Field guide
The Small Shop Front Office
Most local shops do not lose work to a competitor's better craftsmanship. They lose it in the short stretch between a customer deciding they need someone and deciding who to call.
This guide walks that stretch step by step: where customers find you, what makes them pick you, what happens when they reach out, and how you know any of it worked. Each step has the common failure, a test you can run yourself in a few minutes without hiring anyone, and what fixing it involves.
We build and run this exact front office for local businesses, so we have a view into where it breaks. Where we use an example from our own operation, it says so. Where a fix is something you can do yourself in an afternoon, the guide says that too.
Getting found
A customer with a burst pipe does not browse. They search once, read what comes up, and start calling from the top.
The listing sites answer for you, and they are wrong. Old hours from two seasons ago, a phone number from a line you dropped, a website link that dies. We verified the online presence of dozens of shops in this county this summer, and the most common serious defect was not a bad website but a wrong fact on a listing the owner had never seen. One diner we researched had an aggregator listing its website as a domain that is dead and belongs to a different diner entirely.
Search your business name plus your town in a private browser window. Open the first five results. Check three things on each: the phone number, today's hours, and where the website link goes. Then do the same on the map result. Ten minutes, and the wrong ones tend to show themselves.
Claiming your Google Business Profile, then correcting each listing at its source. It is unglamorous work: accounts, verification postcards, waiting. No tool does it all automatically, whatever the ad says.
Measure it by the facts being right, checked on a schedule, not once. Our own site and forms are checked by an automated watch every day, and we hear the day something breaks. The same discipline works at any size, even as a calendar reminder to re-run the ten-minute test monthly.
Getting chosen
Findable is not chosen. Between two plumbers with working phone numbers, the one with recent reviews and a page that answers questions gets the call.
An empty or stale review profile. The research here is unambiguous: BrightLocal's 2025 consumer survey found that only 4% of consumers never read reviews of local businesses, 84% use Google to find them, and 42% trust them as much as a recommendation from a friend. A shop with decades of word-of-mouth and one review online is invisible in the place the next generation of customers checks first.
Read your own Google reviews as a stranger would. Count them, note the newest one's date, and read the unanswered ones. Then read your best competitor's.
Asking every real customer, at the natural moment, with nothing attached to the ask. Never offer anything in exchange for a review: the FTC rule on consumer reviews and testimonials makes incentivised or filtered reviews a legal problem, and Google's own policy prohibits it. Reply to the ones you have, including the unkind ones, because the reply is read by the next customer more than by the reviewer.
Measure it by count and recency together. A shop with forty reviews where the newest is two years old reads worse than a shop with twelve from this season.
Answering
The customer found you and picked you. Now they call at 7:40pm because that is when their day allows it.
The promise and the desk disagree. In our verification sweeps we keep finding a site that promises 24/7 emergency service on one page and posts office hours of Monday to Friday, 9 to 5 in the footer of every page. Both statements are sincere. Only one of them answers the phone.
Call your own business after hours from a number it will not recognise. Listen to exactly what a customer hears. If it is voicemail, listen to your greeting the way a stranger in a small emergency would.
A decision about who or what picks up when you cannot: a person, a call service, or a receptionist system that answers from facts you approve. We build the last kind, so judge our bias accordingly. The honest comparison is that a human answering service runs more per month than most software, and voicemail converts worst of all. Whatever you choose, the test above is the measure: call it, listen, decide if you would leave a message with it.
Measure it by missed-call count per week. If nothing counts your missed calls today, that is the first gap: you cannot manage what nobody writes down.
Booking and following up
An answered inquiry that never becomes an appointment is the quietest way to lose a job. So is the quote that goes out and is never chased.
Follow-up lives in one person's head. The estimate sent Tuesday is forgotten by Friday because three new jobs landed. Nobody decided to ignore the customer. The system was a memory.
Pull up your last ten quotes or serious inquiries. For each: did it get a follow-up, and how many days later? The ones with no answer at all are the list worth working first.
A written list that survives a busy week, at minimum. A calendar block to chase quotes, next. Software that drafts the chase for your approval, only if the volume justifies it.
Measure it by the share of quotes that ever receive a follow-up, and the days between quote and first chase.
Knowing what happened
This is the step almost nobody has, and it is the one that makes the other four improvable.
Nothing is counted, so every decision is a feeling. Was the new website worth it? Did the answering change help? Without a before-count, no one can say, and whoever sold you the fix will happily say yes.
Answer from records, not memory: how many calls did you miss last week? How many quotes went unanswered last month? If the records do not exist, the honest answer is that you cannot know yet, and that is worth more than a guess.
Baselines before changes, and counts after. When we take on a client, the baseline gets written down on day one, before anything launches, because it cannot be reconstructed later. We hold ourselves to publishing only counted results with the client's written consent. Anything else is an anecdote, and we label anecdotes as what they are.
Measure it by having the numbers at all. One page, five numbers, updated monthly: found (searches or profile views), chosen (calls and inquiries), answered (missed-call count), booked (jobs), followed up (quotes chased). That page is the whole front office on one sheet.
Do it yourself, hire a specialist, or us
Swipe the table sideways for the last two columns.
| Step | Do it yourself | Hire a specialist | Where we fit |
|---|---|---|---|
| Getting found | The ten-minute listing test, monthly | A local-SEO service for citations at scale | The free scanner finds the visible defects, audits go deeper |
| Getting chosen | Ask every customer for a review, honestly | A reputation service, checked for FTC compliance | Review replies drafted for your approval, inside our plans |
| Answering | The after-hours self-call, then a decision | An answering service, human and priced likewise | The Receptionist answers from facts you approve |
| Booking and follow-up | The written quote list | An assistant or office manager | The Scheduler and follow-up work inside our plans |
| Knowing | The five-number monthly page | A bookkeeper who will track operations numbers | Every plan emails a monthly count of what it actually did |
The honest column is the first one. Every step has a free version, and a shop that does all five free versions consistently beats a shop that buys tools and ignores them.
One next step
Pick the step above where you already suspect the answer is bad, and run its test this week. All five are free and none of them need us.
If you want a second pair of eyes on the first two steps, our scanner reads your site and listings the way a customer would and names what it finds. It is our own tool, not an AI and not PageSpeed, and the first findings need no signup.
Methodology and change log. Written from BellCode's verification sweeps of local businesses in Washington County RI and New London County CT in the summer of 2026, our own operated systems where they are cited as ours, BrightLocal's Local Consumer Review Survey 2025, and the FTC Consumer Reviews and Testimonials Rule at 16 CFR 465. Figures re-read at their sources before publication. Changes to this guide will be listed here with dates.