Pay-per-click advertising means buying a place in a set of results and paying only when somebody clicks through. It is the fastest way to put a business in front of people who are actively looking, and it stops the day you stop paying, which is the trade at the heart of it. This page covers what it costs, how you know whether it worked, and the situations where we would tell you not to bother.
The mechanic is simple enough to explain in a sentence, and every expensive mistake in this discipline comes from one of the four things underneath that sentence.
The price is set by the auction. The value is set by your conversion rate and your margin, and it is the number most advertisers have never actually worked out. Until you know what a visit from a given search is worth to you, a bid is a guess wearing a decimal point.
Cheap clicks that never convert are more expensive than dear ones that do. What matters is cost per acquisition, and after that the return on the whole spend. Optimizing toward a lower cost per click on its own reliably makes an account worse.
A category with well-funded advertisers in it will cost what it costs. In some markets the top-of-page bids run to tens of dollars a click, which is not a scandal — it is a signal about what a customer is worth in that category, and it should tell you something before you enter.
Setting a daily budget so low that the campaign stops mid-morning does not make the advertising cheaper. It makes it partial, and it removes the data the bidding needs to get any good.
Every one of these is cheaper to fix before the money starts than after, and an account missing any of them will produce numbers that cannot be acted on.
Without it you are buying traffic and guessing. With it half-installed, which is more common, you are buying traffic and being confidently misled. This is the first thing we check in any account we are handed and it is wrong more often than it is right.
One job, the shortest form you can live with, and a message that matches whatever the person clicked. Sending paid traffic to a home page is the most expensive habit in this discipline.
Paid search produces inquiries at whatever hour people search. If calls go unanswered after five, that is not a marketing problem you can bid your way out of, and it will show up as a low conversion rate that gets blamed on the ads.
The first weeks of any account are partly tuition. Automated bidding needs conversions before it is any use, and the spend that produces those conversions is bought at a worse rate than the spend that follows. Budgeting as though month one performs like month six is how accounts get canceled in month two.
We sell this service, so read the list with that in mind. It is here because the accounts that fail tend to fail for reasons that were visible before anyone spent anything, and saying so early costs us one engagement and saves a relationship.
Paid search harvests demand that already exists. If your category is genuinely new, there is no query to buy, and the money is better spent creating the demand somewhere it can be created.
Work out what a customer is worth and what a click costs in your category before you start. Some businesses simply cannot make the arithmetic work at current prices, and the right answer is to say so rather than to run a smaller version of a losing campaign.
If existing traffic is not turning into inquiries, buying more of it multiplies the leak rather than fixing it. Conversion work first, then the media buy, in that order and not the other way around.
The moment the budget stops, so does the traffic. That is not a criticism of the channel, it is what renting means. If the requirement is an asset that keeps producing, that is an argument for organic search alongside it rather than instead of it.
Three passages, quoted as published. The first is the plainest definition of the model we have written; the other two are the two pieces of advice from the era that have aged best.
PPC is an advertising model. If you're an advertiser, you pay the site hosting your ad only when somebody clicks the ad.
Sixteen years on this is still the whole definition. Everything that has changed since sits in how the price of that click is decided and who decides which click you get.
Spending $20 a day may appear to be the way to go for cost-conscious advertisers, but a budget that low will also limit your ability to profit.
This matters more now than it did then. Automated bidding needs a flow of conversions to learn from, and a budget that starves the campaign starves the learning as well as the traffic.
Rather than go for the homerun, aim for the base hit that will help you score the number two, or even number three position.
Written before automated bidding existed, and it turned into the standard argument against chasing the top slot. The positions immediately below it usually convert nearly as well and cost noticeably less.
Around 700 posts, published since 2003
Read the archiveThis page is the ground floor. These three go up from it.
The hub: what decides the price of a click, the five paid surfaces, and one real month of an account reported in full including the numbers that went the wrong way.
/ppc/Paid and organic side by side: what each one is good at, how to split a budget between them, and why the comparison is usually framed wrongly.
/search-engine-marketing/What a managed month contains, and the four problems we find in almost every account we are handed.
/ppc-management/Tell us what you sell, what a customer is worth to you and what you would want to spend. We will look up what clicks actually cost in your category and tell you plainly whether this is a channel worth entering, including when the answer is no.