Google Ads Text Tool

How do you draft and check a complete responsive search ad?

Enter three headlines and two descriptions to meet the tool's minimum, then review the weighted counts before exporting. For a worked example, draft a bicycle repair ad using the text below. Counts include spaces and punctuation; these examples use ordinary Latin characters, each worth one unit.

A complete minimum draft with calculated character counts

Field	Example text	Count / limit
Headline 1	Local Bike Repairs	18 / 30
Headline 2	Book Your Tune-Up	17 / 30
Headline 3	Ride Again This Week	20 / 30
Description 1	Book a bike tune-up with our local team. Choose a time that suits you.	70 / 90
Description 2	From brake checks to gear adjustments, get your bike ready for the road.	72 / 90
Display URL	example.com	11 / 35
Path 1	bike-repair	11 / 15
Path 2	book	4 / 15

The summary reads 3/15 headlines and 2/4 descriptions valid. With these URL fields within their limits, it also reports that all fields are within character limits. The preview displays example.com/bike-repair/book. Enter the actual destination separately in Final URL (for CSV export); the displayed path does not create a landing page.

Which character limits and completeness checks apply?

The workspace contains 15 headline rows limited to 30 weighted characters each, four description rows limited to 90 each, two path fields limited to 15 each, and a Display URL check of 35. The minimum is three valid headlines and two valid descriptions. Empty rows do not count toward either total; filled rows over their limit do not count as valid.

Over-limit asset fields receive a red counter and border, but text remains editable rather than being cut off. The overall success state requires every filled headline and description to fit, the minimum valid counts to be met, and all three display fields to fit. Duplicates and shared pin positions have separate warnings. A green summary does not certify originality, policy compliance, URL validity or landing-page availability. Download CSV remains available even when the summary reports a problem.

How does weighted counting handle spaces and multilingual text?

The counter iterates Unicode code points. Configured East Asian wide and fullwidth ranges count as two units; other code points count as one. For example, 東京 counts as four units, while halfwidth katakana remains one unit per code point. Ordinary spaces, punctuation and Latin letters each add one. A plain string-length check can therefore disagree with this tool.

A single-code-point emoji counts as one here, but a visible emoji made from several joined code points counts each part separately. Combining marks also contribute separately. The counter measures neither visible glyphs nor pixel width. Leading and trailing spaces typed directly into a field affect its counter, although CSV export trims surrounding whitespace. Clean accidental spaces before reviewing counts. For general text statistics outside this ad workspace, use Character Counter.

How do pasted lists and duplicate warnings work?

Use Paste list (one headline per line), then click Fill headlines. The importer trims each line, removes blank lines and accepts the first 15 nonempty lines. Its notice reports how many additional lines were ignored. Commas do not split headlines, and this box does not import descriptions or a CSV file.

Pasting a shorter list overwrites only the corresponding first headline rows. Later rows keep their previous text, and existing pins remain attached to their rows. Use Clear All first when replacing the entire ad. Duplicate checks run separately for headlines and descriptions, ignoring surrounding whitespace and letter case. Every matching row receives a Duplicate badge. Similar wording with different punctuation is not treated as an exact duplicate; review those variations yourself.

How do pinning and the device previews work?

Each headline offers No pin or Headline 1, 2 or 3; descriptions offer No pin or Description 1 or 2. The preview fills up to three headline slots and two description slots. A uniquely pinned asset keeps its assigned slot, while remaining assets are shuffled into vacancies without selecting the same row twice. Shuffle preview requests another draw, which can produce the same combination again.

When two rows share a pin, the first nonempty row claims that slot. Later rows with the same pin enter the free pool and can appear elsewhere if space remains. Resolve the warning before using the illustration to explain a strict placement requirement. Desktop and Mobile use the same selected copy at different widths. The display joins headlines with pipes and descriptions with spaces; it does not simulate exact live truncation or predict which combination will serve.

How do you export copy for an Ads Editor handoff?

Download CSV creates responsive-search-ad.csv with a header and one data row. It includes Headline 1 through 15, Description 1 through 4, Path 1, Path 2 and Final URL. Nonempty pinned assets add their position columns. Display URL is a preview field and is not included in this CSV. Campaign and Ad group columns are not generated; supply the necessary account context during your import workflow and review the mapping before posting.

CSV fields follow RFC 4180 quoting for commas, quotes and line breaks. Values beginning with formula-like characters such as an equals sign receive a leading apostrophe for spreadsheet safety. That protection can change the literal exported value, so review unusual leading punctuation. The download does not publish ads or connect to an advertising account, and it does not block incomplete drafts.

How do you save drafts and create an approval image?

Save draft downloads JSON containing all headline and description rows, their pins, Display URL, both paths and Final URL. Open draft reads a compatible saved file and replaces the current form. It does not merge campaigns. Form edits stay in browser memory without automatic persistence, so save before reloading or using Clear All.

Export for approval produces a dated PNG with Desktop and Mobile versions of the current combination. It also lists filled headline and description text, counts, pins, display-field counts and Final URL. Keep JSON for editable revisions and use the image for review. Neither the image nor a successful count constitutes platform approval.

How can you make the draft useful beyond passing the counters?

Give each headline a distinct role: service, benefit, location or next action. Read several shuffled combinations to catch repeated promises or descriptions that depend on a specific headline. Confirm claims against the actual offer and landing page; this tool checks text structure rather than whether a promise is true.

Keep the ad draft separate from organic search metadata. Use SERP Title Checker and Meta Description Checker for those fields. Their previews address a different publishing task, so passing their checks does not replace this workspace's ad-specific limits.