Q2 2025 Earnings Call Reveals Contradictions in Customer Growth, EBITDA Margins, and Infrastructure Spending

The above is the analysis of the conflicting points in this earnings call

Date of Call: August 28, 2025

Financials Results

• Revenue: $8.8M, compared to $8.9M in Q2 2024 (down ~1% YOY)
• EPS: $0.04 per ADS (non-IFRS), compared to $(0.05) in Q2 2024
• Gross Margin: 63% (non-IFRS), compared to 78% in Q2 2024

Guidance:

• Q3 2025 revenue expected around $12.8M with a ±7% range (~78% YOY growth).
• Q3 2025 adjusted EBITDA expected around $1.1M with ±$0.5M range.
• New large-scale AI data project to contribute ~+$3M revenue in Q3; scope and duration still unclear.
• Near-term profitability from the new project limited; gross margins expected to decline further in Q3.
• Company investing in infrastructure and R&D to support major AI customer demand.

Business Commentary:

* Strong Revenue Performance and AI Market Drivers: - Alarum Technologies reported second quarter revenue of $8.8 million, with a net profit of $0.3 million and adjusted EBITDA of $1 million. - The growth was driven by the huge demand for training data for foundational AI models, with a significant increase in deal size potential from major companies.

• Shift in Customer Base and Marketdriver Impact:
• Alarum experienced a shift in customer base, with major tech giants and e-commerce companies becoming significant customers.

• The AI market drivers for this shift include increased demand for data collection, labeling, and model fine-tuning services, which are critical for AI model development.

• Investment in Infrastructure and Network Expansion:

• Alarum is investing strongly in infrastructure and IP proxy network expansion to meet the growing needs of major customers and support new data demand.

• This investment is aimed at optimizing network infrastructure and product delivery, despite the initial impact on gross margins.

• Large-Scale Data Collection Project Ramps Up:

• The company is working on a large-scale data collection project with a major online marketplace in Asia, expected to contribute approximately $3 million in revenue per quarter.

• This project is being done at lower gross profit margins due to the need to shape infrastructure costs for high-scale demand, impacting short-term profitability.

• Strategic R&D and Talent Development:

• Alarum is investing in R&D to expand its capabilities and broaden its product portfolio, focusing on developing cooperative data collection products for the AI era.
• The goal is to attract new customers and cross-sell to existing customers, thereby meeting more data needs under one roof.

Sentiment Analysis:

• Management guided Q3 revenue to ~$12.8M (~78% YOY) and highlighted strong AI-driven demand and a new large customer adding ~$3M per quarter. They remained profitable in Q2 (non-IFRS EPS $0.04) and hold ~$25M cash for investment. Caveats: gross margin fell to 63% (vs 78%) and is expected to decline further near term as they scale infrastructure for the AI project.

Q&A:

• Question from Brian David Kinstlinger (Alliance Global Partners): Why won’t the large customer ramp drive incremental EBITDA and why are gross margins low? Is this a discounted POC, and will economics improve over time?
  Response: Margins are temporarily lower due to heavy technology infrastructure (servers/cloud) costs for a massive dataset project; as they optimize infrastructure and scale/cross-sell, margins should improve.

• Question from Brian David Kinstlinger (Alliance Global Partners): Is the elevated cost driven by technology or people?
  Response: Technology infrastructure—servers, network, and cloud computing.

• Question from Brian David Kinstlinger (Alliance Global Partners): Do margins recover only with significant volume beyond the ~$3M/quarter addition?
  Response: Recovery will come from cost optimization on this project and adding more standard-margin projects; volume and mix both matter.

• Question from Brian David Kinstlinger (Alliance Global Partners): What is the product and how does it differ from existing offerings?
  Response: It’s a huge-scale data collection/dataset effort requiring extraordinary bandwidth, driving higher network/server/cloud costs versus typical products.

• Question from Brian David Kinstlinger (Alliance Global Partners): Outside this customer, how is broader demand and the new logo pipeline?
  Response: Usage is increasing with strong AI-related demand and many new logos; they’re investing in infrastructure and R&D to meet feature and scale needs.

• Question from William Kingsley Crane (Canaccord Genuity): Given customer mix shift and low NRR, could this lead to higher LTV and more stability?
  Response: NRR method lags and reflects the past; current customers are growing strongly and should improve retention/LTV as AI demand persists.

• Question from William Kingsley Crane (Canaccord Genuity): How long-standing is the large customer and how impactful were they to Q2?
  Response: Engagement began ~1.5 quarters ago; meaningful but not the largest in Q2, using multiple products with a significant ramp now.

• Question from William Kingsley Crane (Canaccord Genuity): How certain is the ~$3M Q3 contribution—committed or variable?
  Response: With two-thirds of Q3 completed, they have high confidence for Q3; beyond that, demand is promising but volatile and less predictable.

• Question from William Kingsley Crane (Canaccord Genuity): Gross margin outlook for Q4/Q1—improvement or continued pressure?
  Response: If growth is from existing products, margins should improve; new large-scale projects could pressure margins; ongoing investment targets long-term gains.