Pinning RSA Headlines in 2026: Five Slots, Five Rules and What Each Pin Costs
A responsive search ad holds fifteen headlines and four descriptions, and shows at most three headlines and two descriptions. Which ones is the system’s call. Pinning RSA headlines is the only lever an advertiser has over that call, and it is also the most misused control in text advertising — Google states outright that pinning “isn’t recommended for most advertisers”. What follows: what a pin actually buys you, which five slots exist, why a pin to headline 3 guarantees nothing, how many combinations each pin costs you, and the four situations where pinning is justified by money rather than habit.
What Google is actually assembling
Before you start locking pieces down, it helps to know what you are interfering with.
- Asset limits: up to 15 headlines and 4 descriptions per RSA. Minimum at creation: 3 headlines and 2 descriptions.
- Lengths: 30 characters per headline, 90 per description, 15 per path field. In double-width languages (Korean, Japanese, Chinese) each character counts as two.
- What serves: the system selects at least one headline and one description for any given ad. Headline 1 and description 1, in Google’s own phrasing, “usually” appear. Headline 2, headline 3 and description 2 are not guaranteed.
- Ceiling in the headline position: three headlines.
Then come the parts people forget. A headline can appear at the start of the description. Up to two unused headlines from the same RSA can serve as link-based assets. Unused headlines and descriptions from a different RSA in the same ad group can also serve as link-based assets, pointing at that ad’s final URL. However, sensitive verticals are not eligible for that asset sharing. And on the Display Network, headline 2 may not show at all in some formats.
First practical conclusion: the ad is not your copy, it is a kit assembled from your copy. A pin fixes one piece of the kit, and the price is every combination where that piece would have sat somewhere else.
Five slots: what a pin guarantees and what it does not
The documentation names five pin positions — headline 1, headline 2, headline 3, description 1, description 2 — and splits them into two unequal groups.
| Slot | Serving guarantee | Suitable for |
|---|---|---|
| Headline 1 | Always shows | Brand, mandated wording, disclaimer |
| Headline 2 | Always shows | Second mandatory line, compliance copy |
| Headline 3 | Not guaranteed | Nothing mandatory belongs here |
| Description 1 | Always shows | Offer terms, regulatory caveat |
| Description 2 | Not guaranteed | Nice-to-have the ad works without |
Google’s wording is direct: content pinned to headline position 1, headline position 2 or description position 1 will always show. By contrast, Google does not guarantee that content pinned to headline position 3 and description position 2 will show in every ad.
This matters more than it looks. If legal requires “18+” or “credit provided by partner bank” in every ad, pinning it to headline 3 solves nothing: the pin is there, the guarantee is not. The only three slots that satisfy a mandatory-text requirement are headline 1, headline 2 and description 1.
Mechanics: what happens to a slot once you pin it
The path: create or edit an RSA → pin icon to the right of a headline or description → choose a position → Save.
Four rules then apply, and all four are worth memorising.
- A pinned asset only shows in its slot. It stops competing for any other position.
- No other asset can fill that slot. The position is now closed.
- Several assets pinned to one position rotate within it. Google’s own example: pin both “Official website” and “The official site” to headline position 1, and ads will show either one as the first headline.
- Also, if you pin every available position, unpinned assets do not show at all. Your other twelve headlines effectively cease to exist.
That fourth rule is the quiet money-burner. An advertiser diligently writes fifteen headlines, pins the three “most important” one per slot, and ends up with exactly one possible ad instead of thousands. Fifteen headlines of work, three headlines of value.
The arithmetic: how many combinations each pin costs
Take a simple model: 3 headlines chosen from 15 and 2 descriptions from 4, ignoring order. That is C(15,3) × C(4,2) = 455 × 6 = 2,730 combinations. This is arithmetic on Google’s published limits, not an official Google figure — the documentation never states a combination count and does not describe ordering.
Now watch what each pin does to that number.
| Configuration | Combinations (model) | Share of unpinned |
|---|---|---|
| Nothing pinned | 2,730 | 100% |
| 1 headline pinned to position 1 | C(14,2) × 6 = 546 | 20% |
| 2 headlines pinned (positions 1 and 2) | 13 × 6 = 78 | 3% |
| 3 headlines and 2 descriptions pinned | 1 | 0.04% |
| 3 unique headlines pinned to each of 3 slots | 3 × 3 × 3 × 6 = 162 | 6% |
That last row is what Google recommends instead of hard pinning: pin two or three unique headlines or descriptions to each position. You keep control over what kind of copy occupies each slot while leaving the system 162 options instead of one. The documentation adds a specific warning: avoid pinning identical or very similar text to the same slot, because it restricts the system and can lower Ad Strength.
