AI MARKETING BUỔI 1

Summary

The session argues that effective AI marketing begins with marketing strategy and customer understanding, not with choosing AI tools. Using AI to generate and publish content without a clear audience, brand direction, or evidence of what works can produce generic content and weak results.

Core Marketing Framework

• Modernized 4Ps: The presenter reframes the traditional Product, Price, Promotion, and Place model:
  • Product → Brand: A good product still needs a distinctive brand.
  • Price → Value: Customers judge the value they receive, not just the price.
  • Promotion → Offer and sales pitch: Create a compelling reason to buy.
  • Place → Coverage: Build visibility and reach across relevant channels.
• Marketing funnel and content mix:
  • TOFU (top of funnel): 50–70% of content, focused on reach, attention, virality, and brand awareness.
  • MOFU (middle of funnel): 20–30%, focused on building trust and preference.
  • BOFU (bottom of funnel): 10–20%, focused on product and sales content.

The presenter warns that many businesses produce too much direct sales content, limiting reach and leaving sales teams with too few prospects.

• Demand creation: The presenter estimates that only 10–15% of viewers may already have a need to buy. Marketing should also educate and persuade people who do not yet recognize that need.
• Customer insight: Understand target segments, needs, motivations, and buying behavior before creating campaigns. Review sales outcomes with the sales team regularly and use actual customer data to refine targeting.

Marketing Roadmap and Execution Playbook

1. Develop customer insight. Identify the most valuable segments and the language and motivations that resonate with them.
2. Choose a communication concept. Examples include education, emotional stories, challenges, celebrities, unusual or controversial topics, useful tips, secrets, humor, trends, animals, and competitions. Choose concepts that fit the business and use them consistently to protect brand identity.
3. Select content formats. Options include short- and long-form videos, sales copy, storytelling, infographics, image posts, and large-text or scrolling-image formats.
4. Design the channel structure.
   • Official channels: Corporate brand and personal brand.
   • Satellite channels: Product or category channels, personal-brand satellites, and other supporting channels.
   • Channels can have different funnel roles. Some may focus on reach, while others carry more educational or sales content.
5. Research proven “winning concepts.” Find content that has already achieved the desired result, such as reach or conversions, rather than asking AI to invent content without evidence.
6. Create, publish, and improve. Adapt the underlying formula into original content, publish it, measure performance, run A/B tests, and refine the approach. The presenter recommends using AI agents for recurring measurement and reporting.

AI Marketing Workflow: “DNA + Dream 100 + Decode”

• Build business DNA for AI. Compile detailed information about the business, products, customers, positioning, strategy, messaging, funnel, content, advertising, and measurement. The presenter suggests a set of 150 questions to create a consistent “AI marketing brain,” rather than relying on generic prompts.
• Create a Dream 100 research list. Identify up to 100 channels, brands, or competitors that perform well against the company’s goals. For viral reach, study successful channels; for conversion-focused content, the presenter recommends using Kalodata.
• Decode proven content rather than copy it. Analyze the hook, structure, pacing, visuals, message, audience fit, and other elements behind successful content. Use those insights alongside the company’s DNA and product to create new variations—not to repost or duplicate the original.
• Measure and iterate. Track results, build a library of successful formulas, improve the content, and adapt effective concepts for additional channels.

Operational Recommendations

• Connect marketing closely with sales. Review which customer segments actually purchased and adjust the next day’s targeting accordingly.
• Don’t treat every channel as a direct revenue channel. Some channels primarily support customer coverage and awareness.
• Use AI with a defined business context and process. The presenter cautions that AI-generated copy can sound mechanical, vague, or off-topic when prompts lack context or content has not been validated.
• The session also demonstrates potential operational uses for AI, such as drafting customer-service replies and organizing inconsistent employee sales reports. These are presented as applications, not measured case studies.

Metrics and Examples Cited

The figures below are examples stated or shown by the presenter; the session does not independently verify them.

• Content volume: The presenter cites more than 2,000 videos per month as a possible benchmark for meaningful online coverage. Their own system reportedly produced up to 7,000 pieces of content in some months, with a team uploading around 4,000 videos per month.
  • At 2,000 videos, the stated funnel mix implies roughly 1,000–1,400 TOFU, 400–600 MOFU, and 200–400 BOFU videos.
• TikTok reach and leads: The presenter reports 44 million reach in 30 days and 77 million reach over 90 days, with approximately 19,700 phone-number leads attributed to organic traffic over the cited period. They also cite nearly 1 million followers across channels, excluding YouTube.
• Affiliate example: A toilet-paper seller is described as generating VND 126 billion in TikTok revenue over 30 days, with around VND 100 billion attributed to affiliate referrals—approximately 80%—and about 120,000 videos per month.
• Channel examples: An orthodontics channel is cited as having 1.3 million YouTube subscribers and videos with tens of millions of views. A Vietnamese tips-and-tricks channel is described as reaching 318,000 followers in about two months; the presenter later claims their “decode” approach helped build a channel system to nearly 600,000 followers in two months.
• Commerce examples: The presenter cites a skincare product with VND 24 billion in monthly sales, a toothpaste product with VND 8.5 billion in monthly sales, a toothpaste video associated with roughly VND 600–680 million in sales, and a livestream generating VND 400 million in one session.

Assignments and Next Steps

Participants are asked to complete two pieces of work over the following week:

1. Create their business DNA profile for AI.
2. Build a Dream 100 research list and a library of winning content, including formulas decoded from examples.

The presenter also suggests using an Excel template to record competitor and channel details, target segments, content performance, and the formulas behind successful posts.

Presenters and Sources

• Presenter: Nguyễn Tất Kiệm (identified in the video context as Kiểm Nguyễn Tất).
• Tools and sources referenced: ChatGPT, Claude, Gemini, Cloud Code and its Chrome extension, Kalodata, TikTok, YouTube, Facebook, and GitHub.