Performance Max Single-Conversion-Signal Steering: Why the 2026 LSA Migration Makes It Worse (and How to Fix It)

The Short Answer

Performance Max single conversion signal steering happens when a campaign optimizes bidding toward one blended conversion signal instead of separate targets per vertical or value tier, so low-value or junk conversions drag the whole Target CPA down. Starting August 1, 2026, Google's Local Services Ads migration removes vertical-level Target CPA, pushing multi-vertical advertisers straight into this trap.

This is not the same steering Google announced in its 2026 "steering and reporting update," which covers audience exclusions and reporting, not bidding-signal weighting (business.google.com). The August 1, 2026 date comes from Google's official community announcement (Google Ads Community, Aug 2026); Google's Help Center article frames the same milestone more generally as "August 2026" (Google Ads Help, 2026).

What "Single Conversion Signal Steering" Actually Means

Performance Max is a goal-based Google Ads campaign type that uses Smart Bidding to automate bidding, budget, creative combinations, and audience targeting toward whatever conversion goal it is given (Google Ads Help). Google's own description stops there: "Performance Max drives performance based on your specified conversion goals, and delivers more conversions and value" (Google Ads Help, 2026). It never covers what happens when several structurally different conversion types feed into the same goal.

Single conversion signal steering describes that gap. When a plumbing lead, an HVAC lead, and a spam form fill all count as the same conversion inside one Target CPA, Smart Bidding cannot tell them apart and optimizes toward the blended average, not the outcome that matters. This differs from the tracking-accuracy question covered in the 7-lever PMax optimization playbook: tracking accuracy asks whether a conversion is recorded correctly; this asks whether structurally different conversion types get blended into one target at all.

As @everdenepublic put it: "Performance Max has exactly one steering input: the conversion signal you send it" (X, Aug 19 2026). Everything sent into that one input counts as equally valuable whether it is or not.

Google's own narrowing mechanism seeds the fix later in this article: "To specify which conversions are most valuable to your business, apply conversion values and set value rules" (Google Ads Help). Most PMax accounts never configure this, which is why blended-signal risk shows up as unexplained ROAS volatility rather than a visible setting. See how attribution models train Smart Bidding for the credit-allocation side of the same training data.

The 2026 LSA Migration Is About to Multiply This Problem

Local Services Ads is Google's pay-per-lead local advertising format for home-service and storefront businesses, verified through a Google Guaranteed or Google Screened badge. Starting August 1, 2026, Google begins migrating these accounts into Performance Max campaigns with pay-per-lead goals (Google Ads Community, Aug 2026); the Help Center article frames the same milestone more generally as "August 2026" (Google Ads Help, 2026).

Three phases. Phase 1, August 1, 2026: select US home and storefront trades, introduced with the qualifier "including": plumbing, HVAC, electrical, appliance repair, house cleaning, lawn care, roofing, pest control, moving (Google Ads Help, 2026). Search Engine Land describes the same rollout more broadly, as pet care, home services, wellness, and education (Search Engine Land, Jul 2026); Google's nine trades all fall under Search Engine Land's "home services" label, but its other three categories are not confirmed on Google's own page. Phase 2, late 2026: service-area businesses without a storefront, plus custom bidding and booking setups. Phase 3, 2027: non-US accounts and remaining categories.

The connecting fact, verbatim: "Google Ads doesn't support vertical-level Target CPA bidding. If you previously managed different Target CPA amounts for separate verticals under a single campaign, a unified campaign-level Target CPA will be calculated and applied to all vertical categories" (Google Ads Help, 2026). An agency running Plumbing and HVAC under separate Target CPAs loses that separation on migration day unless it splits first: "you can achieve this by setting up separate campaigns for each vertical" (same source).

Manual bidding is no longer supported. Weekly budgets convert to a daily average (divided by 7), and the monthly cap equals the daily average times 30.4. BBB callouts are retired; accounts need at least 6 other structured callouts instead (all three, Google Ads Help, 2026). Search Engine Journal sums the impact: "The confirmed changes cluster in the four areas advertisers watch most: budgets, bidding, historical reporting, and day-to-day campaign management" (Katie Morton, Search Engine Journal, Aug 2026).

