AI Brief is Google's new plain-language control layer for AI Max: a Gemini-powered interface that lets advertisers steer automated Search campaigns using written instructions across three areas — messaging, matching, and audience — instead of only toggles and keyword lists. Google announced it on April 30, 2026, and it is rolling out in English for AI Max for Search first, then Performance Max and AI Max for Shopping. For B2B SaaS teams, it is the most useful AI Max development of the year because it finally gives a way to encode strategy the automation had previously been guessing at.
The reason this matters is specific to how B2B SaaS differs from the e-commerce accounts AI Max was largely built around. A SaaS advertiser does not want the widest possible query expansion; it wants a narrow band of high-intent buyer searches, a consultative tone rather than a promotional one, and messaging that speaks to a VP of Engineering differently than to a student researching a free tool. AI Brief is the first native mechanism to express all three. This guide covers what it is, how to write each guideline block for a software business, and how to confirm it is steering toward pipeline rather than cheap conversions.
What Google's AI Brief actually is
AI Brief sits inside AI Max for Search and accepts natural-language instructions that guide how the system generates assets and expands queries. It is organized into three guideline types. Messaging guidelines tell the system what your ads should and should not say. Matching guidelines set boundaries on which searches to capture or avoid. Audience guidelines describe how to tailor the message to particular segments. Google's own launch examples are deliberately simple: messaging can "tell AI Brief exactly what your ads should — and shouldn't — say, e.g., never mention prices", while matching can prioritize or exclude whole categories of query.
Two structural details matter for anyone already running AI Max. First, your existing text guidelines migrate automatically into the messaging guidelines block — you do not rebuild them. Second, before you commit, Google shows previews of sample generated assets and the searches the system expects to match, so you can catch a misfire before it spends. AI Brief is not a new campaign type and it does not replace your conversion goals, bidding strategy, or exclusions; it is a guidance layer that rides on top of the AI Max controls covered in our guide to the newer AI Max controls for B2B SaaS.
Why AI Brief matters more for B2B SaaS than for retail
AI Max's core behavior is query expansion: it takes your existing keywords and finds adjacent searches it believes are relevant. For a retailer selling a well-understood product, broad expansion often finds incremental demand. For B2B SaaS, the same behavior is the primary way budget leaks. Software categories are riddled with queries that look topically related but carry no buying intent — "free CRM template," "is HubSpot worth it reddit," "marketing analyst jobs," "how to learn SQL." Without a constraint, AI Max will happily expand into all of them and report conversions when low-intent visitors fill a form.
AI Brief's matching guidelines are the first native lever to bias that expansion toward buyers before the spend happens, rather than cleaning it up afterward with negative keywords. The messaging guidelines matter almost as much: B2B buyers respond to specificity and credibility, not discount language, and a documented rule such as "lead with outcomes and integration depth; never imply the product is free" keeps generated assets on-brand. This is the same failure the community keeps surfacing — automation optimizing to the cheapest conversion — and it is why we argue you should be optimizing for SQLs, not raw leads before you hand more control to any AI system.
Writing messaging guidelines for a software product
Messaging guidelines govern what generated headlines and descriptions say and avoid saying. Treat this block as your brand's copy policy written for a capable but literal assistant. Be concrete about prohibitions and priorities rather than vague about tone. "Sound professional" gives the model nothing to act on; "never claim the product is free or offer a discount; never name competitors; always describe the product as a platform, not a tool" gives it enforceable boundaries. If your product has compliance-sensitive claims — security certifications, uptime figures, regulated-industry language — spell out exactly which claims are permitted and how they must be phrased.
For B2B SaaS, the highest-value messaging rules encode differentiation and accuracy. State the two or three value propositions you want emphasized (for example, time-to-value, a specific integration, or a measurable outcome) and the things that are factually off-limits because they misrepresent the product post-click. A message-to-landing-page mismatch is one of the most common causes of wasted spend, and AI Brief does not fix your landing page — so pair the messaging rules with a pass over the destination using our landing page audit checklist. The goal is that a generated ad and the page it points to describe the same product to the same buyer.
