Use a Meta instant form when someone can make a useful enquiry from the information in the ad and form. Use your website when they need more explanation, comparison or a working tool before contacting you. Neither route guarantees better leads. Choose the first version around the buying decision, then compare what happens after the enquiry arrives.
This is a different question from whether you need a new landing page. You may already have a suitable website destination. Equally, a campaign may be ready to test a simple callback offer without commissioning another page.
What changes when the form is inside the ad?
An instant form collects details within Meta's advertising experience; a website form takes the visitor to your own page. Meta's lead-generation guide distinguishes these destinations and describes using information already supplied to Meta to help fill form fields. That reduces typing, but a completed field does not tell you whether someone understood the offer. Meta's lead-generation guide.
The practical difference is how much of the decision happens before submission. Write down what the person should know by then: who the service is for, any important limits and what the business will do next. Check whether the proposed destination communicates those points clearly on a phone.
Do not treat either format as a shortcut around an unclear offer. A website can collect poorly informed requests too. An in-app form can set clear expectations when the next step is straightforward.
When I would start with an instant form
An instant form is a reasonable first candidate for a defined callback or introductory conversation. The offer can be explained briefly, a small amount of context is enough to respond, and someone is responsible for the follow-up.
Imagine a service business inviting people to discuss whether it covers their location. If the ad explains the service area and the form makes the callback clear, a separate page may not add a necessary decision. This is a planning example, not a reported campaign result.
Before choosing this route, check three things:
- Expectation: would the person know whether they are requesting a call, a quote, information or a confirmed appointment?
- Handover: can the team see the request and its context in the system it actually uses?
- Capacity: can someone respond within the timeframe promised, including the days when the campaign runs but the usual owner is away?
If these are unresolved, changing the form format will not settle them. A smaller initial campaign with a working response process is a more useful test than generating enquiries nobody reviews.
When the website earns the extra step
I would favour a website destination when the visitor needs to compare options, read substantial conditions, inspect examples or use a tool before submitting. Availability selection, a tailored estimate or a detailed product explanation can make that extra step useful.
My SpainProperty work includes area guides, property discovery and consultation booking. A buyer comparing locations has a different immediate task from someone who already knows what they want to discuss. The website gives the first person room to research. A short consultation request may suit the second. These are possible journeys based on the product's actual scope, not evidence of an instant-form comparison I have run.
The website still has to work. Open the intended page on a phone, read it without the homepage for context, submit the form and check the received request. A slow or broken destination is a repair task, not proof that website forms are the wrong format.
If the existing service page already supports the offer, improve it before ordering a duplicate. The separate guide to campaign landing pages versus existing pages covers that decision.
Separate receiving leads from sending outcomes back
There are two different connections to specify. One brings the enquiry into your working system. The other sends later business outcomes back to the advertising platform. Meta lists CRM lead retrieval and CRM integration through the Conversions API separately in its lead-generation guide. Do not accept “the CRM is connected” as proof that both directions work.
Ask to see a test request with its source, submission time, answers and assigned owner. Then check a repeat request and a changed form. Decide which information must remain attached to the enquiry so that a colleague can continue the conversation.
For a concrete integration example, HubSpot lets businesses map custom lead-ad fields to CRM properties. Its documentation says mapping changes apply to newly synced leads, not previously synced contacts. A setting change therefore does not establish that earlier records have been corrected. HubSpot's lead-sync and field-mapping documentation.
HubSpot also documents a separate setup for sending lifecycle changes or supported form events to Meta through the Conversions API. Its form-event route is limited to forms created in HubSpot; it is not a promise that every external website form works through that feature. Check the actual integration and subscription before commissioning it. HubSpot's Meta conversion-event documentation.
These are examples of product-specific boundaries, not a recommendation to replace your CRM with HubSpot. In my own MarciCRM work, the relevant principle is keeping property enquiries, conversations, appointments and follow-up together. That portfolio example does not establish that a particular Meta integration is implemented there.
Compare the same offer through both routes
If both destinations are viable, write the comparison plan before launching. Keep the offer and qualification rule consistent. Record differences in audience, creative, budget, delivery period and follow-up, because they can explain an apparent winner.
The question should be specific: “Which route produces more suitable consultation requests at an acceptable cost and workload?” Comparing a downloadable guide with a sales appointment would answer a different question, even if both campaigns report leads.
For each route, keep a simple review sheet:
- Spend and received requests: how much was spent, and how many distinct enquiries reached the business?
- Review status: how many were assessed, and how many are still waiting?
- Suitability: how many met the agreed criteria, with a reason recorded for those that did not?
- Progress: how many reached the next meaningful stage, such as a conversation held or a proposal requested?
- Workload: how much time went into contacting, clarifying and processing those enquiries?
Calculate advertising cost per suitable enquiry by dividing the relevant spend by the number that met the agreed criteria. Keep setup and operating costs visible separately. If none qualified, report that directly; there is no meaningful finite cost-per-qualified-enquiry figure to display.
Give both groups the same time to progress. Keep “not reached yet” separate from “unsuitable”, and identify people who submitted through both routes. Do not add form submissions and later appointments together as if they were separate acquired customers.
Two campaigns running alongside each other are not automatically a controlled experiment. Delivery and audiences may differ. Use the comparison to guide the next decision, and describe its limits. A handful of early enquiries does not establish a reliable winner.
Match the next investment to the missing step
An initial review should end with a concrete scope. If people misunderstand the offer, revise the ad and destination copy. If enquiries arrive but lose their answers or owner, fix the handover. If the business cannot tell which requests are useful, agree the review process before paying for more sophisticated optimisation.
Commission a website change when information or functionality is missing from the customer journey. Commission an integration when the handover is the problem. Keep a functioning instant-form route if it supports useful conversations at an acceptable cost. There is no requirement to replace it just because a custom page looks more substantial.
For an enquiry-form review, send the offer, the current ad destination and an anonymised description of what happens to incoming requests. Avoid sending customer names or contact details in the initial brief. I can review the advertising, website and follow-up together. Contact me through the website or at hello@davidbuka.com.
Primary references checked on 30 October 2026. The cover is an original editorial illustration, not a photograph of a client campaign or its results.