Website visits do not always turn into enquiries. Some people are looking for information, others are comparing offers, and some may encounter unclear terms or a technical error.
To find the cause, follow the journey from the traffic source to a received enquiry and a response from your team. At each stage, check a specific action and its outcome.
Below is a diagnostic sequence, an illustrative funnel and a table to help you decide which issue to fix first.
Compare website data with your inbox, CRM, enquiry log and messages in other channels. A visitor may contact you through a messaging app or by phone without using the form.
Then submit a test enquiry from both a phone and a computer. Check the entire journey:
A “Thank you” message on the page does not confirm that the enquiry reached your team. Check the record in the receiving system and notification delivery separately.
W3C recommends informing users when an action has been completed successfully and explaining how to correct form errors. W3C guidance on form notifications.
If the test enquiry is lost, find and fix that error first. Until it is resolved, it will be difficult to judge how well the copy works based on the number of enquiries received.
Track at least four separate actions: a visit to the service page, a move to the form, the start of form completion and a confirmed submission. You can compare receipt of the enquiry and its subsequent status with records in your CRM or another log.
A button click indicates an intention to take action. It is not a form submission. A browser-side form event does not, by itself, confirm that a member of your team received the enquiry.
GA4 provides the recommended event generate_lead to record the generation of a lead, for example through a form. Configure it to match the actual action on your website. Google Analytics event documentation.
Check that the event does not fire before the intended action is successfully completed and is not sent multiple times because of repeated clicks or a page refresh.
Use the same reporting period and clearly defined units of measurement. Do not divide enquiries by advertising clicks and then compare the result with a rate calculated using users or sessions.
If you receive real enquiries but analytics shows zero, fixing measurement is the priority. If analytics shows submissions but the receiving system has no enquiries, check both the event configuration and form processing.
The total number of visits can conceal different intentions. Someone reading a reference article is in a different situation from someone arriving at a service page through an advert.
Compare data by source, landing page, device, language and market. For paid advertising, also review the available search terms, audiences and ad content.
Ask three questions:
For example, an advert for online store development should lead to a page that clearly explains the terms of that particular service. If visitors see a general list of all your services, they have to continue searching within the website.
Google Ads recommends aligning your landing page with the advert and its call to action. Google Ads guidance for landing pages.
When visitors arrive with needs you do not serve, review the traffic source and how the offer is worded. Changing a button colour will not resolve a mismatch between their request and your service.
Visitors should understand the problem you solve, what the work includes, how pricing is determined and what happens after they contact you.
Ask someone unfamiliar with your company to read the service page and explain the offer in their own words. If they misunderstand the scope, terms or outcome, clarify the relevant sections.
Also check whether visitors can assess your suitability: view relevant projects, understand your role in them, and find contact details and company information. General statements about quality need specific explanations and evidence.
A “from” price should be accompanied by a description of the basic scope. If you quote individually, explain what information you need to prepare an estimate. Show important limitations and additional costs before the form.
If the page already contains this information, check how noticeable it is on a phone. The details may be present but lost among long paragraphs or hidden under an unclear section heading.
Open the page as a visitor would: through an advertising link, from search or inside an app’s browser. Where possible, use the devices and browsers your audience uses.
Check how the important content loads and whether the menu, buttons and form work. Look for overlapping elements, buttons shifting during loading and an on-screen keyboard that makes completion difficult.
Core Web Vitals can help assess technical performance: LCP describes the loading of the main content, INP measures responsiveness to interactions, and CLS measures visual layout stability. These metrics help identify problems, but do not explain the absence of enquiries on their own. Core Web Vitals overview.
A single high tool score does not replace checking the user journey. A form can fail to deliver enquiries even on a website that loads quickly.
Explain beside the form what visitors should send and what will happen next. A “Get a quote” button should lead to a clear process for preparing that quote.
Review which fields are essential for the first conversation. If you request a detailed technical brief, consider whether the client can prepare it without an initial consultation.
Check error messages, phone number format requirements, submission from different devices and an available alternative way to contact you.
Then assess how enquiries are handled: who responds, where the status is recorded and how failed contact attempts are dealt with. If enquiries arrive but go unanswered or are unsuitable for the service, investigate the reason during follow-up.
Imagine a service page with the following data for one reporting period. All figures are illustrative; this is not a client case study or a benchmark for a normal conversion rate.
At each stage, count sessions in which the relevant action took place, rather than the total number of events. For this example, assume no more than one enquiry per session and that the stages can be matched.
The share of sessions resulting in a received enquiry is 4 / 1,000 × 100% = 0.4%. Among sessions where form completion began, an enquiry was received in 4 / 40 × 100% = 10% of cases.
These figures identify a stage to investigate. They do not prove that the form is difficult to use: possible causes include a submission error, required fields, a visitor changing their mind or inaccurate measurement.
The practical next step is to check events and delivery, complete the form on the main devices and examine specific difficulties. You can then define a fix and assess its result.
If few people click the button and you have confirmed it works, examine traffic quality, the offer and the visibility of the next step. Be cautious about drawing conclusions from a small number of visits: a few actions can noticeably change the calculated percentage.
Record the observation, the suspected cause and the way to check it separately. This helps avoid presenting an assumption as an established problem.
First, fix confirmed errors that prevent contact or distort the data. Then investigate the most significant issues in the relevant traffic sources and on the main devices.
For a systematic review, you can start with our website audit. For smaller fixes on Tilda, WordPress and Webflow, we offer website and design support billed by the hour. We agree on the scope and time estimate based on the issues identified.
Before making a fix, record the problem, date, baseline data and success criterion. For example: “A test enquiry submitted from a phone is not saved in the system. After the fix, it should be saved and delivered to the responsible team member.”
Verify a technical fix by repeating the same scenario. For changes to copy or the form, monitor confirmed enquiries and their quality, taking visit numbers and traffic sources into account.
Changes to an advertising campaign, the season or the audience mix can affect the result. Comparing two periods does not, by itself, prove that a particular website change caused an increase or decrease.
Keep a record of hypotheses and results. A form malfunction can be fixed straight away; assessing a new offer or a different presentation of content may require more observations.
Finding the weak point starts with a journey you can check: a relevant visitor → a clear offer → a working contact process → an enquiry that is received and handled. Choose fixes based on confirmed problems, and assess the result using real enquiries and their quality.