The budget allocation that is correct for your B2B SaaS Google Ads account is the one where spending one more dollar on any intent tier — brand, competitor, high-intent category, or retargeting — produces the same incremental pipeline value as spending it on any other tier. That is the entire model. Fixed-percentage splits like "15% brand, 50% non-brand, 25% retargeting" are useful as a starting point, but they are a template, not a decision: they ignore the fact that each tier has a different demand ceiling and hits diminishing returns at a different spend level. The job is to allocate to marginal return and rebalance as the margins move.
This post is about the allocation decision specifically — how to divide a capped budget across tiers and when to move money between them. It assumes you have already decided how much to spend in total (covered in how much you should spend on Google Ads) and how to structure the campaigns themselves (covered in B2B SaaS campaign structure by funnel intent tier). Allocation is the layer in between: a fixed pot, four tiers competing for it, and a method for deciding which tier gets the next dollar.
The default split — a starting point, not an answer
A reasonable default allocation for a B2B SaaS account spending $10K-$25K/month is roughly 15-20% brand, 45-55% high-intent category (non-brand), 10% competitor, and 20-25% retargeting and audience. These ranges are consistent with what agencies publish — one 2026 framework for the $10K-$25K band suggests roughly 15% brand, 45% non-brand high-intent, 25% retargeting, 10% competitive conquest, and 5% display prospecting, shifting toward conquest and prospecting as spend rises. Use a table like this to set the opening position, then expect to change it within 60 days.
| Intent tier | Opening share | Demand ceiling set by | Typical role |
|---|---|---|---|
| Brand | 15-20% | How many people search your name | Defensive; capture + protect existing demand |
| High-intent category | 45-55% | Auction depth on buy-intent terms | Primary pipeline engine |
| Competitor | 5-15% (capped test) | Rival brand search volume | In-market interception; scale only on closed-won |
| Retargeting / audience | 20-25% | Qualified site traffic volume | High-efficiency re-engagement |
The reason the opening split is only a starting point is in the third column. Each tier is capped by something you do not control. Brand spend is capped by how many people search your name — you cannot buy more brand demand with budget, only capture more of what exists. Category is capped by how deep the auction runs on buy-intent terms before you are bidding on marginal queries. Competitor is capped by rival search volume. Retargeting is capped by how many qualified visitors you can re-reach. A fixed percentage pretends these ceilings don't exist; the allocation method exists precisely to respect them.
Allocate to marginal return, not to the template
The operating rule is to equalize marginal cost per SQL across tiers. In plain terms: look at what the next dollar buys in each tier, and move budget toward whichever tier produces the cheapest incremental qualified pipeline until the tiers converge. A tier that is "budget-limited" — Google reports it losing impression share because of budget rather than rank — while its cost per SQL is at or below target is telling you the auction has more profitable demand than you are funding. That is the signal to add budget there, funded by pulling it from the tier with the highest marginal cost per SQL.
This is different from the common mistake of chasing the tier with the lowest average CPA. Average CPA always makes brand look best, because brand captures demand that already decided to buy. But the marginal brand dollar — the one spent after you already hold 90% impression share on your own name — buys almost nothing incremental, because those clicks would have reached you organically. Average return says "fund brand more"; marginal return says "brand is nearly maxed, move the money." Allocation decisions made on average CPA systematically overfund the tiers that look efficient only because they harvest existing demand. Decisions made on marginal cost per SQL fund the tiers that actually grow pipeline.
The brand over-funding trap
Brand is the single most over-allocated tier in B2B SaaS accounts, and the reason is a reporting illusion. Blend brand into your overall numbers and the account looks healthy: low CPA, high ROAS. Separate it out and you often find the brand campaign is inflating the blended figure while the non-brand tiers — the ones that actually generate net-new demand — look much weaker on their own. One widely-repeated agency observation is that brand ROAS can mask weak non-brand performance when the two are reported together, which is why brand should always be a separate campaign with separate reporting.
