Account Structure

This page is updated every two months with current best practices for Google Ads account structure. In the Smart Bidding and AI era the old hyper-granular playbook can starve the algorithm of data, so how you consolidate campaigns and feed conversions now drives performance. Each update draws on our own experience plus authoritative industry sources and verified real-time research. Bookmark this page and check back for the latest account structure best practices. Each update includes worked examples with the arithmetic shown.

Last updated: 6 August 2026

In This Guide

  1. Executive Summary
  2. Benchmarks & Numbers at a Glance
  3. Structure in the Smart Bidding Era
  4. Consolidation vs Granularity
  5. Organising by Intent & Product
  6. Campaign Types Together
  7. Ad Group Structure
  8. Naming & Labels
  9. Conversion Data Density
  10. Account Hygiene
  11. Common Mistakes to Avoid
  12. What Changed Recently
  13. References

Reading this to build your own campaign? Skip the learning curve — we’ll build the whole thing for you for $129, delivered within 24 hours as a spreadsheet you own. See how it works →

1. Executive Summary

Google Ads account structure in August 2026 is fundamentally a signal architecture problem, not a manual bid-management exercise. The five principles every senior digital marketer should internalise are:

  • Consolidate around business objectives. Fewer, well-funded campaigns give Smart Bidding the conversion volume it needs to learn and stabilise. Fragmentation is now the primary structural risk, not insufficient granularity.[7]
  • Separate only when economics genuinely differ. Brand versus non-brand, distinct margin tiers, materially different target CPA or ROAS, or hard budget isolation requirements are the legitimate reasons to split campaigns. Match type, device, and geography alone are not.[6][10]
  • Use themed ad groups, not single-keyword ad groups (SKAGs). Google’s current guidance explicitly recommends moving away from SKAG-style structures toward tightly themed ad groups with 5–20 closely related keywords, up to 3 Responsive Search Ads each.[1][6]
  • Feed automation with clean signals. Conversion action hygiene, verified enhanced conversions, Consent Mode v2 compliance, and accurate conversion values matter more in 2026 than adding structural layers. Automation is only as good as the goal you hand it.[9][13]
  • Treat naming conventions and labels as operational infrastructure. A consistent, metadata-rich naming scheme and a disciplined label taxonomy let you manage, report, and audit without creating structural splits that starve campaigns of data.[11]

2. Benchmarks and Numbers at a Glance

Metric Typical range or threshold Applies when Source
Minimum conversions per campaign for Smart Bidding to function at all 15 conversions per month Vendor claim; applies to any Smart Bidding strategy before trusting its decisions [1]
Minimum conversions per campaign for stable tCPA performance 30 conversions per 30-day trailing window Vendor claim; threshold before tCPA bids are considered reliable [52][63]
Minimum conversions per campaign for stable tROAS performance 50 conversions per 30-day trailing window Vendor claim; higher bar required because tROAS optimises value, not just volume [52]
Conversions before expanding with an additional campaign layer 30 conversions in trailing 30 days Vendor claim; applies to DTC ecommerce and lead-gen accounts considering horizontal expansion [64]
Recommended keywords per ad group 5–20 closely related keywords (5–10 per tighter guidance; 10–20 per broader guidance) Vendor claim; both figures are practical heuristics, not hard Google requirements. Where sources conflict, use 5–10 as the more conservative default. [20][45]
Recommended ad groups per campaign 7–10 (general); 3–8 (B2B Search) Vendor claim; lower end applies to B2B accounts with fewer high-intent themes [43][45]
Maximum enabled RSAs per ad group 3 RSAs per ad group Vendor claim; Google Ads platform limit for active RSAs in a single ad group [45]
Restructure trigger: keyword density More than 50 keywords per ad group Vendor claim; ad groups exceeding this threshold typically suffer poor ad relevance and mixed intent [61]
Restructure trigger: campaign count More than 10 active campaigns Vendor claim; accounts above this count should audit for data fragmentation before adding further campaigns [61]
Search CTR benchmark 3%–5% Vendor claim; cross-industry average for standard Search campaigns [62]
Search CPC benchmark USD $2–$4 (approx. AUD $3.10–$6.20 at 0.64 exchange rate) Vendor claim; cross-industry average; highly variable by vertical and geo [62]
Search CPA benchmark USD $50–$80 (approx. AUD $78–$125) Vendor claim; cross-industry average; use as a directional reference only [62]
Conversion rate benchmark 3%–5% Vendor claim; cross-industry Search average [62]
ROAS benchmark 200%–400% Vendor claim; cross-industry average for ecommerce Search and Shopping [62]
Quality Score benchmark 5–7 out of 10 Vendor claim; scores below 5 indicate ad relevance or landing page issues worth investigating [62]
Impression share benchmark 60%–80% Vendor claim; accounts with impression share below 60% should investigate budget or bid constraints before restructuring [62]
Quality Score and CPC effect of better structure Quality Score up 2–3 points; CPC down 15%–25% Vendor claim; directional illustration only; no study-grade sample size provided [61]
Typical campaign count for accounts under USD $50,000/month 1 brand Search + 1–2 non-brand Search + 1 Performance Max or Standard Shopping + 1 prospecting = 4–5 campaigns Vendor claim; applies to single-market accounts with a single product or service line [52]