Matching guidelines: the SaaS budget guardrail
Matching guidelines are where B2B SaaS advertisers should spend the most care, because they control the queries AI Max is willing to serve on. Write them as an explicit include-and-exclude policy in plain language. On the include side, name the buyer-intent patterns you want prioritized: category terms with commercial modifiers ("software," "platform," "pricing," "alternative," "for [industry]"), competitor comparison queries, and problem-aware searches that map to your product's core job.
On the exclude side, describe the categories that consistently waste B2B SaaS budget: free and template seekers, job and career queries, educational and how-to-learn searches, and consumer-grade intent for a business product. Matching guidelines are directional, not absolute, so keep your hard negative keyword lists and audience exclusions running underneath them as a backstop — AI Brief steers, negatives enforce. Confirm the two are working together by reading the search terms report weekly for the first month; if buyer-intent share is not rising, tighten the language. This is the same discipline behind migrating DSA campaigns to AI Max, where controlling expansion is the whole game.
Audience guidelines: talking to buyers vs. practitioners
Audience guidelines let you vary the message by segment, which maps neatly onto the multi-stakeholder reality of B2B SaaS purchases. The same product is evaluated by an economic buyer who cares about ROI and risk, a technical evaluator who cares about integration and security, and an end-user who cares about workflow. Write guidelines that tailor emphasis accordingly — for example, "for senior and executive audiences, emphasize ROI, security, and time-to-value; for practitioner audiences, emphasize workflow and integrations." Google's launch framing uses exactly this pattern of segment-specific emphasis.
Be realistic about what audience guidelines can and cannot do. They influence how AI Max talks to segments it can identify; they do not replace your first-party audience signals or your bidding strategy. For B2B SaaS, the strongest audience inputs remain your CRM-derived lists — customer-match lists of closed-won accounts, high-fit segments, and exclusions of existing customers and churned accounts. Feed those in as signals and let audience guidelines shape the copy on top. The combination of accurate audience data plus message-tailoring guidance is far more powerful than either alone.
Rollout, previews, and what to do before it reaches your account
AI Brief is rolling out gradually, so the practical stance today is preparation. Announced on April 30, 2026, it is arriving first in English for AI Max for Search, with Performance Max and Shopping to follow over subsequent months. Because access lands account-by-account, waiting is wasted time. Write the three guideline blocks now as a plain document: your messaging do's and don'ts, your query include/exclude policy, and your segment-specific emphasis. When AI Brief appears, that document becomes the brief you paste in, and you launch with a considered strategy instead of improvising in the interface.
When you do configure it, use the preview step deliberately. Google shows sample generated assets and expected searches before you commit — review those previews against your SQL definition and your negative lists, not just for tone. If the previewed searches include the free-tool and job-seeker patterns you meant to exclude, your matching guidelines are too loose; revise before launch rather than after the first billing cycle. Treat the first version of the brief as a hypothesis to be validated by data, in the same evidence-based way we recommend for evaluating Google Ads AI recommendations.
How to know the brief is steering toward pipeline
The only verdict that matters is downstream revenue quality. Because AI Brief changes which queries you expand into and how you speak to them, the risk is a campaign that looks better on cost per conversion while getting worse on cost per SQL. Establish a baseline before you apply the brief: cost per SQL, SQL-to-opportunity rate, and the buyer-intent share of your search terms. After the change, hold those same metrics up against the baseline over a 4–6 week window with enough conversion volume to be meaningful, since AI Max re-learns after significant guidance changes.
If cost per SQL falls and buyer-intent query share rises, the brief is doing its job — tighten it incrementally from there. If cost per form fill drops while cost per SQL climbs, the automation is optimizing toward cheaper, lower-quality conversions and your matching or conversion-goal setup needs correction before the brief can help. AI Brief is a genuine step forward in making AI Max controllable for B2B SaaS, but it rewards advertisers who already measure the right outcome. Get your conversion signal pointed at revenue first, then let plain-language guidance do the steering.