Blackpearl Group reported FY26 annual recurring revenue of $26.8 million, up 114% from a year earlier.
The New Zealand-listed software company added $10.6 million in net annual recurring revenue during the year, more than double the FY25 level. The result was ahead of a Bell Potter forecast of $25.3 million.
The update comes as investors scrutinise which parts of the artificial intelligence market are converting infrastructure spending into commercial returns. Blackpearl argues the strongest gains are shifting toward specialist systems built for narrow business tasks rather than broad foundational models.
Third-party benchmarking showed its Pearl Engine could outperform foundational models by up to 25 times in B2B sales and marketing work, which Blackpearl attributes to its use of proprietary data, buyer-intent signals and workflow integration in customer acquisition tasks.
Chief executive Nick Lissette said the latest numbers reflect a broader shift in where AI value is accruing. "$26.8 million ARR - up 114% - is a signal of where the economics of AI are increasingly accruing. Investors are looking beyond foundational AI infrastructure and reassessing the value of application-layer companies built around solving specific operational problems," he said.
Lissette said Blackpearl made an early decision to build its own systems rather than rely solely on external models. "We made a deliberate decision early on to invest heavily in building our own AI infrastructure and intelligence layer rather than simply sitting on top of someone else's model," he said.
Commercial metrics
Alongside revenue growth, Blackpearl reported a customer acquisition cost payback period of 3.5 months, compared with sector averages it said typically range from 18 to 24 months.
Its Data-as-a-Service business accounted for 36% of annual recurring revenue and has recorded zero churn since inception, according to the company.
The DaaS model allows agencies and larger customers to feed Blackpearl's enriched intent data into their own systems and workflows. Blackpearl presented the product as evidence that customers will pay for data and model outputs tailored to specific sales and marketing processes.
Its platform ingests more than 31 billion B2B data signals each day from more than 330 data partners. Blackpearl uses what it describes as an augmented large language model to identify buyer intent, automate outreach and improve customer acquisition performance.
Lissette said that design gives Blackpearl an advantage over software groups that added AI tools later. "This wasn't a software platform with AI bolted on later. Pearl Engine is the business," he said.
He added that the basis of competition in AI is changing. "That matters because the competitive advantage increasingly comes from proprietary data, feedback loops and domain-specific intelligence rather than access to generic models," he said.
Investor view
Blackpearl's comments come amid a broader debate in equity markets over how much value will remain with chipmakers and cloud providers, and how much will shift to software and data businesses. Nvidia's recent results have underscored the scale of global spending on AI infrastructure, while application-layer companies are trying to show they can capture a growing share of the economic benefit.
Blackpearl said analyst interest in that theme is rising. Unified Capital Partners recently initiated coverage with a Buy rating and an AUD $1.30 price target, describing the stock as potentially "the cheapest tech leverage on the ASX" relative to its growth profile.
The company also linked its outlook to changes in digital commerce and advertising. It said AI-generated traffic is making older web-based signals less reliable and increasing the value of systems that can distinguish real buyer intent from automated activity.
Industry commentary cited by Blackpearl suggested bot traffic grew eight times faster than human traffic during 2025. It said that trend is pushing businesses to seek more accurate sales and marketing data.
Blackpearl also pointed to forecasts for agentic AI spending that project the market will rise from about USD $5 billion to USD $236 billion by 2034. Those projections have helped support investor interest in companies offering more targeted AI products for business operations.
Lissette said the market is beginning to draw finer distinctions between different AI business models. "The market is starting to separate AI as a feature from AI as infrastructure. The companies creating durable value are the ones building proprietary intelligence layers that become deeply embedded inside how businesses actually operate," he said.