The correct brand budget is not a percentage of total spend — it is the amount needed to hold a high brand impression share (typically 80-95%) and defend against competitors bidding on your name. Once you are showing on nearly all your brand searches, every additional brand dollar is incremental-value-negative: you are paying for clicks you would have won for free in organic. Size brand to defensive coverage, confirm the impression share is where you want it, and move everything above that line to category or retargeting. If competitors are bidding on your brand, the defense is covered in our post on competitor brand bidding for B2B SaaS.
High-intent category: where growth actually comes from
High-intent category terms — demo, trial, pricing, "[problem] software," and buy-intent queries that do not mention your brand or a competitor — are the primary pipeline engine and should hold the largest single share of budget, usually 45-55%. This is the only tier that generates net-new demand at scale: it reaches buyers who have a problem and are looking for a solution but have not yet decided on a vendor. Everything else in the account either harvests existing demand (brand), intercepts a competitor's demand (competitor), or re-engages people you already reached (retargeting). Category is where you create new pipeline.
The allocation risk in category is diluting it across too many match types and intent levels at once. A dollar spent on a tightly-matched "demo" or "pricing" query is worth far more at the margin than a dollar spent on a broad informational term that happens to mention your category. Within the category budget, concentrate on the proven buy-intent terms first and expand to comparison and mid-intent terms only after the high-intent core is funded to its impression-share ceiling. The waste risk here is real: one analysis found B2B SaaS accounts lose around a third of spend to irrelevant search terms, which is why disciplined negative-keyword management is a prerequisite for scaling category spend — see negative keywords in Google Ads.
Retargeting: the tier teams starve
Retargeting is routinely the most efficient tier in the account and routinely the most underfunded. It re-engages prospects who already visited high-intent pages — pricing, demo, docs — so the audience is pre-qualified and the conversion rate is high. The reason it gets starved is that inside a blended campaign structure it competes for budget against prospecting and loses, and because its volume ceiling is low enough that teams assume it "can't absorb much." Both are allocation errors. Retargeting should have its own budget line and its own CPA target, and it should be funded to its ceiling — the point where you are reaching your full addressable audience of qualified recent visitors — before a single marginal dollar goes to higher-CPA prospecting.
Because the ceiling is set by site traffic, the way to grow retargeting is to grow the pool of qualified visitors, not to raise the budget beyond what the audience can absorb. This is why retargeting allocation is tied to the category and competitor tiers: they feed the audience that retargeting then converts cheaply. A useful way to think about it is that category spend buys first-touch demand and retargeting spend closes it at a fraction of the cost — which is the logic behind our guide to dynamic remarketing for SaaS.
The rebalancing cadence and the triggers
Review allocation monthly and rebalance on signal, not on noise. Monthly is the right cadence because B2B SaaS conversion events accumulate slowly — a tier might generate only 15-40 SQLs a month, so a week of data is too thin to tell a real shift from random variation. The concrete trigger to move budget is a tier reporting "lost impression share (budget)" above roughly 10-15% while its cost per SQL is at or below target. That combination means the auction has more profitable demand available than you are funding, and the money should come from the tier with the highest marginal cost per SQL — almost always brand (once defensive coverage is met) or an untested competitor expansion.
Every one of these decisions depends on measuring cost per SQL, not cost per form fill. If your conversion data stops at the lead and never connects to which leads became qualified pipeline in the CRM, your allocation will drift toward whichever tier produces the cheapest leads regardless of quality — usually broad category and competitor traffic that converts to form fills but not to revenue. Offline conversion data from the CRM is the prerequisite that makes marginal-return allocation possible; without it you are optimizing a budget split against the wrong number. The setup is covered in conversion tracking for SaaS, and the value-weighting that lets Smart Bidding allocate within a tier is in conversion value ladders and value-based bidding.
If you want a second opinion on where your current split is leaking — most commonly an over-funded brand campaign and a starved retargeting line — a structured Google Ads audit maps spend to pipeline by intent tier and shows which dollars are buying incremental SQLs and which are buying demand you already own; our ongoing Google Ads management runs this rebalancing monthly. For a fast self-check first, the free 10-point audit tool flags the most common allocation mistakes in a few minutes.