When manual input makes more sense than the homepage

Use the website for a clear single offer. Describe the business yourself when an umbrella homepage hides the focus.

Two people review a page on a laptop together
Two people review a page on a laptop together Kaboompics · Pexels License · Resized for display
In this article

A homepage can introduce a business accurately and still be a poor starting point for a particular AI visibility check. A regional company might list design, installation, repairs and commercial services on the same page. A visitor can click through to the work they need. The check has to turn the original information into one service and market before asking its buyer questions. The right input depends on whether the business information identifies that focus.

Use the site when it clearly describes one service and the audience or market you intend to test. Prepare manual input when the broader company description obscures that service. In either case, verify the proposed scope before proceeding. An accurate input helps you make sense of the questions and inspect the company's own public information after you see the results.

Read the homepage as a prospective customer would

Start at the top and identify what a first-time visitor can reliably learn. Can they find the specific service, who it serves and the area it covers? Do those facts appear early, or does it take visiting several pages to see them? Give yourself a reasonable answer based on the page rather than assuming a tool will select your preferred service from a long list. Your company's broad brand description may be clear on its own terms.

Check what the business actually offers with the staff responsible for inquiries. If the site describes a service in one metro area and the booking team no longer accepts new projects there, correct your own understanding before running a check. If a site combines household and business work, ask which one you intend to measure. A clear service name without an intended audience can still produce questions from several different kinds of buyers.

Use the publicly accessible page as the source when suitable. A small firm that offers one service throughout a defined area may explain enough on the homepage. If your website directs people to a more specific public page for the offer, read that page as part of your preparation. Check any result from importing the website before deciding what material the test actually captured.

Decide when a manual description will be clearer

Manual input can help when a company serves several markets, uses an umbrella brand, or presents several offers at the same level. It can also focus on an offer with a separate buyer process that's described later on the site. Start by saying what the company does and where, using ordinary terms a staff member can explain. Include the intended audience or scope only where that distinction matters to the purchase.

Suppose a home services company offers kitchen updates, broad remodeling and accessibility improvements. It wants to inspect how people looking for a bathroom remodel might encounter local providers. A manual description of its actual bathroom work and service area can give you a specific sample to review. Avoid adding work the company does not perform or suggesting a geographic area merely because it would make the test more interesting.

Check whether your description leads to a question a customer can answer from public information. If the manual scope offers much more detail than you show on the site, ask why. The page may need additional accurate information or the description may include a claim that depends on an unusual request. Write down which source backs each important point. This keeps manual input connected to the real business.

Confirm the service and market before the checker runs

Mantas Auk created the free AI Visibility Check. The tool accepts a website or a manual business description and lets you confirm the service and market. It then asks Claude 25 buyer questions with live web search, without inserting your brand name in those questions. The report includes a mention count, other names found, and suggested next steps. Treat the check as a sample based on the details you provide.

Read the proposed scope in plain language. If the tool suggests a market or service that does not fit, correct the scope before proceeding or choose manual input. A participating corporate site may have a long list of markets, while the branch you want to inspect handles only one. You can save the input with the outcome. When someone later asks why a particular question appeared, the chosen scope will still be available.

Anthropic’s web search support material describes live search and source references in Claude. The checker uses Claude with web search to answer a defined question set. A result describes those answers. Sources and wording can change across checks, so keep the original input, questions and run date if you intend to compare. Claude web search documentation

Compare the questions with real customer inquiries

Choose a few questions from the report and ask someone who handles inquiries whether a customer would realistically ask them. Look for service type, location, buyer needs and expected next action. A question about an occasional specialty job might be interesting even if few visitors request it. A question assuming emergency availability might be outside the service you intended. The usefulness of a mention count depends in part on whether the underlying questions match.

Mark questions that reveal a gap you can verify. The site might describe what the company does and fail to explain how a customer requests a quote. A service page could use a broad phrase that can mean two different offerings. Confirm the appropriate facts with staff and look at what information a visitor can find. Keep marketing descriptions tied to what the business can substantiate.

Google’s Search Central guidelines discuss helpful, reliable information that serves a real audience. Use them for the question of whether prospects can use the source page. They do not describe how Claude decides which business to mention. A clearer service name, verified area and practical first step are valuable details when the customer actually needs them. Google Search Central on helpful, people-first content

Compare two methods without mixing their results

If you test the homepage and a manual description, save the inputs and questions separately. The runs may not be directly comparable if the content or market differs. Look at which questions each method produced and whether the scope matches your purpose. A manual run could give better questions while an earlier homepage run counted more mentions. Inspect the individual cases instead of selecting a method because of a single number.

For example, a consultant may want one report focused on work for small organizations and another for a separate specialty. A homepage run could retrieve an overview of both. A manual description can let the consultant specify one. The comparison shows whether the questions address the chosen service. The result does not establish that the company should remove its other services from the site or expect a later answer to use its business name.

Build the manual input from information the business already uses

Start with the service phrasing on the most relevant public page and the terms staff use with new customers. Write one plain sentence that describes the actual job and market. Add a second sentence only if it identifies a material distinction, such as whether the work is for commercial property or applies to a defined service area. An adjective that makes a claim about quality will not resolve which service a customer can request.

Read the description back to someone who can verify it. If the company offers many kinds of the work, ask which one its team expects to see in the test. If another branch handles the request, put the correct market or location in the recorded scope. The person checking it should be able to identify the real service the test will ask about and the approved route a customer could use.

Compare which part of the real page could support the proposed service

If the homepage starts with company-wide information, follow its normal links to the relevant service before deciding whether to improve the homepage itself. A visitor might already find a direct route and accurate answer. In that case, the manual input can serve the purpose of a focused test. If multiple links use the same broad words, find out where customers are likely to hesitate and which page the company expects them to use.

Keep the service page and business description consistent where they refer to the same facts. For example, a site may describe certain work broadly while its scheduling staff only takes larger jobs in a limited area. Verify what the business wants to tell prospective customers before editing. An accurate description can leave room for exceptions without implying that a service is available at every address.

Save one example that explains the input choice

Record a representative question from the run and the part of the business description it reflects. If someone compares the results later, they should be able to see which method produced that question and what you wanted to inspect. Keep dates and the market along with the source. This makes the comparison useful even if you retire one version of the homepage or change what the company offers.

Keep the next action close to the real service

Pick one question the relevant buyer could ask and that you can answer with reliable information. Confirm the facts and decide where a visitor would expect to find them. Save the original result and date if you intend to run the checker again. Use the actual scope and questions to interpret a later result. A useful adjustment gives the business a clearer public explanation, even if mention counts vary.

Use the Preston Hollow Today tools section to find related digital guides.

Sources

Share by email