Most Google Ads guides tell B2B SaaS teams what to start doing. This one covers what to stop. Five tactics in wide use across B2B SaaS accounts in 2026 are not just suboptimal — they actively produce bad outcomes and get worse as accounts mature. Each one made reasonable sense under older platform behavior or when measurement infrastructure was simpler. Each one is now a structural drag on CAC that no amount of bid adjustment or creative testing will overcome.
The common thread is measurement quality. When the signal feeding Smart Bidding is wrong, the campaign learns the wrong patterns. When attribution windows do not match real buyer behavior, budget flows to the wrong campaigns. When campaign structure treats dissimilar intent levels as equivalent, the algorithm optimizes for the path of least resistance rather than the highest-value outcome. Stopping these five practices is not a configuration tweak — it is correcting the underlying inputs that determine what Google Ads is learning to do.
1. Optimizing toward form fills instead of qualified pipeline
The most costly habit in B2B SaaS Google Ads accounts is using web form submissions as the primary conversion event for Smart Bidding. It is also the most common: the Google tag fires on a thank-you page, the conversion records, and the campaign shows a cost-per-conversion that looks actionable. The problem is that the conversion being counted — a form submission — and the outcome you are paying to generate — a qualified sales opportunity — are separated by a gap that can be as wide as a 10:1 ratio.
When B2B SaaS companies import CRM-sourced downstream events — SQL, opportunity created, or demo attended — and feed those to Smart Bidding instead of raw form fills, the algorithm retrains on what qualified buyers actually look like: their query patterns, time of day, device, audience signals. Accounts that make this switch typically generate 3x more pipeline at 31% lower cost per lead compared to form-fill-optimized accounts, because the algorithm stops bidding for the patterns that generate cheap form fills and starts bidding for the patterns that generate buyers. The signal change is more impactful than any bidding strategy change. Our guide to offline conversion imports from your CRM covers the technical setup.
2. Running broad match without CRM conversion signals
Broad match in 2026 is not the undisciplined catch-all it was five years ago. Google has materially improved query understanding and the mechanism that decides how broad match expands. The expansion is anchored to your conversion history: the algorithm looks at which queries have historically produced conversions and expands toward semantically related queries where it expects similar conversion likelihood. This mechanism is effective, but it is only as good as the conversion signal that teaches it what "good" looks like.
The failure mode in B2B SaaS is running broad match with form-fill conversion history. The algorithm identifies that certain informational query patterns — "how to improve sales team productivity," "saas pricing models comparison" — generate high form-fill rates. It expands toward those patterns aggressively. Those queries are not purchase-intent queries; they are research-stage content queries that attract a broad audience including students, consultants, and competitors conducting market research. Broad match without offline CRM signals trains the algorithm to serve a wider audience of the wrong people. With CRM-sourced SQL signals, broad match expansion becomes a discovery mechanism for high-intent query variations your keyword list was not covering. The difference between these two outcomes is the quality of the signal, not the match type setting.
3. Applying the 30-day window to 90-day sales cycles
Google Ads default conversion window is 30 days — the period after a click during which a conversion is attributed to that click and credited to the campaign. For B2B SaaS companies with average sales cycles of 60–120 days, this default creates systematic attribution failure at the campaign level. Enterprise buyers who first encounter your brand through a category search keyword in week one are still in evaluation in week five. When they book a demo in week eight, the 30-day window means that click gets zero attribution credit. The demo conversion is attributed to a later retargeting touchpoint or a branded search immediately before booking.
Smart Bidding reads this as evidence that retargeting and branded campaigns are efficient converters while category campaigns are not. Budget shifts accordingly — toward retargeting and brand, away from the category keywords that initiated the buyer relationship. The higher-funnel campaigns that produced the most enterprise opportunities register as underperformers because their conversions fall outside the attribution window. Extending conversion windows to 60 or 90 days (Google allows up to 90 days for click-through attribution) restores the credited volume to category campaigns and corrects the budget allocation signal. Our post on matching conversion windows to B2B SaaS sales cycles covers the specific window lengths by sales cycle and the impact on attributed conversion counts.
