Responsive Search Ads: Best Practices for B2B

Responsive search ads are the only standard search format Google still offers. Expanded text ads were retired in 2022, so if you run Search campaigns today, you run RSAs. That removed a choice and added a problem: the format that gives you the most clicks is not automatically the format that gives you the most pipeline.

Here is the tension for B2B. An RSA mixes up to 15 headlines and 4 descriptions into thousands of combinations, optimized by Google's machine learning. The default optimization target is engagement, usually clicks. A consulting firm or a SaaS company does not need more clicks. It needs the right CFO, the right ops director, the right buying committee member to read the ad, recognize their problem, and fill out a form that a sales team will actually want to call.

This guide covers how to build RSAs that do that: how to write and structure the assets, when to pin and when to let Google rotate, how to read the reporting Google gives you, and how to keep low-quality clicks out. The examples use illustrative numbers, but the methods are the ones that move qualified-lead volume.

How a responsive search ad actually works

You supply a pool of assets: up to 15 headlines (30 characters each) and up to 4 descriptions (90 characters each). Google assembles them into ads, showing up to 3 headlines and 2 descriptions per impression, and learns which combinations perform for each auction. Over time it serves the mixes it predicts will hit your campaign goal.

Two things follow from this that matter for B2B.

First, you are not writing one ad. You are writing a system of parts that have to make sense in any order. Headline 2 might appear next to headline 9, or headline 9 might not appear at all. Every asset has to stand on its own and still read cleanly when combined.

Second, the machine optimizes toward whatever you tell the campaign to value. Feed it a conversion action of "form submitted" and it chases form fills, including the junk ones. Feed it a qualified-lead or pipeline signal through offline conversion import, and it learns to find the people who become opportunities. The format is only as smart as the goal behind it.

Write headlines that qualify, not just attract

In B2C, a broad headline that pulls everyone is often fine. In B2B, a headline that pulls everyone pulls the wrong people. Your job is to write assets that make the right buyer lean in and the wrong one scroll past.

A practical way to fill the 15 headline slots is to cover distinct angles rather than rephrasing one idea 15 times. Google needs variety to test; you need each variant to carry a real message. Group your headlines like this:

Match the query. Include the keyword or a close variant. If someone searches "B2B lead generation agency," a headline with those words signals relevance and helps Ad Strength and Quality Score.
Name the audience. "For B2B SaaS Teams," "Built for Manufacturers." Specificity filters. A small business owner searching a generic term will self-select out, which is what you want when your minimum deal size is $20k.
State the outcome. "Cut Cost Per Lead 30%" or "More Qualified Demos." Mark internal targets as illustrative, never promise a guaranteed result.
Show proof. "Trusted by 200+ B2B Firms," "G2 Leader 2026." Use only numbers you can back up.
Reduce friction. "Free 15-Min Strategy Call," "No Long Contracts."
Handle the objection. "Senior Team, No Juniors," "Setup in 2 Weeks."

A B2B buyer rarely converts on the first click. The headline's job at the top of the funnel is to qualify and create recognition, not to close. Getting this right is the same discipline as writing the rest of your search copy; if you want the full framework, our guide on how to write Google Ads copy that converts goes deeper on message-to-intent matching.

Pinning: control without killing the test

Pinning locks a specific asset to a specific position. Pin a headline to position 1 and it always shows first. This is the single most debated RSA setting, and for B2B the answer is "use it sparingly and on purpose."

When pinning earns its place:

Compliance and legal copy. Regulated industries (finance, healthcare, legal) sometimes need a required phrase or disclaimer in a fixed spot.
Brand or offer integrity. If a headline must always appear, for example a product name that cannot be omitted, pin it.
Protecting message order. If headline 1 is the hook and headline 3 is the proof, and they only make sense in that sequence, pin both.

The cost of pinning: every asset you pin removes combinations Google can test, which shrinks the learning and can lower Ad Strength. Pin all 15 headlines to fixed slots and you have rebuilt an expanded text ad, throwing away the format's main advantage.

A balanced approach that works well in B2B: pin one or two assets to position 1 (your strongest qualifier or brand line) and leave the rest unpinned. Note that you can pin two or three headlines to the same position; Google will rotate among them in that slot, which keeps some testing alive while guaranteeing one of your chosen messages always leads.

RSA pinning approaches (illustrative trade-offs)
Approach	Testing freedom	Message control	Best for
No pinning	Maximum	Low	Most top-of-funnel B2B campaigns
Pin 1 to 2 headlines to position 1	High	Medium	Brand or lead qualifier must lead
Pin multiple positions	Low	High	Compliance, strict message order
Pin everything	None	Total	Rare: legal mandates only

Read Ad Strength for what it is

Google scores each RSA from Poor to Excellent based on asset quantity, variety, and keyword relevance. Treat it as a checklist, not a grade.

Ad Strength rewards having many distinct assets, including keywords, and avoiding duplicate phrasing. Those are reasonable hygiene signals. What it does not measure is whether your ad attracts qualified buyers or burns budget on tire-kickers. An "Excellent" ad can produce expensive, unqualified leads, and a "Good" ad with tightly written B2B copy can outperform it on pipeline.

So: push for Good or Excellent because the underlying behaviors (variety, relevance, enough assets) are genuinely useful. Do not sacrifice a sharp, qualifying message just to turn the meter green. If adding a generic broad-appeal headline lifts Ad Strength but invites the wrong audience, skip it.

Build the asset library, then test it

Strong RSAs come from strong inputs. A few practices that compound over time.