3. Account Structure in the Smart Bidding Era

The fundamental role of account structure has shifted. Before automated bidding, structure was the primary mechanism for controlling bids: you split by match type, device, geography, and keyword to set different CPCs at each level. In 2026, Smart Bidding evaluates hundreds of auction-time signals — device, location, time, audience membership, query context — and adjusts bids in real time without requiring structural separation to do so.[7][10] Structure still matters enormously, but its job is now to direct automation with clean signals, appropriate budgets, and unambiguous conversion goals.

Google’s own account-structure guidance makes this explicit: automated bidding does not require splitting campaigns or ad groups by keyword match type, geo, or device, and recommends pairing broad match keywords with Smart Bidding so Google AI can find additional relevant queries to meet conversion goals.[6][10] The practical implication is that every structural decision should be tested against a simple question: does this split give the business a meaningful control point — different budget, different bid target, different conversion action — or does it merely fragment data? If it is the latter, consolidate.[7]

Four structural variables still matter in the Smart Bidding era:

  • Campaign budget boundaries determine which pool of conversion data informs each bidding strategy. Too many campaigns means each pool is too small to stabilise.[1][7]
  • Primary conversion action assignment tells Smart Bidding what to optimise toward. Mixing high-value and low-value conversions in the same primary action set misleads the algorithm.[3][13]
  • Campaign-level target settings (tCPA, tROAS, or Maximise Conversions) set the economic objective. Separate campaigns are warranted only when those targets genuinely differ.[2][3]
  • Ad group theme quality affects Quality Score components — expected click-through rate, ad relevance, and landing page experience — which in turn affect the cost at which Smart Bidding wins auctions.[6][14]

Worked example

Consolidating a fragmented Search account to feed Smart Bidding

  • Setup: A Melbourne B2B software account spending AUD $18,000 a month across 14 active Search campaigns, each organised by a different keyword match type (exact, phrase, broad) for the same 5 service themes. No campaign reaches 15 conversions per month; the highest records 11.
  • Numbers: 14 campaigns × average 9 conversions per month = 126 total monthly conversions. Consolidated into 5 theme-based campaigns (1 per service): 126 ÷ 5 = 25.2 conversions per campaign per month — still below the 30-conversion tCPA stability threshold[52][63] but above the 15-conversion minimum for Smart Bidding to function.[1] A further merge to 3 campaigns (top 3 revenue-generating themes) yields 126 ÷ 3 = 42 conversions per campaign per month, crossing the 30-conversion tCPA threshold.[52]
  • Decision: Consolidate 14 campaigns into 3 theme-based campaigns; set bidding strategy to Maximise Conversions (no target) for a 4-week learning period before applying tCPA targets.
  • Why: No campaign can reach the 15-conversion minimum for Smart Bidding to function[1] while data is split across 14 campaigns; consolidation is the prerequisite for any automated bidding strategy to stabilise.

