Before Using AI in Marketing, Start with Context

Why Internal Data Alone Is Not Enough

As generative AI becomes more widely adopted, many companies are beginning to ask the same question: if we feed our internal information into AI, could we make our work more efficient?

Past proposals, meeting notes, chat messages, website copy, advertising copy, reports and workshop materials — when these materials are connected to AI, when these are connected to AI, the right information becomes instantly retrievable, accelerating strategy development, content creation, and day-to-day marketing work.

However, in the marketing environments that TAMLO works with every day, simply collecting information is not enough. For global companies to succeed in the Japanese market, it is essential to carefully align the following: what message does the overseas headquarters wants to communicate, what does the Japan-side sales team need, what do Japanese customers want to understand, and what does the brand need to protect?

At TAMLO, we typically begin projects with a two- to three-hour workshop with the client to clarify objectives, challenges and priorities. The purpose of this session is not simply to decide “what needs to be made”.

Instead, we explore questions such as:

• Why is this project needed?
• What challenges exist in the Japanese market?
• What does the global team want to prioritise?
• What is easy, or difficult, for the Japan-side sales team to communicate?
• Which pages, messages or content should be prioritised?

By aligning these perspectives, we define the direction of the project.

In other words, what TAMLO works with is not just words or deliverables. We organise the information held within a company, align the perspectives of different stakeholders, and clarify which decisions should be prioritised.

In the age of AI, the real value does not lie in information itself. It lies in the context: the challenges, assumptions and decisions that explain how that information came to be and will be used.

Aligning the assumptions behind decisions is the shortest path to success

In marketing for global companies entering or expanding in Japan, certain gaps appear again and again.

The overseas headquarters usually wants to maintain a globally consistent brand message. On the other hand, the Japan-side sales team needs explanations that feel clear and relevant to Japanese customers. Meanwhile, the marketing team wants to generate leads through websites, advertising, and content.

And throughout all of this, customers may not yet understand what the service is, why it matters, or how it is different from other options.

In this situation, simply replacing English materials with Japanese-language equivalents is unlikely to produce meaningful results.

What is needed first is alignment.

• Who are we trying to reach?
• What should they understand first?
• Which messages should remain close to the global version, and which expressions need to be adjusted for the Japanese market?
• Which content should be developed or improved first?
• Which performance indicators should guide future improvements?

Our workshops are designed to organise these questions.

What comes out of these sessions is not just a set of meeting notes. It is a decision-making framework for the project.

Feeding information into AI does not automatically create decision-making context

AI can process large amounts of information, such as meeting note summaries, past materials, and draft copy.

However, feeding all internal data into AI does not automatically lead to practical, usable decisions.

For example, a chat tool may contain messages such as “Let’s proceed with the direction discussed last time”, “Let’s put this option on hold for now”, or “This part needs to be adjusted for Japan”. But the chat alone may not explain what “the direction discussed last time” actually was, why an option was put on hold, or what market understanding led to the adjustment.

Similarly, past proposals and articles may show what was eventually produced, but they do not always reveal why a particular structure was chosen, why a certain expression was avoided, or why one page was prioritised over another.

What AI truly needs is not the volume of information, but the meaning attached to that information.

That meaning is often created through workshops, discussion and shared decision-making.

The goal is not to store data, but to make past decisions reusable

When companies think about AI adoption, they often start by asking what data they should collect.

• Should we preserve Slack or Teams logs?
• Should we organise Google Drive or SharePoint?
• Should we automatically save meeting notes?
• Should we connect email and calendars to AI?

These forms of information management are, of course, important — but they are not enough. In fact, collecting data without a clear purpose can increase noise and create more work for the team.

The more important question is: how do we capture our thinking in a way that can be used again and again?

For example, if AI is being used to support marketing strategy, the relevant information may include the client’s business challenges, insights shared in workshops, past proposals, market research, Search Console data, SEO reports and feedback from the sales team.

If AI is being used to support content creation, what matters may be past articles, advertising copy, brand tone, expression rules, client-specific language preferences and editorial direction.

If AI is being used for project management, the key information will be deadlines, owners, pending tasks, next actions, and agreements related to contracts and invoicing.

AI adoption should not mean placing all company data into one large container. It should mean designing the right information and context for each purpose.

We should not create more work for humans just to support AI

One of the risks in AI adoption is creating additional work simply so that AI can use the information later.