What changes and what does not in the August 2026 LSA-to-PMax migration. Source: Google Ads Help, answer 17213585 (2026), unless noted in the cell.

This Is a Different "Steering" Than Google's 2026 Update

Google uses "steering" for something else entirely here. Its own 2026 "steering and reporting" update lets advertisers exclude specific audiences and see deeper reporting breakdowns. It says nothing about how a blended conversion signal gets weighted inside Smart Bidding, which is the problem this article is about. The two threads share a word, not a mechanic.

Business.google.com describes its update this way: "Now, we're launching new updates designed to help you reach your business goals by providing more ways to steer Google AI and see where your budget goes" (business.google.com, modified May 2026). Four items only: first-party audience exclusions (already-converted customers), a budget-projection report, demographic audience reporting, and network segmentation in placement reporting. None touches how conversion types get weighted before a bid is placed; excluding a customer list is audience-targeting control, not signal-weighting control. No specific rollout date beyond "2026" is stated; treat anything more precise as unverified.

Who Is Actually Exposed (a Decision-Grid, Not a Panic List)

Not every PMax account carries blended-signal risk equally. Call it the PMax Exposure Grid: a four-tier read, from highest (forced by a fixed migration date) to low (structurally not possible).

The PMax Exposure Grid. Source for the top row: Google Ads Help, answer 17213585 (2026), which states the unified campaign-level Target CPA. The lower three rows are structural patterns, not cited figures.

Row one is the only migration-triggered tier, and its only defense is splitting into per-vertical campaigns before the date. The middle two are chronic rather than dated: a $20 SKU and a $400 SKU share one Target ROAS unless value rules separate them, and high-intent demos blended with low-intent downloads in one conversion action dilute lead quality inside the same target. That second case is corroborated by why Performance Max fails in B2B marketing. The bottom tier has one conversion type and one value band, so there is nothing structurally different to blend.

For agencies running a mixed book of clients, the grid above is a punch list: run every multi-vertical local-service account against row one before August 1, 2026, and let the rest wait for a scheduled review.

How to Diagnose Whether Your Own PMax Campaign Has This Problem

Blended-signal risk applies to any Performance Max campaign, migration or not. Four checks surface it without waiting for a deadline.

1. Check whether one conversion action counts structurally different outcomes as equal value, a sales call and a newsletter signup, say.
2. If value-based bidding is active, look at the spread of conversion values inside a single campaign for a bimodal distribution, two humps instead of one.
3. Open the Insights tab and look for category-level performance hidden inside one blended total.
4. Confirm whether conversion value rules are actually configured, or quietly left at Google's default equal-weighting.

Performance Max automates bidding, budget allocation, creative-asset combinations, and audience targeting at once, all steered by whichever conversion signal it is fed. That is why signal quality matters more here than in manually-controlled campaign types: automation amplifies whatever target it is given.

Negative keywords cannot fix any of the four checks above. Performance Max negative keywords block specific search queries; they do not repair broken tracking or a blended signal, because that problem lives in what counts as a conversion, not which queries trigger one.

The Fix: Split Signal Before Google Splits It For You

Splitting the signal cleanly is not complicated. It just has to happen on purpose. Three moves get you there, in order of how much control each hands back.

1. Split campaigns by vertical or value tier. Google's own migration FAQ says it plainly: "you can achieve this by setting up separate campaigns for each vertical" (Google Ads Help, 2026). This takes about an afternoon to plan and is the highest-impact move on the list, because it removes the blend at the source instead of working around it.
2. Configure conversion value rules. Google names the mechanism directly: "apply conversion values and set value rules" (Google Ads Help). It is a low-effort change, worth doing even when a full campaign split is not practical yet, though on its own it only moderates the problem rather than removing it.
3. Correct signal after the fact with enhanced conversions and offline conversion adjustments. This one is never really finished. It needs a CRM or lead-status feed feeding back continuously, and it works best paired with the two moves above, not as a standalone fix.

Google's general guidance recommends consolidating campaigns where possible (Google Ads Help), but that advice predates this migration's vertical-Target-CPA problem; the migration FAQ overrides it with a narrower fix: split by vertical when one blended Target CPA would otherwise absorb structurally different conversion types.

