Most conversion rate optimization advice is written for sites with free traffic, where a test that fails costs nothing. When the visitors come from Google Ads or Meta Ads, every one of them was bought, so conversion rate stops being a design topic and becomes a line in the budget. In the accounts I run, the landing page decides whether a given budget produces a profitable month or a polite explanation.
This article treats conversion rate optimization (CRO) the way a business owner paying for clicks has to: what a conversion point is worth, which number to fix first, what to test on a paid landing page, and how to tell when the page is not the real problem.
Why CRO is a budget question when the traffic is paid
Cost per lead equals cost per click divided by conversion rate. Nothing else about the page matters to that formula, which is why a better page lowers your cost per lead without touching the auction. If conversion rate doubles, the same clicks produce twice the leads and cost per lead is cut in half.
How much that is worth depends on what your clicks cost. Diwizi's Paid Media Cost Index (October 2026 edition, live keyword data for the US market from DataForSEO Labs) reports the median Google Ads cost per click for each industry below.
| Industry group | Median CPC, Oct 2026 | What a conversion lift does |
|---|---|---|
| Legal | $98.55 | Each click is expensive enough that small lifts pay for a full redesign |
| B2B and SaaS | $80.43 | Form length and offer clarity move cost per lead more than bids do |
| HVAC and plumbing | $33.26 | Phone and form friction on mobile is the first place to look |
| Home services | $27.09 | Trust signals and a visible booking path matter most |
| Dental | $14.78 | Matching the page to the treatment searched, not a generic practice page |
| Healthcare | $6.37 | Lower click cost, so volume and tracking quality decide the result |
Source for the table: Diwizi Paid Media Cost Index, edition of October 2026, which also lists each keyword behind these medians and states that they are national medians, so a dense metro runs above them. The third column is my reading, not a measurement.
The practical point: the more a click costs, the more a conversion point is worth, so the earlier CRO should come in your priorities. An owner paying a single-digit click price can afford to test slowly. An owner paying the legal or B2B prices in the table cannot.
Conversion rate or click-through rate: which one to fix first
Fix conversion rate first when the clicks are already arriving. Click-through rate (CTR) measures how many people who see your ad click it. Conversion rate measures how many of those who clicked then do what you wanted. They sit on opposite sides of the click and are fixed in different places: CTR in the ad and keyword, conversion rate on the page and in the offer.
Google defines conversion rate as conversions divided by the ad interactions that can be tracked to a conversion, shown as a percentage, and its own Google Ads Help page on conversion rate gives the example of 50 conversions from 1,000 interactions, which is 5 percent. Note the wording: interactions, not people. If you track several conversion actions and count them all, the rate can pass 100 percent in the report, which is a reason to define one primary conversion before you compare anything.
A low CTR with a good conversion rate means the right people click and the ad is underselling. A good CTR with a poor conversion rate means the ad promises something the page does not deliver, and that is where the money leaks. The second case is the more expensive one, which is why I audit the page before I rewrite ads.
CRO versus SEO: same page, different job
SEO gets the right visitors to a page and CRO decides what share of them act. They overlap on the same URL but pull in different directions: SEO tends to add content and keywords, CRO tends to remove everything that is not helping the visitor decide.
For paid traffic the split is cleaner. A paid landing page does not need to rank, so it can be built for one search intent and one action. That is the single biggest advantage paid pages have over organic ones, and many businesses waste it by sending ad clicks to the homepage. If you want the page built for the ad, see my landing page optimization service.
Is my conversion rate good?
A conversion rate is good when your cost per lead sits comfortably below what a lead is worth to you, not when it matches a published average. Averages mix industries, traffic sources and conversion definitions, so a headline average tells you little about your account. A form fill on a free quote and a booked appointment are different events and should never be compared on one scale.
Work the other way around. Decide what you can afford to pay for a lead, divide your click price by that figure, and you get the conversion rate you need to break even. My article on customer acquisition cost against lifetime value shows how to set that ceiling. Then measure your own rate over a full cycle and judge it against your ceiling.
What to test first on a paid landing page
Start with the things a visitor sees before scrolling, because that is where most paid visitors decide. In the order I check them:
- Message match. The headline should repeat the promise in the ad. Someone who searched for "emergency plumber" and lands on a page about "quality home services" has to work to feel they are in the right place.
- One action. One primary button or form, repeated down the page. Menus, blog links and a second offer give people a way to leave.
- Form length. Ask only for what you need to call back. Every extra field is a reason to stop, and for a phone-driven business a tap-to-call button above the fold often outperforms any form.
- Mobile experience. Most local service searches happen on phones. Check the page on a real phone, on mobile data, and tap every button.
- Proof near the action. Reviews, licences, named results and guarantees placed beside the button, not buried in a footer.
- Speed. A slow page loses visitors you already paid for. Google's PageSpeed Insights reports how your page loads for real users.
Examples of what this looks like: a cleaning company that replaced a four-step quote form with a phone number and a two-field request, or a software firm that moved the pricing explanation above the demo form so that only qualified people submitted it. Both changed what the page asked, not how it looked.
How to read a test without fooling yourself
Change one thing at a time and let the test run long enough to cover a full week of traffic, because weekday and weekend visitors behave differently. Stopping a test the day one version pulls ahead is the most common way owners get false winners.
Small accounts often cannot reach statistical significance on a small change. In that case, make bigger changes (a different offer, a different page layout) rather than small ones (button color), and judge them on cost per lead over a few weeks. Keep a written log of what changed and when, so a spike in leads can be traced to its cause instead of argued about.
When the page is not the problem
Poor results are not always a conversion rate problem. Before redesigning a page, rule out three things that look like one.
- Broken tracking. If conversions are not being recorded, the rate looks terrible while the phone rings. My piece on the Google Ads no recent conversions warning shows how to tell a dead tag from a quiet week.
- Wrong traffic. Broad keywords bring people who were never going to buy. Search terms and negative keywords fix this faster than any page change.
- Weak lead handling. If leads arrive and nobody answers quickly, the conversion rate on the page is fine and the business is the leak.
If the account is spending and the leads are not coming, the Google Ads not converting troubleshooting guide walks through the order I check these in, and high cost per lead optimization strategies covers what to do once the cause is the page.
For the official definitions of the metrics used here, Google keeps them in the Google Ads Help Center.
