FUNDAMENTAL 08
KNOW WHERE THE
DEALS CAME FROM.
Most small businesses cannot answer a simple question: which marketing produced the last ten customers. Not approximately. Specifically. Without that answer, every budget decision is a guess wearing a suit.
The five numbers
Elaborate dashboards get built once and never opened. What survives is a small set of counts that someone actually looks at every week.
- Leads. Anyone who raised a hand. Counted by source, with one consistent definition. Half the arguments in small businesses about lead volume are really arguments about definitions.
- Contacts. Leads you actually reached and had a real conversation with. The gap between leads and contacts is a follow-up problem, and it is usually the largest leak in the business.
- Appointments. Whatever the meaningful next step is in your business — a site visit, a consultation, a walkthrough. The gap between contacts and appointments is a conversation problem.
- Contracts. Agreements signed. The gap between appointments and contracts is an offer or pricing problem.
- Cost per each. Spend divided by each stage above, by source. This is what turns activity into a decision.
The value is not in any single number. It is that the gaps between them tell you which problem you have. A business with plenty of leads and few contacts has a completely different problem from one with plenty of appointments and no contracts, and without the stages you cannot tell them apart.
How to structure source tracking
The goal is that every lead carries its origin with it from arrival to closed business. Most systems break because the source gets recorded at the start and lost somewhere in the middle.
- One field, fixed options. A dropdown, not free text. Free text produces six spellings of the same channel and a report nobody trusts.
- Distinct phone numbers or forms per channel where it is practical. A number that appears only on the mailer removes the guessing.
- Tag the campaign, not just the channel. Knowing mail worked is less useful than knowing which list and which message.
- Ask the person. Put a plain question early in the conversation and record the answer verbatim before it is categorized.
- Carry the source to the close. If it disappears when a lead becomes a contract, you can measure cost per lead and never cost per customer.
- Record the date of first contact. Without it you cannot build cohorts, and without cohorts a channel with a long lag looks broken.
This can live in a spreadsheet. What it cannot do is live in several places with different definitions, which is the normal state of affairs.
If this was useful, Cash Property Offers goes deeper.
Why last-touch attribution lies
Most tracking credits whatever happened immediately before the conversion. It is simple, it is the default in most tools, and it systematically misreads what actually happened.
Consider a plausible path. Somebody gets a mailer and keeps it on the counter. Weeks later they mention it to a neighbor who says they have heard of you. They search your name, read two pages, think about it, and call a month after that. Last-touch records a branded search. The mail, the referral and the content are invisible.
Now watch the damage. The branded search line looks efficient, so budget moves toward it. The mail looks unproductive, so it gets cut. Six months later the branded searches dry up, because nobody was creating the awareness that produced them. The business has optimized its way into harvesting a crop it stopped planting.
Long sales cycles make this worse. The slower your customers decide, the more your attribution understates everything that happens early.
What to do about it
You cannot solve attribution. You can stop being fooled by it, which is cheaper and nearly as useful.
Ask, and record the words. Self-reported attribution is imperfect and it catches things no tracking system sees. “My neighbor told me about you, and I think I got a letter” is a sentence no platform will ever report.
Report by cohort, not by month. Group leads by the month they first arrived and follow that group forward. Otherwise this month's closings get compared against this month's spend, and every channel with a lag looks worse than it is.
Use holdouts. Leave a comparable area or segment unmailed for a period and watch what happens to total volume. Blunt, slow, and the most honest instrument a small business has.
Watch the total, not just the lines. If total leads and total closings hold while one channel's reported numbers collapse, you probably have an attribution shift rather than a performance problem.
Change one thing at a time. Two simultaneous changes produce a result you cannot assign to either.
The weekly rhythm
Tracking only works if somebody looks. Put it on the calendar as a short recurring meeting with a fixed agenda.
Read the five numbers by source. Find the largest gap between two adjacent stages. Name the one thing you will change this week, and write down what you expect to happen. Next week, check whether it did.
Two rules keep it useful. Do not change the definitions mid-stream — a redefined metric erases your history. And do not judge a channel faster than it can produce data. Weekly review is for spotting the operational leaks. Channel decisions belong to a longer clock, set in advance, so the decision is not made in a bad week.
This is the least exciting part of marketing and the part that makes everything else improvable.
Frequently asked
Questions people actually ask
What is the minimum tracking a small business needs?
Five counts by source, updated weekly: leads, contacts, appointments, contracts, and cost per each. A spreadsheet is sufficient. Consistency of definition matters far more than the tool you keep it in.
How do I track offline channels like direct mail?
Use a phone number that appears only on that piece, a unique landing page or code, and ask the person directly. Combine all three, because each one alone misses a portion of the response.
Why not just trust the ad platform's numbers?
Platforms report on what they can see and are not neutral about their own contribution. Use them for optimizing inside a channel and use your own records, all the way to closed business, for deciding between channels.
What does 'cost per lead' actually tell me?
Less than people assume. A cheap lead that never becomes a customer costs more than an expensive one that does. Carry the cost through to contracts before you compare channels.
How often should I review the numbers?
Weekly for operations, so leaks get caught fast. Channel-level decisions belong on a longer schedule set in advance, because a weekly view of a slow channel will talk you out of it before it has produced readable data.
What if my numbers are a mess right now?
Start counting today rather than reconstructing the past. Three clean months are more useful than a year of reconstructed guesswork, and the reconstruction usually stalls the whole project.
Make your next move
A year from now, what will you be glad you started today?
You don't need another promise that everything will be easy. You need something useful to learn — and a next step you're willing to take.