How to Generate Quality Leads from Google Ads

Plenty of accounts generate leads. Far fewer generate leads the sales team is willing to call twice. The difference is rarely the bidding, and it is almost never the budget — it is that the account has been optimised towards the wrong definition of success.

This is how to build a Google Ads account that produces leads worth having.

Decide what a good lead is, in writing

Before touching the account, get a written answer to one question: what makes a lead worth having?

For most businesses it is some combination of the enquiry being in the right country, having a budget in the right range, wanting something you actually sell, and having some intention of deciding this quarter rather than next year.

Write it down, because this definition becomes the thing the whole account optimises towards. Vague definitions produce vague campaigns.

Then get the sales side to mark outcomes. Even a crude flag — junk, unqualified, qualified, quoted, won — turns your campaign data from click statistics into a picture of what buys business.

The lead quality problem is usually a measurement problem

Here is the mechanism that quietly wrecks accounts.

Google's automated bidding optimises for the conversions you report. If every form submission counts as a conversion, the system learns to buy form submissions — and the cheapest form submissions come from the least qualified traffic. Six weeks later you have twice the leads, half the quality, and a bidding algorithm working exactly as instructed.

The fix is to report back what actually happened. Two approaches, in increasing order of usefulness:

Conversion values. Instead of counting every lead as 1, send a value reflecting what that lead type is typically worth. A quote request is worth more than a newsletter signup, and bidding should know that.

Offline conversion import. Send the outcome back to Google once sales know it. The lead that became a customer gets imported as a conversion, with its real value. Bidding then optimises towards revenue rather than form fills, which is what you wanted all along.

Offline import takes real setup work — you need to capture the Google click identifier with the lead, store it in your CRM, and send outcomes back. It is the highest-leverage thing most lead-generation accounts are not doing. The mechanics are in the conversion tracking guide.

Intent lives in the search terms, not the keywords

Keywords are what you bid on. Search terms are what people typed. The gap between them is where lead quality is won or lost.

Some patterns worth knowing:

• Searches naming a specific service convert better than category searches. "Emergency plumber" beats "plumbing".
• Searches with commercial modifiers — hire, buy, quote, cost, near me, agency, service — carry buying intent. Research modifiers — what is, how to, tutorial, free, DIY, template, salary — usually do not.
• Brand searches for your competitors convert at low rates and high cost, and are worth testing carefully rather than assuming.

Read the search terms report every week without exception. It is the only place the platform tells you the truth about what you are buying. Every irrelevant term you find becomes a negative keyword; every good one you find becomes a keyword in its own right.

For business-to-business accounts specifically, the recurring offenders are job seekers, students and bargain hunters. This list of negative keywords covers those categories.

Make the ad do some qualifying

An ad that promises everything to everybody produces enquiries from everybody. Some friction is useful.

If you have a minimum engagement size, a specialism, or a service area, say so in the ad. Naming your price band, when you can, is the single most effective qualifier available — it costs you the clicks from people who were never going to buy, which is the point.

This feels counterintuitive because click-through rate falls. Let it. You are not paid for clicks.

The other half is message match: the ad should use the language of the search, and the landing page should open with the language of the ad. Each point where the wording drifts, people decide they are in the wrong place. That is covered in more depth in writing ad copy that matches the search.

The form is part of the campaign

Most lead quality debates end at the form, and most forms are built without any thought about who they filter.

Short forms produce more leads of lower average quality. Longer forms produce fewer, better ones. Neither is right in the abstract — it depends on whether your constraint is volume or sales capacity. Decide which problem you have before changing the form.

A few things that reliably help:

• Ask one or two qualifying questions — budget range, timeline, company size — rather than more contact fields.
• Make the phone number optional if you want volume, required if you want commitment.
• Put a honeypot field in to absorb bot submissions, and check what proportion of your "leads" are automated. In many accounts it is not small.
• Confirm the form actually delivers. Broken notification emails are more common than anyone would like, and they look exactly like a lead quality problem from inside the ad account.

Speed of response beats almost every optimisation

This is not a Google Ads setting, and it usually matters more than any of them. A lead contacted within a few minutes is far more likely to convert than the same lead contacted the next day, because they are still at their desk, still thinking about the problem, and have usually enquired with two or three other suppliers as well.

If enquiries arrive into an inbox nobody watches after 6pm, no amount of bidding strategy will fix the resulting numbers. Route them somewhere they will be seen — a shared inbox, a CRM, a WhatsApp notification — and measure how long it takes to make first contact.

Landing pages that convert the right people

Send each campaign to a page that answers its specific search. Not the homepage.

The order to fix things in:

1. Message match — the page opens with what the ad promised.
2. Speed on mobile — most paid traffic is mobile, and slow pages lose people before they read anything.
3. One clear action, visible without scrolling and repeated further down.
4. Proof — real credentials, real case studies, real names. Anything unverifiable is worse than nothing, because the people you want are the people who check.
5. Only then, layout and colour tests.

Most teams start at step five. The landing pages guide goes through the whole sequence.

Targeting settings that quietly cost you

Three defaults worth checking on every campaign:

Location targeting. The default includes people "regularly in or showing interest in" your target area — which can mean enquiries from well outside it. If you serve one city, set targeting to people present in that location.

The Search Partners network and Display Expansion are on by default in some campaign types. Both extend reach beyond Google Search itself, and both typically convert worse. Segment your reporting by network and judge them separately rather than letting them hide inside a blended cost per lead.

Ad scheduling. If enquiries can only be handled during business hours, there is a real argument for concentrating budget there. Check the data before assuming — some businesses get their best leads at 10pm.

Remarketing, used with restraint

Remarketing to people who visited a key page and did not enquire is among the cheapest useful traffic available. Remarketing to everyone who ever landed on the site, indefinitely, is a way to spend money reminding uninterested people that you exist.

Segment by behaviour — visited the pricing page, started the form, did not finish — and keep the membership windows short enough to match how people actually buy in your market.

What to measure

Track these, in this order:

1. Cost per qualified lead, not cost per lead.
2. Qualified rate — what proportion of enquiries survive first contact. If this falls while volume rises, the account is optimising towards junk.
3. Cost per acquired customer.
4. Revenue per customer, so the first three numbers can be judged against something.

If you only ever look at cost per lead, you will eventually be congratulated for a number that is destroying the sales team's week. The wasted spend calculator works the true cost per customer through from these figures.

The short version

Define a good lead. Report real outcomes back into the platform so bidding optimises towards them. Read the search terms report weekly. Let the ad and the form do some qualifying. Answer fast. Judge everything on cost per acquired customer.

None of that is complicated. It is just less interesting than changing bid strategies, which is why so few accounts do it.