Contents
- AI and the future-revenue test: the market only pays for scale it can verify
- Changes in AI business models are reaching the costs marketing teams must bear
- Vietnam’s AI buying problem: choose tools by total real cost, not by the feeling that they are cheap
- What to do to keep AI worth the money in your marketing plan
- Sources
- Reference sources
The AI market is moving into a phase where value is no longer measured by feature counts or user-growth speed, but by future revenue and the amount of cash companies must spend to catch up. For Vietnamese marketers, this is not just an investor story; it signals that budgets, tools, and the way AI capability is bought will be tightened under the same logic: anything that cannot prove output will be harder to fund.
Sources from Reuters, Financial Times, WSJ and Anthropic point to one common theme: major AI companies must prove that their business model is large enough to support their valuation, while AI technology itself is entering an environment with more agents, more interactions, and more operational risk. In short, AI is no longer being asked only “what can it do?” but also “how much money can it make, and how much does it cost to run reliably?”
- Key point:
- Anthropic is being viewed by the market through a 2028 revenue lens, not just current growth, according to Reuters.
- Financial Times shows investors are betting on an extremely high valuation, so the pressure to prove business scale will be greater.
- Anthropic also acknowledges that AI agent will interact more in the real world, bringing higher risk and control costs.
- For Vietnamese marketers, the important thing is to prepare budgets, processes, and tool-selection criteria based on real cost, not technological hype.
AI and the future-revenue test: the market only pays for scale it can verify
Reuters says Anthropic is being valued with a focus on a projected 2028 revenue of around $190–200 billion, while Financial Times says investors are betting on a $2 trillion valuation in a record IPO. These two data points point to the same thing: the AI market has moved beyond the stage where a growth story alone is enough. Now, that story must come with a revenue path long and large enough for investors to commit capital.
What stands out is how AI valuation is shifting from the present to a farther future. Reuters describes Wall Street looking at revenue two years ahead rather than just current run-rate revenue; Financial Times shows that the expectation premium is large enough to push valuations to rare levels. For marketers, this is a signal that anything tied to AI will soon be judged by the same standard: does it generate cash flow, or only the feeling of progress?
At the operational level, this affects how companies buy AI tools. Products that promise a lot but are hard to tie to KPIs, hard to convert into revenue, or hard to prove as cost-saving will face tougher competition from tools that can show clearer impact. In other words, AI budgets are no longer unlimited experimentation budgets.
Changes in AI business models are reaching the costs marketing teams must bear
Anthropic’s IPO and valuation discussions are not just about one company. They reveal how the entire AI industry is being forced to prove profitability within a limited time. Reuters makes clear that investors are using revenue-forecast-based multiples, while FT shows market expectations are high enough to demand a business scale very different from today’s level.
2028 revenue: marketing teams must connect AI to output, not just experimentation
Reuters says Anthropic projects 2028 revenue at around $190–200 billion, while the run-rate revenue figure the company disclosed in May was only $47 billion. That gap shows the market is not valuing AI based on today’s state, but on the growth trajectory the company says it will reach. For marketing teams, the lesson is that every AI expense must answer the question: how much pipeline does it add, how much production cost does it reduce, or how much faster does it serve customers? Without a way to measure that, AI can easily be seen as a nice-looking cost on a slide.
Valuation by multiples: tool budgets must align with clearer KPIs
Reuters describes bankers and investors using enterprise value-to-revenue multiples to assess Anthropic, while Financial Times shows those expected multiples are being pushed very high. This perspective directly affects procurement: every AI software product will be compared against the value it creates for each dollar spent. In marketing, that forces teams to move from asking “can we use it?” to “after using it, how many hours of work does it save, how many leads does it add, or how many revision cycles does it cut?” Tools that cannot answer that will struggle to keep their budgets.
As agents do more work, control and monitoring costs will rise
Anthropic writes that AI agents will work in shared codebases, markets, and other social systems, where agent-to-agent interaction may eventually exceed human-to-human interaction. It also emphasizes that agents have advantages in speed, working duration, and information processing, but remain prone to confabulation and reward hacking. As systems move beyond simple chat assistants, the cost is no longer just the price of the model; it also includes monitoring, testing, auditing, and error handling. For Vietnamese marketers, this is especially important if AI is being used to generate content, optimize media, or support customer service.
Read more: How Content, Data and Revenue Are Converging for SMB Marketing
Vietnam’s AI buying problem: choose tools by total real cost, not by the feeling that they are cheap
In Vietnam, many companies still see AI as a feature layer that can be bought quickly to boost productivity. But signals from Reuters, FT and Anthropic show AI increasingly resembles a systems-cost item: license cost, integration cost, output-checking cost, data cost, and even reviewer cost. When a company looks only at a tool’s sticker price, the real expense is often elsewhere.
For example, a content team using AI to speed up production may save writing hours, but then spend extra hours editing, fact-checking, and controlling brand voice. A performance team may use AI to support analysis, but still need people to read data, compare dashboards, and block errors caused by prompts or input sources. Therefore, AI purchasing decisions should be based on total actual spend, not the first month’s price.
This is also the moment for Vietnamese companies to split AI budgets into three clear parts: a small experimentation budget to learn quickly, an operating budget for tasks that have proven effective, and a control budget for review, compliance, and risk management. Without that split, AI is very easy to buy as a trend and then leave idle because no one trusts the output.
What to do to keep AI worth the money in your marketing plan
- Set KPIs for each AI use case: hours saved, faster publishing, lower CPA, or higher close rate.
- Calculate total real cost, including license, integration, review, and supervision.
- Prioritize tools that can prove output on a dashboard or through a clear measurement process.
- Limit the scope of testing before expanding into full operations.
- Clearly define what the machine does and what humans review, to avoid the illusion of full automation.
Sources
- Reuters — Anthropic IPO valuation hinges on $190-200 billion 2028 revenue forecast, sources say — https://www.reuters.com/business/anthropic-ipo-valuation-hinges-190-200-billion-2028-revenue-forecast-sources-say-2026-08-15/
- Financial Times — Anthropic investors bet on $2tn valuation in record IPO — https://www.ft.com/content/840ac156-af1c-4a82-b260-ae791072fcfa?syn-25a6b1a6=1
- Anthropic — Patterns and problems in emerging multiagent systems — https://www.anthropic.com/research/multiagent-systems
- Wall Street Journal — Even Claude Is in the Dark About Dario Amodei’s Wife—and Her Influence at Anthropic — https://www.wsj.com/tech/ai/claude-dario-amodei-wife-anthropic-e1eeda7d
See more marketing analysis and guides at https://marketing365.vn.
Follow more analysis from Marketing365 to stay updated on the latest marketing trends.
Read more articles in the same category Digital Trends.
Reference sources
- Reuters — EXCLUSIVE: Anthropic IPO valuation hinges on $190-200 billion 2028 revenue forecast, sources say
- Financial Times — Anthropic investors bet on $2tn valuation in record IPO
- Anthropic — Patterns and problems in emerging multiagent systems
- WSJ — Even Claude Is in the Dark About Dario Amodei’s Wife—and Her Influence at Anthropic
