Spread 60 monthly conversions across 12 campaigns and each one gets five. Smart Bidding can't distinguish a real trend from noise at that volume, so it overcorrects bids on every fluctuation. Google's own guidance says calibration after a structural change can take around 50 conversion events or three full conversion cycles (Google Ads Help). For a B2B SaaS company with a 30-to-60-day sales cycle, that means weeks or months of unstable performance, and legacy account structures make it worse by splitting that already-thin signal into fragments the algorithm can't use.
The Problem Isn't Smart Bidding. It's What You're Feeding It.
Single Keyword Ad Groups, device splits, match-type isolation: these were rational when humans adjusted bids line by line. Smart Bidding needs conversion density at the campaign level to evaluate contextual signals (device, location, time of day, intent). Google's Search Ads product director Brandon Ervin has noted that learning can aggregate conversions tied to the same conversion tracking ID, and he's suggested shared budgets or portfolio bidding when campaigns share the same conversion action and targets (Ads Decoded podcast). Splitting campaigns that share identical goals fragments the data Smart Bidding needs to learn.
Three bottlenecks show up predictably in hyper-fragmented accounts:
- RSA testing stalls. Responsive Search Ads need continuous impression and conversion data to rotate headline combinations. Low-traffic ad groups never generate enough signal for the creative engine to find winners.
- Cross-negation blocks good queries. Dense negative keyword lists built to enforce rigid keyword isolation suppress high-converting long-tail searches. The algorithm matches on intent now, not exact strings.
- Budget gets trapped. Fixed daily budgets on micro-campaigns create silos. A high-intent traffic surge caps out a $20/day campaign by noon while unspent budget sits idle in a quiet segment next door.
Consolidation Has Limits. Respect Them.
Google's own stance, as reported by Search Engine Journal, frames consolidation as appropriate when it removes redundant segmentation, but retaining splits that reflect real differences in budget, business goals, margins, or customer experience. Flattening everything into one campaign is a different kind of mistake.
Merging high-margin and low-margin product lines into a single Target ROAS forces the algorithm toward an average that underserves your best inventory. For B2B SaaS specifically, a demo-request campaign (high friction, longer cycle) and a free-trial campaign (low friction, faster close) need separate structures because the conversion economics and funnel stages are fundamentally different.
The decision rule: if splitting a campaign changes your budget ownership, margin target, conversion goal, or landing experience, keep it separate. If the split only changes how identical customer journeys get labeled in reports, consolidate.
Run It This Week: A Three-Step Audit
Step 1: Map every campaign to a business reason. Pull your campaign list. For each one, write down the budget owner, the conversion action, and the margin or CPA target. If two campaigns share all three, they're candidates for merging. If you can't articulate why they're separate beyond "we set it up that way in 2021," that's your answer.
Step 2: Check conversion density. Google recommends at least 30 conversions for evaluating Smart Bidding performance, and 50 for Target ROAS (Google Ads Help). Anything consistently below 15 per month is almost certainly starving the algorithm. Brandon Ervin's cited guidance of roughly 15 conversions over 30 days is a floor, not a target.
Step 3: Consolidate one product line first. Don't restructure the whole account at once. Pick a single product line or campaign group, merge the redundant splits, and benchmark stability over one to two conversion cycles (not a fixed seven-day window, per Google's bidding FAQ). Sephora's vendor-published case study reported a 42% increase in campaign conversion rate and 13% higher ROAS after reducing campaign count by 85%, but that's a retail example with high volume. Your B2B numbers will look different. Track qualified pipeline, not just form fills.
The hypothesis: if we merge campaigns that share the same conversion action and CPA target, then conversion volume per campaign will increase and CPA volatility will decrease, because the algorithm gets enough signal to distinguish real trends from noise.
Success = CPA variance drops by 20%+ over two conversion cycles. Guardrails = monitor impression share and average position weekly to catch audience suppression. Stop-loss = if CPA rises more than 30% above baseline after two full conversion cycles, revert and diagnose.
The Trade-Off You're Accepting
Consolidation means giving up granular reporting segmentation. You'll lose the ability to see exact-match vs. broad-match performance in separate campaign rows. That feels like losing control. What you're actually losing is the illusion of control: those thin campaigns never had enough data to produce reliable performance reads. Five conversions in a month is a coin flip, not a signal.
The harder adjustment is organizational. Teams built workflows, dashboards, and QBR decks around the old structure. Changing the account means changing how people report, and that's a conversation with RevOps and finance before it's a conversation in Google Ads. Structure your account around business truth, not legacy reporting convenience, and give Smart Bidding the density it needs to do the math.