4. Consolidation vs Granularity

The 2026 consensus is clear: consolidate by default, and preserve granularity only where it creates a genuine business control point.[1][4][7] This is a meaningful departure from the legacy approach of splitting campaigns to enable manual bid adjustments at every level. The risk of over-consolidation — losing reporting visibility or blending incompatible business objectives — is real but manageable with naming conventions and labels. The risk of over-fragmentation — starving campaigns of the conversion volume Smart Bidding needs — is both more common and harder to recover from.[1][7]

The following table summarises the legitimate reasons to maintain granularity versus the situations where consolidation is the better default:

Dimension Keep separate — legitimate reason Consolidate — fragmentation risk
Brand vs non-brand Always separate; different economics, different auction dynamics, different conversion rates[3][10] Never merge brand into non-brand Search
Match type Almost never a reason to split in 2026[6][10] Merge match types into the same themed ad group under Smart Bidding
Device Only if the business genuinely cannot serve mobile users (e.g., a desktop-only booking tool)[6] Use device bid adjustments rather than separate campaigns
Geography Different language, materially different CPA target, separate budget requirement, or distinct serviceability[2][10][14] Use location targeting within one campaign when economics are similar
Margin tier or product line Different tCPA or tROAS targets; separate budget priorities[2][3] Merge product lines with similar economics into one campaign
Funnel stage Informational vs transactional queries need different bids, copy, and landing pages[2][14][16] Do not create a separate campaign for every funnel micro-stage
Performance Max vs Search Always separate; PMax operates across channels and requires distinct budget allocation and asset groups[3][12] Use campaign-level negatives and distinct objectives to prevent overlap

A useful consolidation audit starts with the question: if I merged these two campaigns today, would I lose a real control lever — budget, bid target, conversion action, or compliance boundary — or only a reporting filter? If only a reporting filter, merge and use a label instead.[11][12]

Worked example

Auditing a 12-campaign account for consolidation candidates

  • Setup: A Brisbane home services account spending AUD $9,000 a month runs 12 Search campaigns: 4 match-type splits (exact, phrase, broad, BMM-legacy) × 3 service lines (plumbing, electrical, HVAC). All 12 share a single tCPA target of AUD $85 and the same conversion action (phone call ≥ 60 seconds).
  • Numbers: AUD $9,000 ÷ 12 campaigns = AUD $750 average monthly budget per campaign. At a benchmark CPA of AUD $85[62], each campaign generates approximately $750 ÷ $85 = 8.8 conversions per month — well below the 15-conversion minimum for Smart Bidding to function.[1] Merged to 3 service-line campaigns: $9,000 ÷ 3 = $3,000 per campaign; $3,000 ÷ $85 = 35.3 conversions per campaign per month, crossing the 30-conversion tCPA stability threshold.[52][63]
  • Decision: Consolidate 12 campaigns into 3 service-line campaigns (Plumbing Search, Electrical Search, HVAC Search); combine all match types within each themed ad group; retain tCPA = AUD $85 on each.
  • Why: The match-type splits share identical economics and provide no budget or bid-target control lever, making them a fragmentation risk with no offsetting benefit under Smart Bidding.[6][10]

Worked example

Deciding when granularity is justified — high-margin vs standard product line

  • Setup: A Sydney e-commerce retailer selling both premium coffee machines (average order value AUD $1,200, target ROAS 350%) and consumable coffee pods (average order value AUD $45, target ROAS 600%) currently runs both product lines in a single Shopping campaign with a blended tROAS of 450%.
  • Numbers: At a blended tROAS of 450%, the algorithm bids to return AUD $4.50 for every $1 spent. For machines: $1,200 × 350% target = $1,200 revenue goal; acceptable spend = $1,200 ÷ 3.5 = AUD $342.86 per conversion. For pods: $45 × 600% target = $45 revenue goal; acceptable spend = $45 ÷ 6.0 = AUD $7.50 per conversion. The blended 450% target overspends on pods ($45 ÷ 4.5 = $10 vs the $7.50 ceiling) and underbids on machines ($1,200 ÷ 4.5 = $266.67 vs the $342.86 ceiling).
  • Decision: Split into two campaigns — Premium Machines (tROAS = 350%) and Coffee Pods (tROAS = 600%) — with separate budgets proportional to revenue contribution.
  • Why: Materially different tROAS targets (350% vs 600%) mean the economics cannot be blended without systematically misdirecting Smart Bidding on both product lines.[2][3]

