Nội dung
Two recent AI stories from MarTech highlight a reality that many Vietnamese marketers are also facing: AI technology is not just about choosing better tools, but about evaluating investment models and controlling output quality. On one side is the wave of “agentic commerce” with its many promises of reshaping the shopping journey; on the other is a very familiar pain in content operations: AI produces more, but not necessarily better.
Looking at these two stories side by side, it becomes clear that the marketing challenge today is no longer whether to use AI, but how to use AI to create real value, avoid cost bloat, and maintain brand standards.
When agentic commerce is promoted as the “future,” what lens should marketers use?
The MarTech article opens with a scenario familiar to many CMOs: a flood of vendor proposals all built around a 12-month payback period, very confident in year one but much less clear when looking into year three. With agentic commerce, the story is even easier to get swept up in through impressive demos, technical acronyms such as UCP, MCP, A2A, and promises to “rewrite” the entire sales funnel.
What stands out is that the author does not encourage evaluating these platforms based on hype, but instead suggests looking at a channel that has survived many technology cycles: direct mail. Although it may sound “outdated,” direct mail still exists because it forces marketers to think very practically about cost, measurability, usage lifecycle, and the durability of a channel as the surrounding technology changes.
The implicit message here is very clear: with agentic commerce, the question is not only “can AI do it,” but “will it be sustainable over the next 2–3 years, will it create more technical debt and operational debt, and will it align with the company’s overall strategy?” This is the cautious approach many marketing teams should learn when making AI software purchasing decisions.
Source: MarTech, “3 direct mail lessons for evaluating agentic commerce”.
Why can’t “workslop” be fixed with a few prompts or AI training sessions?
The second story addresses a problem very close to marketing team life: AI tools have been deployed, a prompt library exists, a brand voice guide has been written, AI training has been held, yet the output still includes half-baked drafts, slides that look polished but are empty inside, or newsletters that match the brief but miss the audience.
The main message of this MarTech article is that “workslop” is only a symptom; the root cause lies deeper in organizational structure and how work is designed. In other words, if a company only focuses on teaching users to prompt better without clarifying output standards, review processes, content decision rights, and professional accountability, then AI will only make output volume grow faster — not automatically improve quality.
This is especially important for marketers. In a content environment, AI can help speed up idea generation, summarize information, personalize messaging, or support large-scale production. But without editing, without customer positioning, and without someone ultimately accountable, teams can easily fall into “fake productivity”: they look busier, produce more, but spend even more time fixing things than before.
Source: MarTech, “Better prompts won’t fix your workslop problem”.
Two common signals from the two stories: AI marketing is entering a phase of discipline
When the two stories are placed together, it becomes clear that AI in marketing is moving from an enthusiastic experimentation phase to a phase where it must prove real-world effectiveness. On the technology-buying side, marketers need to evaluate agentic commerce using longer-term criteria rather than just looking at demos and expecting a quick payback. On the content operations side, teams need to understand that prompts are only a very small part of the quality system.
In short, AI cannot replace strategy, decision-making structure, or editorial capability on its own. If a company treats AI as an acceleration layer placed on top of a weak process, the result is often that old problems are produced faster. Conversely, if AI is placed within a system with clear standards, it can become a tool that amplifies the marketing team’s capabilities.
A perspective for the Vietnamese market
For Vietnamese businesses, these two stories are especially worth noting in a context where many marketing teams are simultaneously under pressure to cut costs, speed up content production, and experiment with AI. Before buying another agentic commerce platform or expanding content-generation tools, marketers should answer three questions: what exact business problem does this tool solve, who is ultimately responsible for output quality, and will it make the system more complex over the next 12–24 months?
At the operational level, instead of focusing only on prompts, businesses should standardize briefs, content approval workflows, editorial checklists, and quality measurement criteria. At the strategic level, AI needs to be evaluated as a long-term investment rather than just a short-term productivity hack. That is the only way AI can truly create a competitive advantage for Vietnamese marketers instead of just creating more workload.
See more marketing news and guides at https://marketing365.vn.
Follow more updates from Marketing365 to stay up to date with the latest marketing trends.
Read more articles in the Digital Trends category.
This article focuses on AI marketing with a perspective for the Vietnamese market.
References
- 3 direct mail lessons for evaluating agentic commerce
- Better prompts won’t fix your workslop problem
