You Can't Improve What You Don't Measure. Here's What's Cheap to Start.

Most owners don't avoid measurement because they think it's useless. They avoid it because every time they look into it, someone tries to sell them a platform.

You don't need a platform. You need to know five things, and every one of them can be started this week with tools you already have and roughly zero dollars. The software conversation can come later, once you know which numbers actually change your mind about something.

Here's the cheap version of a measurement system, in the order I'd build it.

1. Count your leads. By hand if you have to.

Start a spreadsheet. One row per inquiry. Six columns:

Date. Name. Phone. Source. Service they asked about. What happened.

That's it. That's the whole thing. Every phone call, every form, every walk-in, every text, every "hey my buddy said to call you."

This feels too simple to matter. It is the most valuable measurement asset a small business can own, and most don't have it.

Within a month you'll be able to answer questions you currently guess at:

How many leads do I actually get in a week?
Which source produces the most, and which produces the ones that buy?
What percentage of inquiries turn into money?
Which service do people ask about most?

You've been running the business on a feeling about all four of those. The feeling is usually wrong, and it's usually wrong in an expensive direction.

Cost: free. Time: about thirty seconds per lead.

2. Get a tracking number and find out how many calls you miss

This one costs a few dollars a month and it is the highest-return measurement purchase available to a small business.

Get a call-tracking number, put it on your website and your listings, and forward it to your real phone. Now every inbound call is logged. Time, duration, and — critically — whether anyone answered.

Then go look at the missed-call log.

I have never seen an owner look at that log for the first time and not be unhappy. Everybody underestimates this number. Everybody. You think you miss a couple a week. You're missing several a day, most of them during work hours, when you were up a ladder or on another call.

Do the arithmetic yourself. Take your missed calls last month. Assume some fraction of them would have converted at your normal rate — use your own numbers, not mine. Multiply by your average job. That's the size of the leak, in dollars, computed from your data.

You now know exactly how much a missed call costs you, and you can decide what it's worth to fix.

3. Write down where every customer came from

Not the lead. The customer — the ones who paid.

One field on the job record or the invoice: source. Filled in every single time, by whoever takes the job.

Do it for ninety days and then sort by source and add up the revenue. What you'll usually find is that your money is concentrated in one or two channels, and your effort is spread evenly across six. That gap is the most actionable thing you'll learn all year.

The best way to fill this field is to ask out loud, on the first call: "How'd you hear about us?" People answer honestly and they don't mind being asked. It takes five seconds and it beats any attribution software, because it captures things software can't see — the referral, the truck they saw, the guy at church.

Cost: free. Time: five seconds per customer.

4. Time-stamp your first response

Two timestamps: when the lead arrived, and when a human actually reached them. Subtract.

You can do this by hand in the same spreadsheet. Add two columns.

Within two weeks you'll have an average, and you'll have the distribution, which is the part that stings. Your average might be forty minutes. But look at the spread. There's a lead in there that waited two days. Find out why. That's not a data problem, that's a process problem, and it's costing you jobs.

Response time is my favorite number to hand a small business, because it is:

Entirely within your control
Directly tied to whether you win the job
Free to measure
Cheap to fix

Almost nothing else on the list has all four.

5. Do the two pieces of arithmetic nobody does

Set aside one hour. Not this week — right now, today.

What does a customer cost you? Add up everything you spent last month trying to get customers. Ads. Listing fees. The retainer you pay somebody. A fair value on the hours you personally spent chasing work. Divide by the number of new customers you got. That's your acquisition cost.

What is a customer worth? Average job value, times the number of times a typical customer buys from you over a few years, minus roughly what it costs you to deliver the work. Rough is fine. Rough is infinitely better than unknown.

Put the two numbers next to each other.

That single comparison will settle arguments you've been having with yourself for years — whether to spend more on ads, whether that lead service is a ripoff, whether it's worth taking small jobs, whether a discount to win a first-time customer makes sense.

It's a napkin and ten minutes of honest addition. No tool required. Nobody does it.

What you'll notice after a month

You'll have four things you didn't have before: a lead count, a missed-call count, a source breakdown, and a response time. Plus two numbers on a napkin.

And you'll notice something else. The act of measuring changes behavior before you fix anything.

The moment your team knows response time is being tracked, response time improves. The moment you can see missed calls on a report, you start answering more of them. This isn't a trick — it's just that most operational failures survive because nobody's looking at them, and looking is most of the cure.

When to buy software

Buy software when the manual version has proven the number matters and the manual version has become annoying. Not before.

If you've been logging leads by hand for two months and it's genuinely tedious and you keep referring to the sheet, buy the CRM. You'll know what you need it to do, because you've done it by hand and you know where it hurt.

If you buy the CRM first, you'll configure it around numbers you haven't thought about, your team won't use it, and in six months you'll be paying $200 a month for a system that holds a partial list of people who called you.

Measure by hand. Feel the pain. Then automate the specific pain.

Start here, today

Open a spreadsheet. Six columns. Log every lead starting now.
Order a call tracking number this afternoon.
Ask every customer this week where they heard about you, and write it down.
Add two timestamp columns to the sheet.
Spend an hour tonight computing what a customer costs and what one is worth.

That's a full measurement system. Total cost: a few dollars a month and about twenty minutes of your attention a day.

You'll know more about your business in thirty days than you have in the last three years.

If you'd rather have all of this running automatically instead of living in a spreadsheet — that's exactly the kind of quiet plumbing I build. Get in touch once you've got a month of data and know what hurts.