5. Organising by Intent, Product and Geography

The most durable structural principle in 2026 is to organise campaigns around what the user intends to do and what the business can profitably deliver to them, rather than around keyword mechanics.[2][14][16] Three primary organising dimensions apply to most accounts: intent/funnel stage, product or service line, and geography.

Intent and funnel stage

Informational queries (e.g., “how to fix a leaking tap”), commercial comparison queries (e.g., “best emergency plumber Sydney reviews”), and high-intent transactional queries (e.g., “emergency plumber Sydney book now”) require different bids, different ad copy, and different landing pages. Separating these into distinct campaigns or at minimum clearly distinct ad groups within a campaign lets Smart Bidding learn the conversion rate of each intent tier without averaging them into noise.[2][14]

Product or service line

Each major product or service offering should have its own campaign or tightly aligned campaign set when budgets, bid targets, or landing pages differ materially. This prevents Smart Bidding from allocating budget away from a high-margin line toward a high-volume but lower-value line simply because the latter generates more raw conversions.[14][18]

Geography

Campaign-level geographic separation is warranted when locations have materially different CPAs, different languages, different serviceability, different pricing, or different store coverage.[10][14] For accounts serving a single country or metro area with uniform economics, use location targeting within one campaign rather than geographic campaign splits — the latter fragments data unnecessarily.[6]

Worked example

Organising a national vs local campaign split for a franchise network

  • Setup: An Australian franchise network of 22 tyre and auto service centres spending AUD $55,000 a month on Search. Locations in metro Sydney and Melbourne deliver a CPA of AUD $62 for a booking conversion. Regional Queensland and Western Australia locations deliver a CPA of AUD $110 for the same conversion. A single national campaign currently runs with a tCPA of AUD $80.
  • Numbers: At tCPA = AUD $80, Smart Bidding systematically underbids metro (fair price AUD $62, set ceiling AUD $80 — leaving headroom) but overspends in regional markets ($80 target vs $110 actual cost, meaning the campaign loses money on every regional booking). Regional locations represent 8 of 22 centres = 36% of locations but are generating losses on every conversion.
  • Decision: Split into two campaigns — Metro Search (Sydney + Melbourne, tCPA = AUD $62) and Regional Search (QLD + WA, tCPA = AUD $95 as a test ceiling, with a review at 30 conversions); assign budgets of AUD $38,500 and AUD $16,500 respectively.
  • Why: A CPA differential of AUD $48 between geo segments (AUD $62 vs AUD $110) is material enough that a blended tCPA misdirects Smart Bidding in both directions, warranting separate campaign economics.[2][10]

Worked example

Structuring intent tiers for a B2B SaaS lead-gen account

  • Setup: A Perth B2B SaaS account selling project management software at AUD $299/month per seat spends AUD $22,000 a month on Search. Currently all queries — from “what is project management software” (informational) to “buy project management software Australia” (transactional) — run in a single campaign with Maximise Conversions optimising to a free-trial sign-up.
  • Numbers: The account records 80 free-trial sign-ups per month. Analysis of search terms shows 35% of spend is on informational queries (approx. AUD $7,700/month) generating 8 sign-ups (CPA = AUD $962.50). Transactional queries account for 65% of spend (AUD $14,300/month) and 72 sign-ups (CPA = AUD $198.61). The blended CPA is AUD $275. Separating intent tiers allows setting tCPA = AUD $220 on transactional (72 conversions/month, above the 30-conversion stability threshold[52]) and a separate Awareness campaign for informational traffic with no CPA target.
  • Decision: Split into two campaigns — Transactional Search (tCPA = AUD $220, budget AUD $14,300) and Informational/Awareness Search (Maximise Clicks, budget AUD $7,700 reduced to AUD $4,000 with $3,700 reallocated to the transactional campaign).
  • Why: An intent-tier CPA spread of AUD $763.89 ($962.50 vs $198.61) means informational traffic is consuming 35% of budget to produce only 10% of conversions, and the blended CPA obscures this from Smart Bidding.[2][14]

