What it really costs

Here is the whole bill. Two things about it are worth noticing before you read it: almost every figure belongs to somebody else and changes without telling you, and the largest line item is not money.

The bill that isn't money

The first site is the slowest thing you will build, and the reason is structural rather than motivational. There are about a dozen decisions to get through — the phrase to target, how the pages are structured, what goes on the profile, how the tracking is wired — and search results answer you in weeks. So you get about one lesson per month. Most of your early guesses will turn out to be wrong, and you will find out slowly, one at a time, in the order the results happen to reveal them.

That is the actual mechanism behind “the first one takes forever”. It is not that a search engine punishes beginners. It is that a beginner's feedback loop is a month long, and there are a dozen loops to run. The second site starts with all twelve answers already in hand, which is why the people who stay in this describe site two as a different job entirely.

The money you should not spend

• Bought links and private networks. Buying links is against Google's own guidelines, and the thing you are paying for can be devalued or deindexed on a schedule you do not control. You would be renting a ranking, not owning one.
• A designed brand. Nobody has ever chosen a plumber on typography. The page needs to load, say what it does, and put the phone number where a thumb lands.
• Six domains at once. Your feedback loop is a month long. Six sites means six wrong guesses running in parallel instead of one guess you correct.
• Paid ads “to test the site”. Ads test whether the offer converts paid traffic. They tell you nothing about whether the page can rank, which is the only thing you are actually trying to learn.
• Software subscriptions before there is a single call. The only tool you need on day one is a tracked number.

What the worst case looks like

Be specific about it, because a vague worst case is one you never really price in. The realistic bad outcome is that you spend a few hundred dollars across several months, the site is indexed, and it sits on the second page for a phrase that is contested by three businesses who have been there for a decade. The phone never rings. Nothing dramatic happens. You simply stop, and you are out the domain, a few months of a phone number, and the evenings.

The upside of a small bill is that it makes the decision honest. Nothing here is expensive enough to justify quitting your judgement about the trade and the town. If you find yourself reasoning about sunk cost after two months, notice that the sunk cost is the price of a couple of takeaways and the real thing you are protecting is your time.

Sort the costs by what happens when you stop

Split the rows above into two piles, because they behave completely differently when you walk away. The domain and the pages are paid once and keep existing — a domain you stop renewing is gone at the end of its term, and until then the site is still there to sell or to point at something else. The hosting and the phone number are running meters. They keep charging you every month whether or not anything is happening, and they are the reason a dead project quietly costs money for a year after you stopped thinking about it.

That distinction is why the last lesson of this course is about writing a kill date. It is also a reason to be relaxed about the one-off costs and strict about the recurring ones: pay for the domain without agonising, and know exactly which day of the month the number bills, because that is the number that has to stop the day you decide it is over.