but the transition is incredibly expensive and volatile. Investors must accept structural margin degradation and sequential revenue lumpiness in exchange for a seat at the AI infrastructure table.
Key Themes
Gross Margins in Freefall
The gross margin trajectory is reversing sharply. IFRS gross margin collapsed from 75.1% in FY24 to 58.5% in FY25 (and 53.8% in Q4 specifically). Management cited the necessity of larger server volumes and stronger infrastructure to service large-scale AI clients. Until Alarum can bring third-party product dependencies in-house, COGS will continue to heavily weigh on profitability.
AI Data Infrastructure Pivot
The massive macro tailwind of AI model training is driving the company's growth. Alarum is evolving from a simple proxy provider to an AI data infrastructure platform. The demand for large-scale public web data used to train and refresh AI models has led to a significant increase in data workloads, structurally changing the company's growth trajectory.
Product Portfolio Expansion
Growth is no longer reliant on just the core proxy network. Alarum is seeing accelerating adoption of newer product lines, specifically the Website Unblocker, search engine results page (SERP) solutions, and large-scale dataset delivery. This diversification allows Alarum to capture more value from complex web scraping operations.
Net Retention Rate Collapse Contradicts Growth Narrative
Management boasts of strong year-over-year revenue growth, but the Net Retention Rate (NRR) reveals a reversing trend, plummeting to 0.83 in Q4 (down from 1.27 a year ago and 0.98 in Q2 25). An NRR below 1.0 means revenue lost to churn and down-sells outpaced revenue gained from up-sells. The AI vertical's explosive growth is currently masking a rapid deterioration in Alarum's legacy customer base.
Sequential Revenue Deceleration
While YoY growth looks phenomenal (+60%), the sequential trend is decelerating. Following a massive spike to $13.0M in Q3 2025 (driven by a few huge AI customers), Q4 revenue cooled to $11.8M, and Q1 2026 is guided even lower to $11.0M. Management notes that AI developers are in the experimentation phase, causing lumpy, unpredictable demand patterns.
Other KPIs
Accelerating. Up from $5.0 million a year ago. Full-year Opex jumped to $23.6M from $17.2M. This reflects a deliberate strategy to aggressively scale the organization, hiring for R&D, product development, and customer success to handle enterprise-grade workloads.
Stable. The balance sheet remains healthy, down slightly from $25.0 million at the end of 2024. The company generated $1.0 million in net profit for the year, allowing it to fund its massive infrastructure expansion largely out of operating cash flow without dilutive capital raises.
Guidance
Decelerating sequentially from Q4's $11.8M, though it represents a 54% YoY increase. This confirms the volatility warning issued by management regarding AI customer R&D phases, where data consumption occurs in massive, irregular bursts rather than smooth SaaS-like MRR curves.
Accelerating sequentially from Q4's $1.0M, though slightly down from $1.5M in the same quarter last year. This suggests that some of the heaviest upfront infrastructure costs from late 2025 might be normalizing, allowing slightly better flow-through to the bottom line.
Key Questions
NRR Stabilization Timeline
With Net Retention Rate dropping to 0.83 due to legacy market churn, at what point do you expect the legacy attrition to bottom out and the NRR to reflect the net expansion of the new AI enterprise cohort?
Margin Floor
Gross margins collapsed to 53.8% in Q4. As you transition third-party tools in-house, what is the long-term structural gross margin target for the business, and when will we see the trough?
Sequential Revenue Volatility
With Q1 2026 revenue guided lower sequentially for the second quarter in a row, how much of this is driven by completion of specific AI training runs versus losing share to competitors?