6. Campaign Types and How They Fit Together

In 2026, most accounts will run a combination of Search, Performance Max (PMax), and in some cases Standard Shopping or Display campaigns. Understanding how these campaign types interact — and how to structure them so they complement rather than cannibalise each other — is one of the most important structural decisions a practitioner makes.[3][12]

Search campaigns

Search remains the most controllable campaign type and the right home for intent-driven keyword targeting. Brand Search should always be isolated in its own campaign to protect budget, measure brand conversion rates accurately, and prevent non-brand spend from subsidising brand traffic.[3][10] Non-brand Search should be organised by business objective, product line, or funnel stage — not by match type.[6][7]

Performance Max

PMax runs across Search, Shopping, Display, YouTube, Gmail, and Maps from a single campaign. Its structural implications are significant: it can cannibalise non-brand Search traffic if not managed with campaign-level negative keywords and clear asset group differentiation.[3][12] Best practice in 2026 is to segment PMax campaigns by distinct product lines, margin tiers, or business objectives — not to fragment them into many small campaigns — and to use brand exclusions or Search campaign brand terms to prevent PMax from capturing brand queries where you want precise control.[2][3][12]

Standard Shopping

Standard Shopping campaigns remain useful when retailers need granular product-level bidding control that PMax’s asset-group structure does not yet provide. For most accounts under AUD $50,000 per month, a single well-structured PMax campaign with segmented asset groups is preferred over maintaining parallel Standard Shopping campaigns that fragment conversion data.[52]

How campaign types fit together

Campaign type Primary role Key structural rule Cannibalisation risk
Brand Search Defend brand queries; measure brand conversion rate Always separate from non-brand; use exact and phrase match Low if isolated; add brand terms as negatives in PMax
Non-brand Search Capture intent-driven queries; drive conversions at target CPA/ROAS Organise by objective, product line, or funnel stage; combine match types Moderate; PMax can serve on same queries — use Search Themes in PMax and campaign negatives
Performance Max Extend reach across all Google channels; capture demand at scale Segment by product line or margin tier; use distinct asset groups per theme High if brand terms not excluded; monitor Search Insights report weekly
Standard Shopping Product-level bid control for specific SKUs or categories Use only when PMax asset groups do not provide sufficient product-level control Moderate with PMax; use campaign priorities and negatives to manage overlap
Display / Demand Gen Awareness, retargeting, upper-funnel reach Separate from conversion-focused campaigns; use audience-based targeting Low for Search; manage view-through conversion attribution carefully

Worked example

Setting up brand protection when adding Performance Max to an existing Search account

  • Setup: An Adelaide legal services firm spending AUD $12,000 a month on non-brand Search (tCPA = AUD $180, 55 conversions/month) adds a PMax campaign in September 2026 with a budget of AUD $4,000 a month. The firm’s brand name generates approximately 200 brand queries per day. Without brand exclusions, the PMax campaign begins serving on brand queries and records 40 of the month’s 70 total conversions — inflating PMax’s apparent performance.
  • Numbers: 40 brand conversions attributed to PMax at an estimated AUD $3.00 CPC × 40 clicks = AUD $120 brand spend within PMax. True incremental PMax conversions = 70 total − 55 Search − 40 PMax brand = the account is double-counting. Removing brand conversions from PMax: 70 − 40 brand = 30 genuinely new PMax conversions at a cost of AUD $4,000 − AUD $120 = AUD $3,880; PMax incremental CPA = $3,880 ÷ 30 = AUD $129.33.
  • Decision: Apply brand name as a campaign-level negative keyword in PMax (exact match); set Brand Search campaign budget to AUD $2,000 with tCPA = AUD $45; restate PMax tCPA target at AUD $140 based on incremental data.
  • Why: Without brand exclusions in PMax, the campaign cannibalises brand Search traffic and overstates its own performance, distorting tCPA targets and misallocating budget.[3][12]

