How to fix: Imbalanced budget allocation across funnel stages

TL;DR

When the Google Ads budget is concentrated almost entirely at the bottom of the funnel — brand Search, retargeting, exact-match harvest — top- and mid-funnel campaigns receive no spend and the pipeline of in-market users that eventually convert stops being refilled. The damage does not show in week one. It shows 60-180 days later as branded search demand flattens and the remarketing pool ages out. The fix is to rebalance toward a 50/30/20 TOFU/MOFU/BOFU split for accounts in growth mode, or 30/30/40 for steady-state harvest accounts, and to do it in two passes so Smart Bidding does not all re-enter learning at once [1].

Why it matters

The funnel exists in Google Ads whether the account structure names it or not. Every paying customer was once a stranger who never heard of the brand. Top-of-funnel (TOFU) and mid-of-funnel (MOFU) campaigns are what move that stranger into the pool of in-market searchers that bottom-of-funnel (BOFU) campaigns then harvest. When 90 percent of spend sits in brand Search and retargeting, the harvest works beautifully on month one and looks fantastic in the dashboard. By month three the harvest yield starts dropping because no fresh demand is entering the pool, and the account manager rebalances toward "what's working", which is the BOFU surfaces, and the spiral tightens.

The mechanism is straightforward. BOFU campaigns convert at high rates (brand Search converts at 5-15 percent on most accounts; retargeting at 3-8 percent) and produce low CPAs because the user is already aware and intent-rich. TOFU campaigns convert at much lower rates (0.3-1.5 percent for Demand Gen and broad-match prospecting) and produce CPAs 3-8x higher than BOFU. Last-click attribution makes TOFU look unprofitable per click. But last-click attribution misses the entire causal chain: the user who clicked a Demand Gen ad in March, never converted there, then googled the brand directly in May and converted on brand Search is attributed entirely to brand Search. The TOFU campaign that planted the seed gets credit for nothing in the column the operator looks at.

Smart Bidding amplifies this. Target CPA and Maximize Conversion Value on BOFU campaigns will happily spend every available dollar because every dollar converts efficiently inside the small pool of brand-aware users. The bidding algorithm has no view of long-term funnel health — it optimizes the immediate auction. So if the operator lets BOFU campaigns claim spend on their own merits, they will, and the account silently strips itself of TOFU and MOFU [2].

The financial impact is measurable but lagged. On audited accounts with severe BOFU concentration (BOFU at 80 percent or more of total spend), pipeline metrics flatline at the 90-180 day mark: branded search impressions stop growing, new-customer revenue share drops, and total revenue plateaus despite spend rising. The fix is not philosophical — it is to set a deliberate budget split per funnel stage and pace each stage independently.

The two baselines that work for most accounts:

• Growth mode (active customer acquisition push): 50% TOFU / 30% MOFU / 20% BOFU. The job is filling the funnel; BOFU still gets enough to harvest existing demand but does not crowd out prospecting.
• Steady state (mature account, established brand): 30% TOFU / 30% MOFU / 40% BOFU. The brand-aware pool refreshes naturally; the priority is efficient harvest with continued top-of-funnel maintenance.

These are starting baselines, not laws. The right split depends on sales cycle, brand maturity, and channel mix. But 0% TOFU is almost always wrong, regardless of phase.

How to fix

1. Tag every enabled campaign with a funnel stage — TOFU (Demand Gen, YouTube Reach, broad-match Search prospecting, PMax without first-party seed lists, top-of-funnel display), MOFU (non-brand Search exact/phrase, Shopping non-brand, PMax with audience signals, in-market display), or BOFU (brand Search, branded Shopping, retargeting display, Customer Match remarketing). Use a campaign-label convention so the segmentation persists (Campaigns → Labels → New label → tofu/mofu/bofu).
2. Pull 30-day cost per stage (Campaigns view → Filter by Label → Sum of Cost) and compute the percentage split. Anything more extreme than 70/20/10 in any direction signals imbalance.
3. Compare against the baseline for your growth phase. Growth-mode accounts target 50/30/20; steady-state target 30/30/40. If BOFU is above 60% of spend on a growth-mode account, the funnel is starved. If TOFU is above 60% on a steady-state account with stable demand, you are over-investing in awareness.
4. Diagnose the missing stage. The most common failure pattern is zero TOFU spend — no Demand Gen, no YouTube prospecting, no broad-match Search experimentation. The second most common is zero MOFU — heavy brand spend plus generic display, with no qualified non-brand Search in between. The diagnosis determines which campaign type to add.
5. Rebalance in two passes, not one large edit. The first pass moves 10-15% of total spend from the overfunded stage to the underfunded stage. Wait 14 days for Smart Bidding to re-stabilize, then a second pass moves another 10-15% if the first pass did not produce drift in core metrics. Large single-pass rebalances trigger simultaneous learning phases on multiple campaigns and break attribution baselines for the next 21-30 days. Budget edits live at the campaign level (Campaign → Settings → Budget); shared budgets at Tools → Shared library → Shared budgets if multiple campaigns share a pool.
6. Set stage-appropriate bidding goals. TOFU campaigns should not optimize toward the macro Primary (purchase, MQL) — there are not enough conversions per TOFU campaign per 30 days to clear the Smart Bidding learning gate of 30/30. Optimize TOFU toward a Secondary engagement action (qualified visit, video 75% view, scroll depth) using Maximize Conversions, or use Target CPM for pure Reach campaigns. MOFU optimizes toward a mid-funnel Secondary (form start, demo request, add-to-cart). Only BOFU optimizes directly to the macro Primary. The funnel-stage rule and [[fix-funnel-coverage-gaps]] both depend on this Primary/Secondary separation.
7. Re-measure at 30, 60, and 90 days. At 30 days, branded Search impression volume should be rising (the new TOFU and MOFU traffic feeds back into brand demand). At 60 days, MOFU conversion rate should hold inside its pre-rebalance band. At 90 days, downstream pipeline metrics (form starts, MQLs, new-customer revenue) should show measurable lift. If TOFU spend produces zero downstream pipeline at 60 days, the channel mix is wrong rather than the split — revisit with the campaign-mix-for-niche rule.

