Your website gets visits. Where do potential customers get stuck?

1. Check whether the visit matches the job

Choose one service you want more inquiries about. Compare the ad, search listing or referral link with the page it opens. Does the page explain the same service, for the same customer and service area?

Illustrative example, not a client result: A commercial cleaning company promotes office cleaning, but its link opens a homepage dominated by house-cleaning photos. An office manager must guess whether commercial work is available. A dedicated office-cleaning section could answer that question before they contact anyone.

Record the mismatch you can actually see. If you have search-query or campaign data, use it to check the assumption. If you do not, mark traffic relevance as unknown. A quiet inbox does not prove that the website is the problem.

2. Answer the questions that determine fit

On the landing page, look for:

• The specific work offered and any important exclusions.
• The locations or customer types served.
• Evidence relevant to that work, such as a clearly described project or a genuine review used with permission.
• What happens after someone gets in touch.

“Quality service you can trust” gives a visitor little to act on. A concrete description of the job, area and process is more useful. Publish only details you can stand behind. Do not add unsupported response-time promises, credentials or availability.

3. Choose a next step that fits the request

The main action should tell people what they are starting. “Ask about office cleaning” is more specific than “Submit.” A call may suit a time-sensitive request when someone answers the phone. A booking link may suit a service with defined appointments. A message or short request form may suit work that needs an initial description.

Offer a sensible alternative without making every option compete for attention. Ask someone unfamiliar with the website which action they would choose and what they expect to happen. If they hesitate, note their question before changing the design.

4. Try the path on a small screen and with a keyboard

Check whether menus, sticky banners or chat controls cover the action. Enlarge the text and look for clipped instructions. Use Tab and Shift+Tab to move through links and fields; make sure the focused control is visible and usable.

For forms, use clear labels, explain required information and show errors that help people recover. Ask only for what is needed at this stage. These checks follow W3C’s accessible-forms guidance; a complete accessibility assessment goes further.

Illustrative example: A garden-maintenance business asks for an exact property measurement before allowing any inquiry. If the team can first establish the area and type of work, move the precise measurement to a later step. Keep it required if it is genuinely necessary to assess the request.

If a page feels slow, record the device, connection and point where you had to wait. Ask the person maintaining the site to measure that path before deciding what to replace.

5. Separate an attempted contact from a real inquiry

A phone-button click does not prove a connected call. Opening a messaging app does not prove a message was sent. Compare the interaction you can measure with the inquiries your team actually receives, without copying personal information into analytics.

If you use GA4, its enhanced measurement can record form starts and submissions when configured. Those signals still need testing against your actual form and receiving system. Google's documentation also requires keeping personally identifiable information out of analytics. See enhanced measurement events.

Arrange a clearly labeled test with your own team. Confirm that the request arrived and that someone knows who handles it. If delivery is the failure, fix that before rewriting the page.

Pick the next repair

Use the worksheet below. Mark each check works, blocked or unknown, and attach a concrete observation.

Fix a blocked contact path first. Next, address a clear mismatch between the visitor’s task and the page. Then test clarity and ease of use. If the evidence is missing, choose an observation to collect rather than declaring a cause.

Write one sentence: “We saw ___ on ___, so we will change ___ and repeat ___.” Compare the same path afterward. When reviewing business results, note changes in traffic sources, seasonality or availability; a before-and-after difference alone does not establish what caused it.

Start with the customer task. Finish with one verified improvement.