Mine your own data for headlines. Your search terms report, your old expanded text ads, your top-converting landing page copy, and your sales team's language all hold phrases that resonate. The words prospects actually use in queries belong in your headlines, which is why RSA performance and keyword research are connected: the same intent mapping that builds your keyword list should feed your headline pool.

Use asset reporting. Google labels each asset Low, Good, or Best based on relative performance. Once you have enough data, pause the Low assets and replace them with new variants on the same theme. This is your test loop. Replace a few assets at a time so you can attribute changes, and give each change a couple of weeks before judging it.

Run at least two RSAs per ad group. One RSA per ad group is the minimum, but a second one, with a meaningfully different angle, gives you a real comparison and protects you if one ad underdelivers. Keep them distinct, not near-duplicates.

Keep descriptions doing real work. Headlines hook; descriptions expand. Use the 90 characters to add the proof, the differentiator, and a clear next step. Restate the audience and the outcome so any headline-description pairing holds together.

Connect RSAs to qualified-lead signals

This is where most B2B accounts leave money on the table. If your conversion action is a raw form fill, Google's automation will optimize toward volume, and a chunk of that volume will be students, job seekers, vendors, and out-of-profile companies.

The fix is to feed the algorithm a better goal. Import offline conversions from your CRM so Google learns from "became an MQL" or "became an opportunity," not just "submitted a form." Once the machine optimizes toward leads that your sales team accepts, the RSA combinations it favors shift toward the assets that attract those buyers. The ad copy and the bidding work together; if you are pairing RSAs with automated bidding, our piece on Smart Bidding in Google Ads explains how to train it on the right signal.

Equally important is keeping the wrong traffic out before it ever clicks. A disciplined negative keyword list (blocking "jobs," "salary," "free," "course," "intern," and consumer-grade modifiers) protects the budget your RSAs spend and keeps the conversion data clean enough for the algorithm to learn from.

Measure RSAs by pipeline, not by CTR

A high click-through rate feels like success and often is not. The metrics that matter for a B2B RSA, in rough order:

Qualified leads (MQL or SQL), not raw conversions. The only volume number worth optimizing.
Cost per qualified lead. CTR and CPC are inputs; this is closer to the outcome.
Lead-to-opportunity rate by ad. Which messages bring people sales actually wants to talk to.
Conversion rate. Useful, but only against a clean conversion action.
CTR and Ad Strength. Diagnostic signals, not goals.

Give RSAs time before you judge them. The format needs impressions to learn, and B2B volume is often low. Pausing an ad after a few hundred clicks usually means acting on noise. Look at four to six weeks of data and enough conversions to mean something.

Common B2B mistakes with RSAs

A short list of what burns budget:

Over-pinning into an expanded text ad. You lose the testing the format exists to do.
15 headlines that say one thing. No variety means nothing to learn from.
Chasing Ad Strength over message quality. Green meter, wrong audience.
Optimizing to form fills instead of qualified leads. The single most common cause of high-volume, low-quality B2B accounts.
Generic, swap-anywhere copy. A landing page promising "great service" matched to a vague ad converts no one. Tight ad-to-page message match still matters.
Judging too soon. Low B2B volume plus early pausing equals decisions made on noise.

FAQ

How many headlines should a B2B responsive search ad have?

Fill all 15 if you can, but only with distinct messages. Variety is what the algorithm tests, so 15 sharp, different headlines beat 15 rephrasings of one. If you genuinely have fewer real angles, a smaller set of strong assets is better than padding with filler.

Should I pin headlines in my RSAs?

Sparingly. Pin only when a message must appear or must lead: a brand name, a compliance phrase, or a qualifier your buyer needs to see first. Pinning every asset rebuilds the old expanded text ad and discards the format's main advantage, so leave most assets unpinned.

Does Ad Strength affect performance?

Ad Strength measures asset variety and relevance, not whether your ad attracts qualified buyers. Aim for Good or Excellent because the behaviors behind the score are useful, but never trade a sharp, qualifying message for a higher rating. A focused "Good" ad can beat a generic "Excellent" one on pipeline.

How long before I can judge an RSA?

Give it four to six weeks and a meaningful number of conversions. B2B search volume is usually low, so a few hundred clicks rarely tells you anything reliable. Acting earlier means optimizing on noise.

Can I still use expanded text ads?

No. Google retired expanded text ads in June 2022. You can no longer create or edit them; responsive search ads are the standard search format. The skill now is writing RSA assets that behave the way you want.

What is the single biggest RSA lever for B2B?

The conversion goal you optimize toward. Feed Google raw form fills and it chases volume; feed it qualified-lead or pipeline signals through offline conversion import and the asset combinations it favors shift toward the buyers you want. Copy matters, but the signal behind it decides what the machine learns.

Conclusion: a quick checklist

Responsive search ads reward accounts that treat them as a qualification system, not a click machine. Before your next campaign goes live, run through this:

15 distinct headlines covering query match, audience, outcome, proof, friction, and objection.
Pinning used only where a message must lead or must appear.
Ad Strength at Good or better, without sacrificing a sharp message.
At least two distinct RSAs per ad group.
Offline conversions imported so Google optimizes to qualified leads.
A negative keyword list keeping unqualified traffic out.
Performance judged on cost per qualified lead, over four to six weeks.

Most B2B accounts get the copy roughly right and the signal completely wrong, then wonder why more clicks never turned into more deals. If you want a second set of eyes on your Search campaigns, we offer a focused 30-minute audit of your RSAs and conversion tracking, where we look at your assets, your pinning, and whether Google is actually optimizing toward leads your sales team wants. Bring your account and we will show you where the budget is leaking.