The short answer
A lead capture audit is a self-test: send an inquiry through each of your own channels, the website form, the phone after hours, email, and social DMs, then track where each one lands, who gets notified, and how long a real reply takes. Four test inquiries and a notepad show you exactly what new customers experience.
You know your business from the inside: the jobs, the crew, the calendar. What you almost never see is the outside view, the one every new customer gets, because you have never had to inquire with your own company. The audit in this post fixes that in about five minutes of doing and a day of watching. You send yourself four inquiries, one per channel, and follow each one like a package: where it lands, who is told, what happens next, and when. No dashboard required, and the results tend to surprise even owners who are sure their intake is fine. It is worth knowing what today's customers compare you against: in Zendesk's CX Trends research, 72 percent of customers say they want immediate service, and over half will switch to a competitor after a single unsatisfactory experience. The first inquiry is the experience.
Why test your own lead capture?
Because intake problems are invisible from the inside. A broken form throws no error you would see. An inbox rule quietly files inquiries under a folder nobody opens. The after-hours voicemail box filled up in March. Each channel was set up once, by someone, and has been assumed to work ever since. Meanwhile every inquiry that goes astray does so silently: the customer does not call to report that your form never answered them. They just book with the next company on their list.
The audit works because it converts assumptions into observations. You are not asking "do we respond quickly?", a question everyone answers optimistically. You are watching one specific inquiry with a timestamp on it, and the timestamp does not negotiate.
What are the four tests?
Run each test as a plausible customer, not as yourself. Use a personal email, a family member's phone, a real-sounding project. Note the time you send each one.
Test 1: Submit your own website form
Fill in your contact form with a normal request and watch three clocks. Did a confirmation of any kind arrive, and did it say anything useful? Where did the inquiry land internally, a shared inbox, one person's email, a CRM, and who was notified? And when did a real, human-or-equivalent reply arrive that moved the conversation forward? While you are there, do the form itself honestly: count its fields, try it on a phone, and notice whether anything about it made you hesitate.
Test 2: Call your own number after hours
Call at 8pm or on a Sunday, when a meaningful share of real inquiries arrive. Listen to what a customer hears: how many rings, what the voicemail greeting actually says, whether the mailbox has room. Leave a message with a callback number, then time the callback. If the honest answer is that after-hours callers get a full mailbox or a greeting recorded years ago, you have found the channel where inquiries end quietly; it is the exact gap an AI receptionist exists to close.
Test 3: Send an email inquiry
Email the address on your website, the one on your Google listing, and any info@ alias, with a short project inquiry. The question is not whether the address works but who owns it: which inbox it feeds, whether that inbox is anyone's named responsibility, and how long a message sits there before someone reads it. Shared inboxes fail exactly the way shared responsibility always fails, a pattern we unpack in Who Owns the Lead?
Test 4: Send a DM
Message your business on whichever platforms you maintain a presence: Instagram, Facebook, Google Business Profile chat. Social DMs are where intake assumptions go furthest unchecked, because the notifications go to whoever set up the page, possibly on a personal phone, possibly years ago. If your ads run on these platforms, this test matters double, since the inquiry arrives on the same platform as the ad.
How do you score the results?
For each channel, grade three things you observed:
| What you measure | Pass | Needs work |
|---|---|---|
| Landing spot | Named system or person, visibly logged | Shared inbox, personal phone, or unknown |
| First response | Substantive reply the customer could act on | Silence, or a bare autoreply with nothing useful |
| Time to human follow-up | Same business day | Next day or later, or never |
A channel that fails the landing-spot test will eventually fail the other two, whatever this week's result, because an inquiry without a defined landing spot is captured by luck. Most businesses that run the audit find at least one channel where the honest grade is "never found out where it went," and that discovery is the entire value of the exercise.
What should you fix first?
Fix in order of volume, not embarrassment. The channel where most of your inquiries actually arrive, usually the form or the phone, gets fixed first even if the DM test failed harder. The fixes themselves come in tiers. The free tier is process: give every channel a named owner and route every inquiry into one place, however manual. The structural tier is automated lead capture: a system that answers every channel in seconds with a real response, asks your qualifying questions, and files the inquiry with its source attached, so the audit's three grades pass by design rather than by effort. Repeat the audit quarterly either way; intake decays quietly as tools, staff, and platforms change.
One distinction worth making: this self-test tells you where inquiries land today. It does not tell you what fixing the gaps is worth or what to build first; that is a scoping question. If you want the professional version, the free AI audit traces your channels the same way, then comes back with a prioritized plan and a fixed price for the build.
FAQ
How do I test my lead capture without confusing my team?
Tell one person, or tell nobody and keep the inquiries small. Use a personal email and an unfamiliar phone number, send plausible but low-stakes requests, and withdraw each one politely once you have your timestamps. The point is to observe the normal process, so the less the team performs for the test, the better your data.
How often should I run this audit?
Quarterly, and after any change to your website, phone system, or staffing. Intake breaks when things change, a form plugin updates, a notification email points at someone who left, and each break is silent. A recurring 15-minute calendar block, four inquiries and a scorecard, catches what no dashboard reports.
What is a reasonable response time for a small business?
Aim for a useful reply within minutes on every channel, automated or not, and a human follow-up the same business day. Customer expectations are set by the fastest company they contacted, not by your staffing; the practical way to meet them without hiring a dispatcher is automation on the first touch and a clear owner on the second.
Ran the self-test and found a channel that goes nowhere? The free AI audit is the professional version: we trace every intake path, measure the gaps, and come back with a prioritized fix list, a scope, and a fixed price.