A prospect is someone who fits your customer profile but has not identified themselves to you. A property manager in Worcester who could use your service is a prospect whether or not they have ever heard your name. Prospects exist in the market. You can estimate them, target them with advertising, and build pages for the questions they search, but you cannot list them by name, because they have not told you who they are.

A lead is a prospect who has identified themselves. They submitted a form, called the office, replied to an email, or started a chat. The defining event is the handoff of contact information paired with some signal of interest. Before that moment you have an audience. After it you have a name, a way to follow up, and a record you can track to a closed sale or a dead end.

The distinction matters because the two groups answer different questions. Prospects tell you how big the opportunity is. Leads tell you how well your website and marketing convert that opportunity into conversations. When a team uses one word for both, those two questions collapse into a single blurry number that answers neither. So what is a lead, in the shortest usable form? A person who fits, who has raised their hand, and whose hand-raise you recorded.

Consider a hypothetical services firm whose site drew 2,000 visitors last month. If everything counts as a lead, the report says 2,000 and the number is useless: most visitors were competitors, job seekers, existing customers, or otherwise the kind of traffic a form built to filter unqualified leads would never pass through. If only sales conversations count, the report might say six, and the website looks like it did nothing, even though it produced every one of those conversations.

Counted separately, the same month reads differently. Two thousand visitors, of whom some fraction were genuine prospects. Twenty-five form submissions and calls: those are the leads. Six of them turned into qualified conversations. Now the business can see three distinct performance questions in one view: whether the site attracts the right visitors, whether it converts them into leads, and whether the sales process converts leads into revenue. When one of those numbers drops, the report shows which stage moved, and the fix lands in the right place. A traffic problem gets a visibility fix. A conversion problem gets a page or form fix. A close-rate problem gets a sales process fix. Businesses that conflate the stages tend to fix the wrong one, because the blended number cannot say where the loss happened.

This is the practical core of sales funnel stages: not a diagram for the conference room wall, but a set of boundaries that make each number mean one thing.

Measurement starts at the moment a prospect identifies themselves, because that is the first event your systems can record about a specific person. Everything before it is aggregate: traffic, impressions, rankings. Those numbers are worth watching, but they describe crowds, not people, and no follow-up action attaches to them.

On most business websites that first measurable moment takes one of a few forms: a contact or quote form submission, a phone call, an email to a published address, or a chat session. Each is a conversion point, and together they define the boundary between prospect and lead for your business. A useful early exercise is simply listing them. Many owners are surprised by the inventory: forms added by a previous developer that route to an unmonitored inbox, a phone number in the footer that nobody attributes, a quote request PDF that gets printed and never logged. Every unrecorded conversion point is a lead the report never sees, which is one of the quieter ways a website’s flaws cost it leads, and it means the report undercounts the website’s contribution and overcounts everything else’s.

Website lead measurement does not require a marketing department. A minimum setup covers three things.

Form submissions come first because they are the easiest to capture. Each form on the site should do two things on submit: deliver the inquiry to a monitored inbox and record the submission as an event in your analytics. Modern analytics tools track form completions as conversion events; once that is configured, every submission is counted automatically rather than reconstructed from memory at month end.

Calls are the leak in most small-business counts. When the number on the website is the same line the office publishes everywhere, there is no way to know which calls the site produced. Two workable approaches exist. Call tracking services display a forwarding number to website visitors and log each call with its source. The lower-tech alternative is a front-desk habit: asking callers how they found the company and tallying the answers. The tally is imperfect, but an imperfect count beats an absent one.

Traffic source is the third piece. A lead without a source tells you the site works; a lead with a source tells you which effort earned it. Analytics platforms attribute each session to a channel (organic search, paid ads, referral, direct), and when form events are configured, that attribution carries through to the lead. Over a few months the pattern typically becomes plain: which channels produce leads, and which produce only traffic.

That is the whole starting kit. Lead tracking can grow into attribution modeling and dashboards later; it cannot grow out of records that were never kept.

A CRM formalizes the boundary you have already drawn. The common mistake is the opposite order: adopting a CRM first, then trying to satisfy its default pipeline of half a dozen stages with names nobody uses in conversation. Records rot in stages that do not match how the business actually sells, and within a quarter the team stops updating them.

The workable pattern is to mirror reality. If your process is “inquiry comes in, we qualify it, we quote it, we win or lose it,” the CRM needs roughly those stages and no more. New leads enter at the first stage automatically when the form connects to the CRM, or by a five-minute daily habit of logging calls. The prospect-versus-lead line maps cleanly: prospects are not in the CRM at all (they are the audience your marketing addresses), and every CRM record began as a recorded hand-raise. When a stage exists that no one can define by an observable event, it usually gets deleted within a few months anyway. Starting without it saves the cleanup.

Once prospects and leads are counted separately, the monthly review changes shape. Instead of one number that provokes an argument, there is a short chain: visitors, leads by source, and qualified conversations. Each link has a rate, and the rates hold reasonably steady month to month, which is what makes movement meaningful. A drop in leads with steady traffic points at the site or an offer. A drop in traffic with a steady conversion rate points at visibility. Neither diagnosis was available when the numbers were blended.

The chain also gives marketing spend a denominator. Cost per lead by channel is a directly computable figure once leads carry sources, and it is the number that makes budget conversations short. None of this requires more data than the minimum setup above produces. It requires the distinction, applied consistently, starting at the moment someone raises a hand on your website.

Most businesses cannot fix this from a blog post alone, because the gaps are specific to their site: which forms fire events, whether calls are attributed, where the CRM boundary sits. Ask us for a measurement review of your site’s forms, calls, and analytics, and you will get back a plain accounting of what is being recorded today, what is leaking, and the shortest path to a lead count you can trust.