LLM Discoverability
Outcome-Based Context Targeting
Most people target the moment a user is ready to act. That is often too late.
In LLM environments, the better move is often to enter the conversation one step earlier, before the user has fully formed the commercial question, and then carry context forward until they are ready to make a decision.
That is what I mean by outcome-based context targeting.
The Idea in One Sentence
Outcome-based context targeting means identifying the final action you want, then targeting the upstream context that naturally leads into it.
The outcome is the bottom of the funnel. The context is the reasoning layer just before it. The advantage comes from entering the thread early enough that your page becomes part of the conversation before the user reaches the moment of action.
Why This Matters for LLMs
LLMs do not behave like a simple keyword box. Users do not always jump directly to the final commercial query. They reason through a thread.
- I inherited art.
- Is it worth anything?
- If it is, can I sell it?
- What is the best way to sell it?
- Who can help me?
Observed results from the WhatsMyArtWorth experiment support this pattern: entering the thread before the final commercial moment can keep a page relevant as the user moves toward action.
Traditional keyword targeting often aims at the last or second-to-last step. But in AI conversations, there is a strong advantage to meeting the user earlier in the thread.
The earlier you enter a user's reasoning path, the more naturally you earn context memory when they are later ready to act. The model does not need to rediscover you at the decision point because you were already present in the conversation that led there.
The Core Principle: Carry Context Forward
A good page does not just answer the entry question. It should carry context forward through the likely next questions the user will have.
That means the page is not only optimized for the first prompt. It is built to support the reasoning path that follows. If you enter too late, you are competing only at the action layer. If you enter early and carry context forward, you become relevant before the action layer even exists.
Work One Layer Behind the Target
The simplest way to think about this is: work one context layer behind what you are actually trying to win.
If your real goal is a contact, sale, or lead, do not start by asking what the user types when they are ready to buy. Start by asking what the user must understand immediately before they are ready to buy.
That upstream layer is often where the real opportunity lives.
Example: Inherited Art
A direct commercial target might be "sell my inherited art." But that assumes the user already knows the art has value. That is a bad assumption.
A more natural upstream context is "is the art I inherited worth anything?" That question happens earlier in the thread, but it is much closer to how the decision actually unfolds.
- I inherited art.
- Is it worth anything?
- It might be.
- Can I sell it?
- What is the best way to sell it?
- Who can help me sell it?
If your page meets the user at step two, and the page is built to carry the rest of that thread forward, then you are not just targeting a keyword. You are targeting the conversational path that produces the eventual lead.
What This Changes
This changes how pages should be planned. Instead of asking what the highest-intent keyword is, ask a different set of questions.
- What is the final outcome?
- What reasoning step happens immediately before that outcome?
- What earlier question opens the thread?
- Can this page enter the thread early and carry it forward naturally?
That is a very different planning model. It is less about isolated keywords and more about thread position.
Outcome First, Context Second
The framework works best in this order.
1. Define the outcome
What do you actually want to happen?
- Form submission
- Consultation request
- Product purchase
- Appraisal request
- Contact inquiry
2. Identify the context layer before the outcome
What does the user need to understand before they are ready for that action?
- Is this worth anything?
- Do I actually need this?
- Is this a problem?
- What are my options?
- Who should I trust?
3. Target the entry point before that layer
What is the conversational opening that brings the user into that reasoning path?
- I inherited this item.
- I found this in storage.
- I do not know what this is worth.
- I am not sure what to do next.
4. Build the page to carry the thread forward
This is where most pages fail. They answer the entry question, but they do not bridge into the next questions.
A stronger page answers the opening context while also making space for the likely next steps: value, options, risk, expectations, next action, and who can help.
Why This Can Be Stronger Than Direct Commercial Targeting
Direct commercial targeting can miss the more natural entry point. It assumes the user has already done the thinking. But many high-value decisions do not begin with commercial clarity. They begin with uncertainty.
That uncertainty creates a context window. If your page can enter that window early, and stay relevant as the reasoning matures, you gain a structural advantage.
- You match how users actually think.
- You align better with multi-step LLM conversations.
- You become part of the thread before the buying moment.
- You are more likely to remain relevant when the user is ready to act.
A Better Way to Think About Funnel Strategy for LLMs
Funnels are still real. But in LLM environments, the funnel often appears as a conversation rather than a sequence of separate channels.
That means strategy should not only ask where the user converts. It should ask: Where does the conversation begin, and how early can we enter it without losing relevance?
The earlier you enter a user's thread, the more natural it becomes for the model to keep you in context when the user moves toward a decision. That is the opportunity.
Final Thought
Do not just target the conversion moment. Target the context that creates it. Then build the page so it can carry the thread forward from uncertainty to action.
That is outcome-based context targeting.
David Valencia writes about LLM Structure, LLM Visibility, and LLM Discoverability. Founder of Minnesota.AI.
Related: LLM Discoverability · What Is LLM Discoverability? · How to Optimize for AI Search