Recording every conversation carefully, just in case AI might need it one day. Entering every decision into a fixed format. Moving every important chat exchange into a database.

These processes may seem ideal in theory. But if they create too much work for the team, they will not last.

AI should not exist to increase human workload. It should help people focus on the decisions that only humans can make.

That is why, when thinking about AI adoption, TAMLO places more importance on purpose than on recording everything.

What matters is making the right information available when it is needed:

• Challenges identified through workshops
• Market assumptions uncovered through client discussions
• Brand tone reflected in past deliverables
• Customer reactions valued by the sales team
• Decisions actually adopted during the project

The goal is not to preserve everything. The goal is to make the right context accessible at the right moment.

For TAMLO, AI does not replace contextual judgement

AI can generate text, summarize information, search past materials, and produce first drafts of ideas.

But in Japan-market marketing for global companies, there are areas that AI alone cannot easily handle: understanding business contexts, aligning stakeholder perspectives, and deciding what should be prioritized to name a few.

In many cases, assumptions held by overseas headquarters differ from what local sales teams experience in the market. Global messaging may feel too abstract for Japanese audiences, while excessive localization can weaken brand consistency.

Making the right decision requires more than information. It requires dialogue.

It requires a process in which someone sits between stakeholders, aligns their perspectives and defines priorities.

TAMLO’s value does not lie simply in refining AI-generated text. It lies in designing the context and decisions that AI itself must rely on.

In the age of AI, workshops become even more important

At first glance, it may seem that the rise of AI will reduce the need for workshops and interviews.

In reality, the opposite is true.

The faster AI becomes at processing information, the more important it becomes for humans to define the right questions at the beginning.

• What should be treated as the real challenge?
• Whose perspective should be prioritised?
• Which information should be trusted?
• What outcome are we trying to achieve?
• Which decisions should AI be able to refer back to?

If these questions remain unclear, AI may produce outputs that sound convincing but are difficult to use in practice.

That is why TAMLO places importance on dialogue at the beginning of a project.

Through workshops, we identify challenges, align stakeholder perspectives and establish the decision-making criteria for the project. This process then informs the direction of website improvements, content creation, advertising, SEO, sales materials and AI adoption itself.

What is needed in the age of AI is not simply the use of AI. It is the design work that comes before AI is used.

Companies need to preserve more than final deliverables

When companies think about knowledge, they often think of final outputs: proposals, articles, reports and presentations.

These are important.

But in the age of AI, the real value is not found only in the finished deliverable. It is found in the decisions that led to it.

• Why was this page prioritised?
• Why was this message chosen?
• Why was this expression avoided?
• Why was this KPI selected?
• Why was the global message adjusted for the Japanese market?

These decisions are often invisible if we only look at the finished output.

Yet when a similar project appears in the future, they become extremely valuable references.

If companies want to use AI effectively, it is not enough to make past deliverables searchable. They need to be able to access and find the thought processes behind these decisions made

Start with decision logs

There is no need to begin by building a large knowledge management system.

It is often better to start small.

For example, after a workshop or meeting, companies can preserve a short record of:

• What was decided
• Why that decision was made
• Which alternatives were not chosen
• What risks were considered
• What thinking could be reused in a similar case

This can be treated as a decision log.

The important point is not to create more manual work for the team. It is to use AI to extract the essence of decisions from existing conversations and materials.

This approach is more realistic for day-to-day work. It also allows a company’s distinctive judgement to accumulate gradually.

Conclusion: AI needs context, not just information

When companies begin adopting AI, they often start by asking what data they should collect.

But there are more important questions that come before that.

• Why are we using AI?
• Which decisions do we want to make reusable?
• What information is needed to support those decisions?
• What context would be lost if AI only saw the data?

Without this order of thinking, AI adoption can become little more than information management.

TAMLO works with more than words. We work with the market, culture, brand, customers, sales realities and stakeholder misalignments behind those words.

That is why, at the beginning of a project, we engage in dialogues with clients, identify challenges, share perspectives and align the assumptions behind decisions.

In the age of AI, companies should not focus only on preserving the volume of information they hold. They should preserve the context that explains how that information was understood, judged and used.

Before deciding what to hand over to AI, companies must first look closely at how humans make decisions. Only then can AI become truly useful in practice.