You cannot tell from the account-level number, and almost every published answer tells you to fix both at once. Segment conversion rate by search-term intent instead. If high-intent searches convert and the rest don't, it's targeting. If nothing converts, it's the page or the offer.

On this page

• How many clicks do I need before zero conversions means anything?
• How do I tell whether it is my targeting or my landing page?
• What should I fix first in each case?

How many clicks do I need before zero conversions means anything?

Do not run this analysis until you have enough clicks to see anything. At a 2% conversion rate you need roughly 150 clicks before zero conversions means anything at all — below that, zero is the expected outcome of a working funnel. Most "no conversions" panics are 40 clicks old.

And before either branch: confirm the conversion is actually firing. Send a test transaction. A broken tag looks exactly like a broken page.

How do I tell whether it is my targeting or my landing page?

Bucket your search terms by intent:

• High intent — the product or service plus a buying word. "emergency electrician near me",

"buy X online", brand plus "price".

• Mid intent — the category with no buying signal. "how much does an electrician cost", "best X for Y".
• Low intent — research, definitions, DIY, jobs, free, competitors' brands.

Then compute conversion rate per bucket.

High intent converts, everything else doesn't → your targeting is buying the wrong traffic. The page is fine. Fix match types, negatives and the query mix.

Nothing converts, including high intent → the page or the offer is the problem. This is the case where people waste months on bid adjustments.

Everything converts but the volume is tiny → neither is broken. You have a demand problem, which is a different conversation.

What should I fix first in each case?

Targeting branch. Negatives validated across multiple windows (see the negatives answer — do not negate on one window). Match-type tightening. Check whether exact match is matching to things that are not close variants, which it does more than people expect.

Page branch. In order: does the page answer the exact search, is the action obvious above the fold, is there a reason to act now, and is the form asking for more than the offer is worth. A form asking nine questions for a quote will out-lose any bid strategy.

Related questions

Why am I getting fake leads and spam form fills from Google Ads, and how do I stop them?

Most spam form fills are bots harvesting public forms, not a Google Ads problem — but if they are counted as conversions they corrupt your bidding, which turns a nuisance into a real cost. Read the full answer.

My Google Ads campaign has spent thousands with zero conversions — is the traffic bad or is something set up wrong?

Check what the campaign is allowed to bid on before you touch the traffic. A campaign spending thousands with zero conversions is far more often pointed at the wrong goal, or at a broken tag, than buying bad traffic. Read the full answer.

I took over a Google Ads account — how do I check whether the existing negative keywords are blocking good traffic?

Run the negative audit backwards. Pull every negative list attached to the account and check its terms against the account's own converters across multiple windows — greedy single-token negatives on shared lists are the ones doing damage. Read the full answer.

Written by Usama Khan, who runs Google Ads for e-commerce brands and home-services companies at AdventurAds. Figures are from accounts under active management and are anonymised. Last reviewed 2026-08-26.