GEO for beginners: five basics to know before you start

Introduction: Getting your first step in GEO right

Many people who have become interested in GEO (AI search optimization) are not sure where to start.

Because GEO is a new field, the reality is that initiatives carried out without the right background knowledge tend not to produce results. This article explains five basics to know before you start GEO.

Basic 1: "AI search" is not just one thing

"AI search optimization" may sound like a single task, but there are several platforms to optimize for. The major ones alone include:

ChatGPT (OpenAI): One of the largest user bases in the world. Also has web search
Gemini (Google): Google's AI, integrated with Google Search. Also appears as AI Overviews
Claude (Anthropic): An AI growing rapidly in business use
Perplexity: An AI-native search engine. Strong on real-time information
Copilot (Microsoft): An AI connected to Bing search

Each of these AIs has different data sources, algorithms and answer tendencies. As a result, it often happens that a brand is recommended by ChatGPT but not mentioned by Gemini.

We explain each platform's characteristics in how AI search recommends brands, but you do not need to optimize for every platform from the start. It is efficient to start with the AI your target users use most.

Basic 2: GEO is not an extension of SEO (but it is not unrelated either)

As we explain in the differences between SEO and GEO, GEO calls for an approach that is fundamentally different from SEO. That said, the two are not completely separate.

Where SEO know-how still applies

Skills for creating high-quality content
Keyword research methods
The concept of E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness)
Knowledge of structured data

New perspectives GEO requires

Designing the information ecosystem around your whole brand, not individual pages
A new KPI, mention share, in place of search rankings
Monitoring across multiple AI platforms
Cross-industry competitor analysis based on Jobs-to-be-Done (JTBD)

If you already do SEO, we recommend building on that foundation and adding the perspectives that are specific to GEO.

Basic 3: The field has no established "right answer" yet

GEO is a younger field than SEO. That makes it important to understand the following:

What we know

AI gives weight to information from trustworthy sources
The volume and quality of mentions on the web influence AI answers
Structured information is easier for AI to reference
When information is consistent across multiple sources, AI trusts it

What we do not yet know

The specific details of each AI platform's algorithm (a black box)
Quantitative data on the direct correlation between mention share and final sales or conversions
Accurate information on when AI training data is updated
How market share in AI search will shift in the future

That is why there is value in starting early and building know-how in-house. In SEO, too, companies that got started in its early days gained a major advantage. We think GEO is likely to follow the same pattern.

Basic 4: Understand the metrics and methods for measuring results

Before you start GEO, decide which metrics you will use to measure results.

Key metrics

Mention share: Your brand's share of mentions among the tracked brands in AI answers (detailed explanation)
Mention position: The order in which your brand is mentioned in the answer
Mention sentiment: Whether the mention is positive, neutral or negative
Number of AI platforms: How many different AIs mention your brand

How to measure

You can ask each AI questions manually and record the results, but that approach has limits for ongoing monitoring. Ideally, you want a setup like this:

Ask the AIs the same questions at regular intervals (ideally every day)
Build a system to record and tally the results
Chart changes over time and track them

A tool like AIOPulse automates this work for questions asked in Japanese. For more on measuring results, see how to measure GEO ROI.

Basic 5: Treat GEO as a medium- to long-term effort, not a short-term one

The most common failure in GEO is the "We tried it for one month, saw no results, and stopped" pattern.

Why it takes time

AI training data takes time to update: Content you publish today will not necessarily show up in AI answers right away
Brand trust is not built overnight: Increasing mentions across multiple sources takes time
Statistically meaningful data takes time to collect: You need enough data to judge changes in mention share with confidence

A realistic timeline

Months 1–2: Assess where you stand and measure a baseline. Plan initiatives
Months 3–4: Start executing initiatives, such as content creation and PR
Months 5–6: Early effects start to show in mention share
Month 7 onward: Run the PDCA cycle and keep improving

The key is to look at medium- to long-term trends instead of reacting to every short-term result.

Practical checklist: Getting ready to start GEO

Based on the five basics above, here is a checklist for getting ready to start GEO:

Have you researched the AI platforms your target users use often?
Do you know where your SEO stands today? (It is the foundation for GEO)
Have you defined the metrics (KPIs) for measuring effectiveness?
Do you have a resource plan that assumes at least three months of sustained effort?
Can you explain the need for GEO to management or your team?
Have you defined your keywords (jobs) from a Jobs-to-be-Done perspective?
Are you aware of competitors from outside your industry?

Summary: The right understanding leads to the right first step

When you approach GEO with the right background knowledge, you can get results efficiently and effectively.

In particular, remember to:

Consider multiple AI platforms
Take a new perspective instead of treating GEO as an extension of SEO
Accept that the field is still developing, and go after the first-mover advantage
Measure results with the right metrics
Plan GEO as a medium- to long-term effort

For a more complete picture, also read What is GEO?

Start by understanding where your brand stands today. To see how ChatGPT, Gemini and Claude answer Japanese users' questions about your brand, and which concrete actions to take, contact us.