Your Website Should Capture Leads

A contact form that emails an inbox is not lead capture. It is a slower inbox.

Most business websites are brochures. They say who you are, list what you do, and finish with a contact form. The form emails somebody. That is where the site's job is understood to end.

It is a reasonable model and it leaves most of the value on the table, because the site stops working at precisely the moment a visitor becomes valuable.

A form that emails an inbox is not capture

Here is the test. Without opening an email client, answer three questions about last month:

How many enquiries came in?
What was the median time to first reply?
Which enquiries are still unanswered right now?

If you cannot answer all three, you do not have lead capture. You have a message channel. Email is a delivery mechanism, not a record system: it has no state, no owner, no queue, and no reporting. Anything that arrives there becomes somebody's personal responsibility to remember, and human memory under load is not a system you should be running a sales process on.

Capture means something narrower and more useful. Every enquiry becomes a record with a timestamp, an owner, a status, and a history. It can be counted, sorted, filtered and reported on. Nobody has to remember it, because it is sitting in a queue that visibly still contains it.

That distinction is where most of the practical value sits, and it is almost entirely invisible in a design review.

Own the path, do not rent it

The second thing that separates a working site from a brochure is who owns the mechanism.

There is a widget for everything: reviews, forms, chat, booking, popups. Each costs a monthly fee, each embeds a third-party script, and each holds a piece of your customer relationship on someone else's terms. Individually they are cheap and reasonable. Collectively they mean the site only works while every one of those subscriptions is current, and your data lives in five places you do not control.

Reviews are the clearest case. If your review display is a rented widget, you are paying a recurring fee to show reviews your own customers wrote about your own business. Stop paying and they vanish from your site. The reviews are yours. The display should be too.

When we rebuilt OzLedger's site, self-hosting the reviews was a deliberate call for exactly this reason: owned outright rather than rented, with no recurring fee to display their own data. The same logic drove lead capture, which goes into a dashboard they own rather than a platform they subscribe to.

This is not an argument against ever using a third-party service. It is an argument for knowing which parts of your customer relationship you are renting, and making that a decision rather than an accumulation.

Structure is what makes you findable

The third difference is the one nobody sees and everybody feels.

A site needs a sitemap so search engines know what exists. It needs clean, semantic markup so the content can be parsed rather than guessed at. It needs structured data (schema) so that a machine reading the page knows this is a business, that is an article, those are frequently asked questions. It needs URLs that are stable, because every link anyone ever makes to you depends on them not changing.

None of this is visible to a visitor. All of it determines whether visitors arrive at all.

This has become more consequential, not less, as AI assistants have started answering questions that people used to type into a search box. Those systems read structured data. A business with no schema is not invisible to them, but it is harder to summarise correctly, and the practical result is being described inaccurately rather than not being described. Being wrong is worse than being absent.

The correctness of this layer should also be checked mechanically rather than remembered. On this site, every build runs a check that asserts the sitemap exists and is complete, that every page has a canonical URL, that schema is present and parses, and that no page has drifted into a state where a crawler would see something different from a visitor. That check fails the build rather than emailing a warning, because a warning nobody reads is the same as no check at all.

What to fix first, in order

If you are looking at your own site and wondering where to start, this is the order that usually returns the most:

Make enquiries into records. Anything that produces a time-stamped, owned, queryable entry beats the best-designed form emailing an inbox.
Say what happens next, then do it. The moment after someone submits is the moment they are most likely to be reassured or lost. Silence at that point undoes the work that got them there.
Fix the structural layer. Sitemap, canonicals, schema, stable URLs. Unglamorous, cheap, and it compounds.
Take back anything rented that should be owned. Reviews first, usually.
Then redesign. Not because design does not matter, but because a beautiful brochure is still a brochure, and you will have to touch the design again anyway once the first four are done.

The order is deliberately the opposite of how most website projects are scoped.

Key takeaways

A form that emails an inbox is not captureCapture means a time-stamped, owned, queryable record. Email is none of those things.
Own the lead path end to endRented widgets hold your data on someone else's terms and stop working when you stop paying.
Structure is what makes you findableSitemaps, clean markup and schema are what let search engines and AI assistants read you at all.

Frequently asked

We already have a contact form. Is that not lead capture?

Only if you can answer three questions from it: how many enquiries came in last month, what the median response time was, and which ones are still unanswered. If the answers live in somebody's inbox, you have a message channel rather than a capture system.

Does the design matter at all then?

It matters for credibility, which is a real conversion factor, particularly for professional services where the site is often the first evidence of competence. It is necessary and it is not sufficient, and most businesses over-invest in it relative to what happens after someone decides to get in touch.

Is a website builder good enough?

For a brochure, often yes. The limits show up when you want the site to talk to something else. Builders are designed to keep you inside their ecosystem, so lead data, review display and integrations tend to be available only as paid add-ons you rent indefinitely.

What is the single highest-value change to most business sites?

Making it obvious what happens next after someone gets in touch, and then actually doing that. Most sites go quiet at exactly the moment the visitor has committed, which is the worst possible time to be silent.

How much of this matters for AI search?

More than it used to. Assistants that answer questions about businesses read structured data and clean markup. A site with no schema and no sitemap is harder for them to summarise correctly, which increasingly means being described inaccurately rather than not at all.