For the mechanics behind why a clean signal changes bidding behavior this fast, see how Target CPA and Target ROAS actually work.

A Pre-Migration Checklist for LSA Advertisers

That is the general playbook. LSA advertisers are working against a fixed date, so here are five steps, each tied to a fact already sourced above.

1. 1

   Export historical LSA performance reports before migration day

   They do not transfer automatically; download or screenshot through the redirection page first (Google Ads Help, 2026).
2. 2

   Identify every vertical under one LSA account and decide split-versus-merge before the advance notice arrives

   Admins get a notification 14 days before migration, plus a final reminder 7 days after (Google Ads Help, 2026).
3. 3

   Recalculate budgets for the weekly-to-daily conversion

   Daily average = weekly divided by 7; monthly cap = daily average times 30.4 (same source).
4. 4

   Set or confirm conversion value rules per vertical before the unified Target CPA takes over

   Same mechanism Google names for narrowing PMax's conversion signal generally (Google Ads Help).
5. 5

   Replace BBB callouts with at least 6 structured callouts

   Business hours, specialties, payment methods, all in place ahead of the cutover.

What This Costs If You Do Nothing

Blended-signal accounts do not fail visibly on day one. They drift. For an agency owner, that drift becomes an end-of-month call explaining why cost-per-lead crept up with no clean story. Nothing broke. The account just optimized toward the wrong average. @TronWatch3r's framing is direct: "Performance Max isn't underperforming. It's doing exactly what your conversion data told it to do. That's the problem" (X, May 26 2026).

Other practitioners describe the same drift. @LeeCaston2: "20-40% of budget disappears inside PMax, broad match, 'smart' bidding" (X, Apr 14 2026). @Aria_Nawi: "ROAS dropped overnight across 4 campaigns" (X, Aug 4 2026, 247 engagements). None of that is a controlled study, individually attributed sentiment is not data, but it lines up closely enough with the sourced mechanics above to take seriously.

This is a structurally correct setup gone wrong, not a bug, and that distinction matters for where the fix actually goes. If tracking itself is broken, the 90-second conversion tracking triage is the right first stop. If tracking is accurate but the campaign still blends unlike conversions, the 9-step Performance Max conversion troubleshooter covers the tactical symptoms; this article covers the architecture underneath. Results vary by account size and vertical mix, so treat any number above as directional, not a guarantee.

Frequently Asked Questions

What is Performance Max in Google Ads?

A goal-based Google Ads campaign type that uses Smart Bidding to automate bidding, budget, creative combinations, and audience targeting toward specified conversion goals (Google Ads Help). It runs across Search, Display, YouTube, Discover, Gmail, and Maps from one campaign.

What are three benefits of a Performance Max campaign?

Cross-channel reach from a single campaign, automated bid, budget and creative optimization, and unified reporting (Google Ads Help). The tradeoff is less manual control over how each conversion type gets weighted.

What do Performance Max ads look like?

Standard PMax assembles assets automatically across Search, Display, YouTube, Discover, Gmail, and Maps. The pay-per-lead variant from the LSA migration is narrower: Search and Maps only (Google Ads Help, 2026).

What is automatically optimized with a Performance Max campaign?

Bidding, budget allocation, creative-asset combinations, and audience targeting, all driven by whichever conversion signal the campaign is fed. A blended signal degrades all four at once.

Where kampaio Fits

Signal-quality monitoring and per-vertical campaign structuring are unglamorous account hygiene that decays without a dedicated watch. Inside kampaio, Buzz (bid strategy) and Echo (reporting) surface blended-signal risk and value-rule drift automatically, before a migration date or a slow ROAS bleed forces the issue. Neither agent auto-splits a campaign or auto-migrates an LSA account without approval; they flag exposure and hand you the decision. This hygiene sits inside the complete Google Ads optimization framework. See Buzz and Echo in the pricing plans.

Migration dates, budget arithmetic and bidding changes are cited to Google Ads Help (answers 17213585 and 10724817) and to Google's Ads Community announcement (thread 456909801), all verified on August 27, 2026. Mascot figures are illustrative, not client results. Practitioner posts on X are individual sentiment, not measured data. This article is informational and does not constitute professional advice.