4. Launching Performance Max before Search is profitable
Performance Max is an inventory-access product, not a conversion optimization product — it reaches Display, YouTube, Gmail, Maps, and Search inventory from a single campaign. Its budget allocation is guided by the conversion signals it receives, and in B2B SaaS accounts without high-quality offline conversion data, those signals are usually form fills from a heterogeneous audience. PMax with form-fill optimization in a B2B SaaS account will typically allocate heavily toward Display and YouTube because those channels generate high volumes of cheap engagement signals — video views, image ad clicks — that look like conversion precursors to an algorithm without downstream revenue data.
The practical result is a campaign that spends efficiently by its own metrics and inflates CAC by the metrics that matter. Benchmark data from 42 Agency across B2B accounts in 2026 shows Search campaigns delivering 553% ROAS versus 436% for Performance Max — a gap that narrows substantially when PMax has access to CRM-sourced conversion signals and audience assets. The readiness criterion for adding PMax to a B2B SaaS account is not "we've run Google Ads for six months." It is "our Search campaigns are profitable at our CAC target and we have consistent offline conversion volume that PMax can learn from." Without that foundation, PMax adds inventory reach to an account that has not established what a buyer looks like. Our analysis of when B2B SaaS should run Performance Max covers the specific readiness criteria.
5. Using one bid target across brand, competitor, and category
Brand campaigns, competitor campaigns, and category campaigns represent fundamentally different buyer intent levels. A brand campaign capturing clicks on your company name serves buyers who already know you — they are in a later evaluation stage, they have higher baseline purchase intent, and they convert at a disproportionately high rate. A category campaign targeting "project management software for remote teams" captures buyers who are problem-aware but have not shortlisted vendors yet — they are earlier in evaluation, they convert at a lower rate, and they represent a wider opportunity if your content and landing experience converts them.
Applying one tCPA or tROAS target across all three campaign types instructs the bidding algorithm to compare them on equal terms. Brand campaigns will hit any reasonable tCPA target easily because their baseline conversion rate is high. Category and competitor campaigns will appear to underperform against that same target because their conversion rates are lower by design. The algorithm responds by shifting budget toward brand — the path of least resistance to hitting the target — while reducing bid aggressiveness in category. The long-term effect is an account that wins on branded searches it would have converted anyway and gradually cedes category market share to competitors who understand the intent distinction. Setting separate tCPA or tROAS targets calibrated to the realistic conversion rate and buyer value at each intent level allows each campaign type to bid competitively for its respective audience without being penalized for the conversion-rate reality of upper-funnel traffic.
The underlying fix: signal quality before tactics
All five of these anti-patterns have the same root cause: the algorithm is being asked to optimize toward an objective that does not accurately represent what the business needs. Form fills instead of SQL events. Wrong attribution windows. Equal treatment of campaigns with unequal buyer intent. When the input is wrong, tactical adjustments — new ad copy, landing page tests, bid modifiers — optimize around a broken objective rather than toward the right one. Getting measurement right is not preliminary to running Google Ads well for B2B SaaS; it is the work.
The sequence that tends to produce the fastest CAC improvement: first, connect CRM offline conversion imports so downstream events are flowing to Google Ads. Second, extend conversion windows to match your median sales cycle. Third, assign conversion values to imported events that reflect deal potential by segment. Fourth, separate brand, competitor, and category campaigns with independent budgets and bid targets. Fifth, evaluate PMax only after Search campaigns are hitting CAC targets with offline data feeding them. That is not a quick checklist — it is a measurement infrastructure project. But each step is a one-time build that permanently improves signal quality, and the compounding effect on CAC over a 6–12 month horizon is typically larger than any creative or structural optimization running on a broken measurement foundation.