For much of the past two years, the global artificial intelligence race has revolved around one question:
Who can build the most powerful model?
Today, investors appear to be asking a different one:
Who can actually make money from AI?
Marketingforce Management Ltd. (HKEX: 2556), a China-based AI-native enterprise applications platform, may have offered one of the clearest answers so far.
The company recently issued a positive profit alert for the first half of 2026, expecting a significant increase in net profit compared with the same period last year. While headline earnings naturally attract investor attention, the broader significance may lie elsewhere: the announcement suggests that enterprise AI applications are beginning to transition from revenue growth to sustainable profitability.
A Shift Beyond Model Competition
The AI industry has evolved rapidly.
The first wave rewarded semiconductor manufacturers supplying GPUs.
The second favored foundation model developers racing to build increasingly capable large language models.
The third phase now appears to focus on commercialization.
As inference costs continue to decline and foundation models become increasingly commoditized, differentiation is moving away from model performance toward enterprise implementation, workflow integration and measurable business outcomes.
This transition is particularly important for enterprise software providers.
Unlike consumer AI applications, enterprise customers typically evaluate AI investments based on return on investment rather than technical novelty.
Selling Business Outcomes Instead of Software
Marketingforce’s strategy reflects this change.
Instead of positioning AI as a standalone software product, the company increasingly packages AI capabilities into industry-specific agents that automate sales, customer service, marketing and operational workflows.
The company’s “Scenario Token” model represents an evolution beyond traditional SaaS pricing.
Rather than charging solely for software licenses or API usage, pricing increasingly aligns with business outcomes generated by AI-powered workflows.
This approach potentially creates stronger customer retention while expanding pricing flexibility as AI performance improves.
For investors, that distinction matters.
Recurring enterprise value generally commands higher long-term valuation than one-time implementation revenue.
Internal AI Adoption Improves Operating Leverage
Another notable aspect of the earnings announcement is operational efficiency.
Marketingforce disclosed that internally developed AI employees have been widely deployed across marketing, sales, customer service and employee training.
This reflects a broader trend among enterprise software companies globally.
Leading software vendors increasingly use their own AI products internally before commercial deployment.
Doing so not only validates product capabilities but also improves operating leverage by reducing labor-intensive processes.
For software businesses, operating leverage often determines whether revenue growth ultimately translates into sustainable earnings growth.
Enterprise Knowledge May Become the Next Competitive Advantage
As foundation models become more widely available, competitive differentiation is likely to shift toward proprietary enterprise knowledge.
Marketingforce reports serving more than 210,000 enterprise customers across 721 industries.
Over time, these deployments create industry-specific knowledge bases that can be incorporated into AI agents and enterprise workflows.
Unlike model parameters, enterprise knowledge accumulated through years of customer engagement cannot easily be replicated.
This may become one of the most valuable assets in enterprise AI over the coming decade.
Looking Ahead
Investors will ultimately judge the sustainability of Marketingforce’s earnings improvement after the company’s interim financial results are released.
Key indicators will include:
AI application revenue contribution;
Gross margin trends;
Customer retention;
Operating cash flow;
Platform scalability.
Nevertheless, the latest earnings forecast may represent more than a company-specific milestone.
It could signal a broader turning point in which enterprise AI applications begin demonstrating that commercial adoption—not model capability alone—will define the next stage of value creation in artificial intelligence.