7. Ad Group Structure

Ad groups are the unit of creative relevance in a Google Ads account. Their structure determines the quality and coherence of the match between a user’s query, the ad they see, and the landing page they reach. In 2026, Google’s guidance has moved decisively away from Single Keyword Ad Groups (SKAGs) and toward tightly themed ad groups that consolidate closely related keywords, reducing structural overhead while improving the data density that informs ad serving decisions.[1][6]

Keyword density

Two vendor benchmarks exist for keywords per ad group and they conflict slightly: Wordstream cites 10–20 keywords per ad group[45], while Sheaf Media Group and others cite 5–10 closely related keywords.[20] Where sources conflict, the more conservative figure — 5–10 keywords per ad group — is the safer default, because it forces genuine thematic tightness and reduces the risk of mixed-intent queries degrading ad relevance. Ad groups exceeding 50 keywords per ad group should be treated as a restructure trigger regardless of source.[61]

Responsive Search Ads

Responsive Search Ads (RSAs) are the only search ad type that can be created or edited in standard Search campaigns as of 2026.[3] The platform limit is 3 enabled RSAs per ad group.[45] Best practice is to run 2–3 RSAs per ad group with meaningfully different headlines and descriptions — not slight variations of the same message — so Google’s ad rotation system has genuine creative signals to test. Pin only the headlines that must appear for compliance or brand-consistency reasons; over-pinning limits performance data.

Ad group naming within a campaign

Each ad group name should communicate its theme immediately without needing to open it. A convention of [Theme] — [Intent] (for example, “Emergency Plumbing — Transactional” or “Hot Water Systems — Commercial Comparison”) takes seconds to scan in bulk and supports rapid budget and bid reviews.[11]

Worked example

Rebuilding SKAG-style ad groups into themed ad groups for a plumbing account

  • Setup: A Sydney plumbing account spending AUD $6,000 a month runs 28 ad groups in a single non-brand Search campaign, each containing exactly 1 keyword in exact match (SKAG structure). Average conversions per ad group per month: 2.7 (75 total ÷ 28 ad groups). Quality Scores average 5.2 across the campaign.
  • Numbers: 28 ad groups × 1 keyword each = 28 keywords. Grouped by theme: Emergency Repairs (8 keywords), Hot Water Systems (6 keywords), Blocked Drains (5 keywords), Bathroom Renovations (5 keywords), Gas Fitting (4 keywords) = 5 themed ad groups. Each themed ad group averages 28 ÷ 5 = 5.6 keywords — within the 5–10 keyword guideline.[20] Conversions redistribute: 75 ÷ 5 = 15 conversions per ad group per month, meeting the 15-conversion minimum for Smart Bidding.[1]
  • Decision: Consolidate 28 SKAGs into 5 themed ad groups; add 2 RSAs per ad group with distinct headline sets; add phrase and broad match variants to each themed group; retain all existing exact match terms.
  • Why: Google explicitly recommends moving single-keyword ad groups into themed ad groups[1][6], and 28 groups averaging 2.7 conversions each are below the 15-conversion minimum needed for Smart Bidding to function at the ad group level.[1]

8. Naming Conventions and Labels

A well-designed naming convention is operational infrastructure, not cosmetic tidiness. It lets any team member — or any future auditor — understand an account’s structure, budget allocation, and strategic intent without opening individual settings.[11] In 2026, naming conventions are also a substitute for structural splits: information that would previously have required a separate campaign to track (for example, funnel stage or promo period) can instead be encoded in a name or label, preserving data density.[11][12]

Recommended naming fields

A practical naming template for campaigns uses six fields in a consistent order:

[Channel] | [Brand/Non-brand] | [Product/Service Line] | [Geo] | [Funnel Stage] | [Objective]

Examples:

