Why Consolidating Google Ads Campaigns Can Lower Lead Volume But Improve Lead Quality (B2B SaaS Case Study)

I want to walk you through an account I still manage today, because it doesn't end the way case studies usually do.

There's no clean before-and-after chart. There's a real constraint I couldn't consolidate my way around, three things I tried that failed, and a client conversation still happening right now that hasn't resolved.

If you run paid media for B2B SaaS, this is probably closer to your reality than the polished version.

The account

Multi-product B2B SaaS, $800K+ in annual spend across Google, Microsoft, and LinkedIn. Long sales cycles, high deal values, CPA that sits between $1,200 and $2,500 depending on the campaign. Several distinct product lines, each with its own budget owner who wants to see spend against leads for their specific line.

That last detail matters more than it sounds like it should. I'll come back to it.

What I inherited

When I took over, the account had 13 campaigns, most of them built around single-keyword ad groups. That structure made sense for control and reporting, but it split conversion volume into slivers too thin for the algorithm to learn from.

On top of that, informational keywords were pulling in traffic that technically converted to an MQL but never had any business qualifying as one.

So the first job wasn't optimization. It was cleanup.

Consolidating, and hitting a real floor

I merged campaigns down from 13 to 8 and cleared out the informational keyword bloat. The disqualified MQL rate dropped noticeably. That part worked the way it's supposed to.

Then I hit a wall. I couldn't consolidate further, because the business genuinely needs spend and leads reported separately by product line. That's not a tracking failure I can engineer around. It's a legitimate reporting requirement, and it means this account will never pool conversion volume the way a single-product account could.

Eight campaigns is the floor here, not a milestone on the way to fewer.

Three things I tried to pool signal anyway

I didn't stop at consolidation. I tried three more ways to get more conversion volume flowing to fewer, smarter decision points.

Shared budgets. Dropped this one first. It made it impossible to control spend per business line, which defeated the entire reason the campaigns were separated in the first place.

Shared bidding strategies. This one was more interesting. Some campaigns, competitor terms especially, brought in MQLs at a much higher cost per lead, but the quality justified it. Once I pooled bidding, the algorithm quietly suppressed spend on those expensive-but-valuable campaigns in favor of the cheaper ones. I was optimizing myself out of the leads that mattered most. Dropped it.

Target CPA. Also dropped, in favor of Maximize Conversions. This is the one change that actually moved the needle. Conversion volume improved once I stopped asking the algorithm to hit a fixed cost target and let it prioritize volume within budget instead.

The test that didn't work

At one point I tried optimizing purely toward MQLs instead of blending MQLs and leads as the primary conversion signal. The logic was sound on paper: give the algorithm a cleaner, higher-value target.

In practice, the campaigns underspent, lead volume tanked, and CPA spiked. I reverted to the blended target within weeks. Sometimes the theoretically cleaner setup starves the algorithm of the data volume it needs to function at all.

What actually improved

Here's the part that's easy to miss if you're only watching lead volume. Qualified-to-lead rate moved from roughly 30-32% in the two years before the cleanup to around 56% so far this year. Volume is down. Quality, as a rate, is meaningfully better. Worth noting 2026 is a partial year, so I'm treating this as a directional signal, not a finished result.

That's the part that doesn't show up if the only metric on the dashboard is lead count.

The takeaway

Not every structural limit is a mistake to fix. Sometimes it's a legitimate business requirement, and the job is optimizing well inside it, not engineering it away. If your account looks worse on the surface after a cleanup, check what happened to quality before you assume something went wrong.

There's a bigger point underneath all of this, though. Difficult accounts don't have an easy fix. I tried shared budgets, shared bidding, a cleaner conversion target, and only one of those four changes actually worked. That's normal. Theory doesn't always survive contact with a real account, and the only way to find out is to test it and give it time to show a result.

What actually moved this account wasn't a clever technique. It was consistent management: cleaning the keyword set, watching what the algorithm did when I changed one variable at a time, reverting fast when something didn't work. And none of that happens without genuinely growing into the account, understanding the product, the competitors, the sales motion, the reason a POC deal is different from a self-serve signup.

That kind of understanding only builds if the relationship is a real partnership.

A brand that treats its PPC provider as a vendor to be handed a target and left alone gets surface-level execution.

A brand willing to share what's happening on their side, like this client did with the POC churn context, gets a partner who can actually connect the ad account to what's happening downstream.

This account has stayed hard. But it's stayed honest, and that's what's kept it improving instead of stalling.