A pin is not “make this headline stronger”. A pin forbids every combination in which that headline is not in that position. Frame every pinning decision that way: what exactly are you banning, and why.
Pinning RSA headlines and Ad Strength
Google’s position is unambiguous: pinning is not recommended for most advertisers and can affect Ad Strength. Pinning a single keyword headline, though, is acceptable — what drags the score down is near-identical copy stacked into one slot.
The figures Google cites (footnoted as internal data, 15–20 August 2025 — a reference point, not a promise for your account):
- Moving Ad Strength from “Poor” to “Excellent”: on average 15% more clicks and conversions.
- Going from one RSA to two in an ad group: +6.6% conversions at a similar cost per conversion.
- From two to three: +3.7% conversions, again at similar cost.
The structural recommendation: at least two RSAs with “Good” or “Excellent” Ad Strength per ad group, each with a unique final URL. We explain how to read the asset report behind that score in Ad Strength and RSA asset reporting.
One caveat about the metric itself. Ad Strength is not an auction factor and not part of Quality Score; it is editorial guidance. Practitioners call it a vanity metric, and in individual cases that is fair — a “Poor” ad can outperform an “Excellent” one. But Google never says the score is irrelevant; the careful phrasing is that pinning “can affect” it. The sensible stance: do not break an ad to chase the score, and do not shrug off a drop from “Excellent” to “Poor” right after you pinned five slots.
Four cases where pinning RSA headlines is justified
Pinning is a tool for obligations, not preferences. The working list is short:
- Legally mandated wording. A regulatory line, age restriction, licence reference. Headline slots 1 and 2 or description slot 1 only. If the line is long and eats thirty characters of headline, there is a purpose-built alternative: text disclaimers in search ads, which do not steal space from the offer.
- A brand name that must come first. Common for franchises and accounts spanning several legal entities.
- Preventing a factually wrong combination. “Next-day delivery” may only be true alongside a description naming a specific region, and alone it becomes a false promise. Here a pin insures against a complaint, not against a CTR dip.
- A structured test. You are deliberately testing “offer in headline 1” against “brand in headline 1”. The pin is then an experimental condition, and you remove it when the test ends. How to set that up properly is in A/B testing RSAs.
Everything else — “I just think this headline is better” — is not a reason to pin. For “I think”, there is the combinations report.
Campaign-level text assets: a pin that behaves differently
A separate mechanism, routinely confused with ordinary pinning. At campaign level you can associate up to 3 headlines and 2 descriptions, and they apply to all enabled RSAs in that campaign.
The path: Campaigns icon → Assets drop-down → Assets → “Associations” table view → filter icon → “All asset types” or “Headlines or Descriptions” → Add → select campaigns → enter text → pin icon if needed → Advanced Options for scheduling → Save.
Rules for campaign-level pins
What to know before switching this on:
- Pinning is supported. And all assets pinned to the same position — campaign level and ad level alike — may be eligible to show in that slot.
- Ad Strength does not factor them in at all. The ad’s score will not move either way.
- Policy fallback: if a pinned campaign-level asset is not eligible for policy reasons, the system uses other creative and campaign-level assets, so the ad keeps running.
- Scheduling exists: start and end dates. “None” at the start means now; “None” at the end means indefinitely.
- Also, you cannot edit them. To change the text, remove and recreate. To change the pin or schedule, remove and recreate, or create a new asset with the same text and updated settings — it replaces the old one.
- They are invisible when creating RSAs at ad level, which regularly baffles whoever is editing the ad and cannot find where a stray line is coming from.
Reporting: assets live under the Associations tab (impressions, clicks, cost, conversions, conversion value); combinations live in the RSA under “View assets details” → Combinations tab. There is no ad-level asset reporting for campaign-level assets, and if a campaign-level asset carries the same text as an RSA-level asset in the same campaign, the two share stats.
The scenario where this beats a normal pin: a seasonal line that must appear across every ad in a campaign for two weeks. Schedule it once, leave Ad Strength untouched, edit zero RSAs.
Reading the combinations report instead of guessing
The combinations report is the strongest argument against “I know which headline works”.
- Open the RSA → “View assets details” → Combinations tab.
- Sort by impressions and read the top five rows. If the same headline sits in position one across all five, the system has already made your choice — pinning it adds nothing.
- Cross-check against the asset report. Asset-level ratings in the interface are Pending, Learning, Low, Good, Best, and a dash for unrated.
- If most assets show a dash, there is not enough data yet and any pinning decision is premature. Practitioners report the performance column appearing after roughly two thousand impressions in the top Search position over 30 days; treat that as an outside estimate, since Google does not publish the threshold.
The report almost always reveals the same thing: the system settled on three or four stable combinations a while ago, and the remaining eleven headlines exist as insurance for query drift. A pin removes exactly that insurance.
