Google Search Ads Management Built on Query Control

What is the difference between keywords and search terms?

Keywords are what you bid on. Search terms are what people actually typed. They are not the same thing, and the gap between them is where most wasted spend lives.

A keyword on broad match can trigger on queries that share intent only loosely — sometimes usefully, often not. The search terms report shows the reality, and on accounts that have not been reviewed in a while it routinely contains research queries, competitor names, job seekers and people looking for free versions of what you sell.

This is why our audits start there rather than with bids. You can tune a bidding strategy perfectly and still be funding an audience that was never going to buy.

Should I use broad match or exact match?

Broad match with a well-built negative list and reliable conversion data can find queries you would never have thought to bid on. Broad match without those two things is an expensive way to discover what your competitors are called.

Exact and phrase match give control at the cost of reach. On smaller budgets that trade is usually right, because there is not enough spend to fund a discovery phase.

The practical answer for most accounts is a mix: exact and phrase for the terms you know convert, broad in a separate campaign with its own budget cap where you can watch it. What matters is that broad match is a deliberate decision with guardrails, not a default that nobody revisited.

Why is my cost per click so high?

Usually competition and Quality Score together. You cannot change how many competitors bid on your terms, but Quality Score is partly within your control and it directly affects what each click costs.

The components are expected click-through rate, ad relevance and landing page experience. Weak ad relevance — ads written for the brand rather than the query — and a landing page that does not obviously answer what was searched will both raise your costs against a competitor bidding the same amount.

The other common cause is bidding on head terms with high commercial competition when longer, more specific queries would convert better for less. A query with clear buying intent is usually cheaper and more valuable than the broad category term above it.

How is search engine advertising different from SEO?

Paid search buys placement immediately and stops the moment you stop paying. SEO earns placement slowly and keeps it. They intercept the same demand at different costs and on different timelines.

The useful way to run both is to let each do what it is good at. Paid search is the right tool for testing which queries actually convert, for seasonal pushes, and for terms where organic ranking is unrealistic. Organic is the right investment for the informational and comparison queries that feed demand in the first place.

They also feed each other. Search term data from a paid account is genuine query research for content, and a page that ranks organically usually makes a better paid landing page than one built for the campaign alone.

What does Quality Score actually change?

It affects what you pay for a given position, and it is a diagnostic more than a target. Google publishes the three components — expected click-through rate, ad relevance and landing page experience — and each points at a different fix.

The common mistake is treating it as a score to raise. It is more useful as a signal of mismatch: a low ad relevance component means the ad does not match the keyword, and a low landing page component means the page does not match the ad. Both are structural problems the number is reporting rather than causing.

Chasing the number directly produces odd behavior — pausing keywords that convert because their score is low, or splitting ad groups until each contains one keyword and there is no data to learn from.

The productive version is to read it per keyword, find the ones where the component scores disagree with your assumptions, and fix the mismatch. The score follows, and so does the cost.

How should automated bidding be used?

With the conversion data it needs, and with an understanding of what it is optimizing toward — which is whatever you told it, not whatever you meant.

Automated strategies need volume to work. An account with a handful of conversions a month does not give the system enough to learn from, and it will produce erratic results that look like the strategy failing when it is really the data being too thin.

The bigger risk is optimizing toward the wrong thing. If form submissions are the conversion and half of them are unqualified, the system will efficiently buy more unqualified submissions. It is doing exactly what it was asked to do.

Which is why conversion quality gets settled before bidding strategy. Feeding back real outcomes — qualified leads, actual sales rather than form fills — changes what the system optimizes for, and it is the single highest-leverage change available in most accounts.

How should an account be structured?

Around how you would answer the query, not around how your website is organized. The two are frequently different and the website usually wins by default.

Ad groups exist so that the ad and the landing page can match the intent behind a set of queries. When an ad group contains queries that want genuinely different things, the ad has to be vague enough to cover all of them, and vague ads perform accordingly.

The opposite failure is over-segmentation. Splitting until every ad group has three impressions a week means no group ever accumulates enough data to optimize, and the account becomes unmanageable without becoming more effective.

The workable middle is grouping by intent and by the page that serves it. If two queries would send someone to the same page and be answered by the same sentence, they belong together.

What do broad match and smart bidding do together?

They expand reach and hand a great deal of control to the system, which works well with good conversion data and badly without it.

Broad match now interprets intent rather than matching words, which means it will find queries you would not have thought of — some genuinely valuable, some unrelated. The value of the good ones and the cost of the bad ones depends entirely on how well the bidding signal distinguishes them.

That is why the two are discussed together. Broad match with manual bidding tends to spend widely with no mechanism for learning what worked. Broad match with well-fed automated bidding can find demand that exact match never reaches.

The control that remains is the one people neglect: negatives, informed by reading the search terms report regularly. That report is the only place the account tells you what it is actually buying, and it is worth more attention than any other screen — which is why we wrote a guide on reading it.

What makes a landing page work for paid search?

Answering the query it was reached by, quickly, and asking for something proportionate to where the person is in their decision.

The mismatch is the most common and most expensive fault. An ad group about a specific service sending traffic to a general services page forces the visitor to find their own way to what they asked for, and a share of them will not.

Speed matters more here than almost anywhere, because you are paying for each arrival. A page that takes several seconds on a mobile connection loses people you already bought, which is the most avoidable waste in a search account.

The form should match the intent. A high-intent query about a specific service can support a fuller inquiry form; a broad research query cannot, and asking for too much converts nobody.

And the page has to answer the obvious objection. If price is what people want and it is nowhere on the page, they leave to find it — usually on a competitor's site that published it.