Common mistakes

• Defending the imbalance with last-click ROAS data. "BOFU has the best ROAS so it should get the budget" is the trap. Last-click ROAS over-credits the closest touch and structurally undervalues TOFU. Switch the attribution model to Data-Driven Attribution before making spend decisions based on column comparisons; see [[fix-attribution-last-click]].
• Rebalancing everything in one push. Moving 30% of total spend across stages in a single edit triggers simultaneous Smart Bidding learning phases on five to ten campaigns and contaminates the next month of data. Two passes, 14 days apart, is the standard.
• Treating PMax as a single funnel stage. PMax spans the funnel by design. PMax with strong first-party audience signals plus product feeds behaves closer to MOFU/BOFU. PMax with no audience signals and broad asset groups behaves closer to TOFU. Label by configuration, not by campaign type.
• Calling brand Search "free demand" that doesn't need budget management. Brand Search has the best ROAS in nearly every account and will still happily absorb 50% of total spend if allowed. Cap brand Search at the level that captures available branded query volume; spend beyond that goes to BOFU surfaces with diminishing returns. See [[fix-brand-vs-nonbrand-separated]] for the separation pattern.
• Forgetting that TOFU needs different conversion tracking. A TOFU campaign optimizing toward a Primary purchase conversion will spend forever and never clear learning. TOFU needs an engagement-level Secondary action to optimize against, and the operator must add it first. The article on funnel coverage gaps owns this setup.

FAQ

Where do the 50/30/20 and 30/30/40 baselines come from?
They are the Whitead audit baselines synthesized from 2024-2025 account reviews across B2B SaaS, ecom, and lead-gen verticals. The exact numbers shift by vertical (B2B SaaS often runs 60/25/15 in growth mode due to longer education cycles; ecom runs closer to 40/30/30), but the principle — every stage gets meaningful spend, BOFU does not eat the budget — holds across verticals.

What if my BOFU campaigns are massively profitable and TOFU has never worked?
Two diagnoses are possible. First, the TOFU campaigns may have been set up to optimize toward the wrong conversion (macro Primary instead of an engagement-level Secondary) and were never given a fair signal. Fix the conversion setup before deciding the channel doesn't work. Second, the BOFU ROAS is real but it is harvesting a finite pool. Track branded Search impression growth month-over-month: if it is flat or declining, the pool is shrinking and TOFU investment is the structural fix.

How does this interact with Performance Max?
PMax is a special case because it spans the funnel internally. The rule still applies at the account level — the total PMax budget plus the total TOFU/MOFU/BOFU Search and Shopping budgets needs to be classified. Audience signals plus product feeds typically push PMax toward MOFU/BOFU classification; broad PMax with no signals is closer to TOFU. The [[fix-pmax-conversion-goals]] rule owns the within-PMax tuning.

Does this apply to accounts under $5K/month?
Less mechanically. Small accounts often cannot run TOFU campaigns that clear Smart Bidding learning gates and have to concentrate on BOFU and MOFU. The 50/30/20 baseline is realistic above roughly $15-20K/month. Below that, expect 70/30 MOFU/BOFU as a working baseline and treat TOFU as a separate quarterly experiment, not an always-on allocation.

How do I detect this on an audit?
The audit rule pulls 30-day cost per campaign, classifies by campaign type and naming convention, and flags accounts where any stage is below 10% of total spend or any stage is above 60%. A medium-severity finding fires when the split is more than 20 percentage points off the appropriate baseline; high severity when one stage receives zero spend.

Sources

1. Google Ads Help — About average daily budgets. Defines how daily budgets pace and how the 2x daily / 30.4x monthly cap mechanics work; the procedural basis for moving budget between campaigns without overspend surprises.
2. Google Ads Help — Budgets and bidding. Documents how campaign budgets interact with bidding strategies and how shared budgets distribute across campaigns; the reference for the rebalance-in-two-passes workflow.
3. Google Ads Help — About Performance Planner. Explains the Performance Planner tool for forecasting how budget changes affect conversions across Search, Shopping, PMax, Demand Gen, and App campaigns; the modeling surface for previewing a rebalance before deploying it.