What the documentation does not say
- A hard cap on how many assets can be pinned to one slot. The “two or three” figure is a recommendation, not a limit.
- How pinned and unpinned assets are weighted when a combination is chosen.
- How many combinations an ad can produce. Not published; anything computed outside is arithmetic on limits.
- The Ad Strength formula or thresholds. Instead, Google gives only the level names.
- A separate asset minimum when pinning. The creation minimums of 3 headlines and 2 descriptions still apply.
A rollout plan for an existing account
| Day | Action | Output |
|---|---|---|
| 1 | Export every RSA, mark which pins exist and on which slots | A map of current restrictions |
| 2 | Write one line of justification per pin. No justification means candidate for removal | List of unjustified pins |
| 3 | Check mandatory wording: is all of it in headline 1, headline 2 or description 1? | Legal risk closed |
| 4–7 | Remove unjustified pins in half the ad groups, hold the other half as control | A clean comparison |
| 8–35 | Leave it alone. Compare CTR and cost per conversion between halves | An answer from your own data |
| 36 | Where unpinned ads did not lose, remove pins in the second half too | A decision, not a belief |
Four weeks of observation is not excessive caution: removing a pin effectively turns the ad into a new set of combinations, so you cannot trust the first week of data.
Common mistakes
Pinning mandatory copy to headline 3. The most expensive item on the list, because it creates the illusion that a requirement has been met. Headline 3 is not guaranteed.
Pinning all five slots. You have rebuilt an expanded text ad. Sometimes that is a deliberate choice; more often it is an accident, and twelve headlines are doing nothing.
Pinning three near-identical headlines to one slot. Google names this explicitly as a cause of lower Ad Strength: the system got a choice between three versions of the same thing.
Deciding by instinct instead of the combinations report. Five minutes in that report usually shows the system already picked the headline you were about to pin — without surrendering 80% of your combinations.
Forgetting campaign-level assets during a diagnosis. A line that appears in the SERP but not in the ad almost always comes from there.
Adjacent controls
Dynamic substitution is a different mechanism entirely: if you need a live price, city or countdown in the headline, that is not a pin but ad customizers on business data feeds. And when Performance Max or AI Max generates the copy, pins do not set the boundaries; instead, you set them with text guidelines for AI-written ad copy.
Zoomed out, pinning is the second of three control layers that stay with a human. The first is what you are allowed to show at all — Google Ads Policy Manager and handling disapprovals. The third is what any of it will cost — a Google Ads media plan built before the first click.
If ad copy is your main lever rather than a side task, the structured version of these rules sits in the Google Ads course, and the short reference version in the Google Ads guide.
FAQ
How many headlines and descriptions can an RSA hold?
Up to 15 headlines and 4 descriptions. The minimum at creation is 3 headlines and 2 descriptions. Headlines are 30 characters, descriptions 90, path fields 15.
Which positions can be pinned?
Five: headline 1, headline 2, headline 3, description 1 and description 2.
Does a pin guarantee the asset will show?
Only for headline 1, headline 2 and description 1, which always show. Pins on headline 3 and description 2 carry no guarantee.
What happens if I pin several headlines to one position?
They rotate within that position. Google recommends two or three unique options per slot and warns that identical or very similar text in one slot can lower Ad Strength.
Does pinning hurt Ad Strength?
Google says pinning is not recommended for most advertisers and can affect Ad Strength. Still, pinning a single keyword headline is acceptable.
What happens to unpinned headlines if every slot is pinned?
They never show. When all available positions are pinned, the ad excludes unpinned assets entirely.
Where should a mandatory legal line be pinned?
Headline 1, headline 2 or description 1 only — the three slots with a serving guarantee. For longer compliance copy there is a separate text disclaimer mechanism.
How do campaign-level assets differ from ad-level ones?
Up to 3 headlines and 2 descriptions apply to every enabled RSA in the campaign, support pinning and scheduling, but Ad Strength does not count them and you cannot edit them — you remove and recreate them.
Why does my ad show text that is not in the ad?
Usually a campaign-level asset, which is invisible at ad level. The other possibility is an unused headline from a different RSA in the same ad group serving as a link-based asset.
How many combinations does one RSA produce?
Google does not publish the number. Arithmetic on the documented limits (3 headlines from 15, 2 descriptions from 4) gives 2,730 — a calculated reference point, not an official metric.
How many RSAs should an ad group hold?
Google recommends at least two with “Good” or “Excellent” Ad Strength and a unique final URL each. Its internal data shows +6.6% conversions moving from one RSA to two and +3.7% from two to three — a reference point measured in August 2025.
How do I know whether a pin is needed at all?
Open the combinations report and sort by impressions. If the headline you had in mind already leads the top combinations, a pin adds nothing and costs you combinations.
