01 / Listen
Understand your offer, customers and commercial goals. Identify the questions that matter and establish how your business currently appears.
← Home
Our approach
A clear question. A better next move.
A focused cycle that connects discovery, implementation and learning.
02 / Prioritise
Choose the changes most likely to resolve a real gap. Agree on the scope, ownership and evidence that will guide the work.
03 / Build
Improve accessible content, consistent business information and supporting sources. Prepare paid tests where they fit the opportunity.
04 / Learn
Review visibility alongside traffic and enquiries. Keep what helps, investigate what does not, and update the plan.
What you bring to the first conversation
A clear description of your offer, your priority customers and a few questions your sales team hears repeatedly are enough to start. Existing website reports, customer feedback and examples of inaccurate AI answers can help identify where to investigate.
What a working plan should contain
Each priority needs a reason, an owner and a way to review progress. A plan can combine content updates, business-information corrections and a paid test, but the scope should reflect your team’s capacity. Agree on the next review before implementation begins.
A closer look
From a question to a working plan.
Each stage should leave the next person with enough context to move the work forward.
01
Define the decision before collecting more information
Start by choosing an offer, audience and business outcome. “Improve AI visibility” is a direction; “understand why our commercial installation service is missing from relevant provider comparisons” is a question that can guide research. The narrower question helps decide which pages, answers and business facts need attention.
Agree what is already known and where the uncertainty lies. Existing customer feedback, service documents and website information can provide a starting point. Avoid collecting information simply because it is available. Each input should help clarify the decision or reveal a gap that affects it.
02
Build a backlog people can actually use
A useful task explains the problem, the intended change and the evidence behind it. Instead of “improve the service page,” specify the missing information and where it belongs. Include any dependency, such as approval from a product specialist or an update to a public listing.
Choose a manageable first group of tasks. Separate urgent factual corrections from longer-term content work and optional experiments. This lets a team make progress without waiting for every possible investigation to finish. Review the sequence when new evidence changes the relative importance of a task.
- Problem: what is unclear, missing or incorrect?
- Change: what should be different when the task is complete?
- Review: who checks it, and against which evidence?
03
A sample first cycle
For a hypothetical local service business, the first cycle might focus on one service area. The team checks how the offer is described, confirms coverage with the business owner and improves the relevant page. Related listings are reviewed for contradictions. The research record is then revisited to decide what to investigate next.
This example describes a sequence, not a promised delivery schedule or result. Real scope depends on the starting information, approvals and work required. The useful principle is that each action has a reason and each review can refer back to what changed.
04
Keep the learning when the work moves on
Store the reasoning with the deliverable. A future editor should be able to see why a passage was added, which evidence supports a claim and who can confirm that it remains current. This prevents the next redesign or campaign from accidentally removing an important clarification.
At each review, distinguish completed work from demonstrated outcomes. A page can be updated successfully before there is enough evidence to assess its effect on discovery. Record both honestly, then choose whether to continue observing, revise the approach or address another priority.
Good to know
A few good questions.
The details that help you decide on your next step.
What happens when my business starts working with NapoleonGPT?+
We begin with your offer, priority customers and growth goals, then review how your business currently appears in relevant AI answers. That baseline informs a focused plan with agreed deliverables, responsibilities and review milestones.
Can my business begin with one service or location?+
Yes. We can focus the initial work on a specific service or market. This gives your business a manageable starting scope and makes the before-and-after comparison more relevant to the opportunity you want to pursue.
What involvement will NapoleonGPT need from my team?+
We need accurate service details, relevant website access and a person who can approve business facts and proposed changes. NapoleonGPT organises the research and agreed implementation; your team supplies the knowledge and approvals that only your business can provide.
How will NapoleonGPT keep my business informed?+
We report what has been completed, what the visibility review shows and which action we recommend next. The comparison uses a documented baseline so your business can distinguish completed work from observed changes in AI answers or enquiries.
Does my business have to buy advertising as well?+
No. Organic AI visibility can be a complete project on its own. We propose paid advertising only when it fits your goals, available access and budget, with a separate scope and media-spend decision.
When can my business expect the work to be reviewed?+
Your proposal sets delivery and review milestones around the agreed scope, access and approvals, including when the final measurement is taken. Changes to answers on external AI platforms remain variable.
The next conversation starts here