  • SRC | NB | Plumbing-Emergency | SYD | Trans | tCPA85 — Search, non-brand, emergency plumbing, Sydney, transactional intent, tCPA $85
  • PMAX | NB | HotWater | AU | All | tROAS350 — Performance Max, non-brand, hot water systems, Australia-wide, all funnel stages, tROAS 350%
  • SRC | BR | Brand | AU | All | MaxConv — Search, brand, all products, Australia-wide, all funnel stages, Maximise Conversions

The key principle is consistency: every campaign in the account must follow the same pattern so teams can sort, filter, and allocate budget without spreadsheets or institutional memory.[11]

Labels as a reporting layer

Labels should carry any metadata that does not require a structural split. Recommended label categories include:

  • Funnel stage: Awareness, Consideration, Transactional
  • Experiment status: Control, Variant-A, Variant-B
  • Promo period: EOFY-2026, BlackFriday-2026, Christmas-2026
  • Audience type: Remarketing, Prospecting, CustomerList
  • Owner: team member initials or squad name for multi-manager accounts
  • Reporting cohort: for pulling segmented reports without campaign-level splits[11][12]

Worked example

Using labels to track a Black Friday 2026 promotion without creating new campaigns

  • Setup: An Australian online homewares retailer running 5 Search campaigns and 2 PMax campaigns wants to measure incremental performance during the Black Friday promotional period (27 November – 3 December 2026) without creating 7 duplicate campaigns that would each receive less than 15 conversions in a 7-day window.[1]
  • Numbers: The account averages 210 conversions per month across 7 campaigns = 30 conversions per campaign per month. A 7-day promo window is 7 ÷ 30 × 30 = 7 days, expected to generate approximately 210 × 7 ÷ 30 = 49 total conversions across all campaigns — only 7 per campaign if split into new promo campaigns, well below the 15-conversion minimum for Smart Bidding to function.[1]
  • Decision: Apply the label “BlackFriday-2026” to all 7 existing campaigns on 27 November 2026; remove the label on 4 December 2026; filter by label in the reporting dashboard to isolate promo-period metrics. No new campaigns are created.
  • Why: Creating 7 new promo campaigns would produce only 7 conversions per campaign during the 7-day window, preventing Smart Bidding from functioning;[1] labels provide equivalent reporting segmentation without fragmenting data.[11][12]

9. Conversion Data Density

Conversion data density — the volume of conversion signals flowing into each campaign’s bidding strategy — is the single most important structural variable in a 2026 Google Ads account. Smart Bidding cannot function below a minimum conversion threshold, performs unreliably between that minimum and a stability threshold, and stabilises properly only once a campaign generates sufficient volume in the trailing 30-day window.[1][52][63]

Three thresholds from vendor guidance should inform every structural decision:

  • 15 conversions per campaign per month: the minimum for Smart Bidding to function at all. Campaigns below this threshold should be considered for consolidation or should run Maximise Clicks while building volume.[1]
  • 30 conversions per campaign per 30-day trailing window: the practical stability floor for tCPA. Below 30, the algorithm’s bid decisions are insufficiently grounded; treat any tCPA performance below this threshold as provisional.[52][63]
  • 50 conversions per campaign per 30-day trailing window: the stability floor for tROAS. tROAS requires more data because it optimises for revenue value, not just conversion count.[52]

These figures are vendor claims rather than published studies with stated sample sizes. Where a practitioner’s own account data conflicts with these thresholds, apply conservative judgement and treat the lower threshold as a trigger for investigation rather than an automatic action.

Conversion action hygiene

Even a campaign with sufficient conversion volume can be misdirected if its primary conversion actions are poorly defined. Best practice is to assign one clearly scoped primary conversion action per campaign — or at most a small set of closely related actions — and demote micro-conversions (page views, scroll depth, time on site) to secondary status.[3][9][13] Mixing primary and secondary conversions in the campaign’s optimisation target inflates apparent conversion volume while diluting the economic signal Smart Bidding uses to set bids.[3]

Accounts should also confirm that enhanced conversions and Consent Mode v2 are implemented and verified. Signal loss from unverified measurement reduces the effective conversion data available to Smart Bidding even when raw conversion counts appear adequate.[2]

Worked example

Diagnosing a tCPA instability problem caused by mixed primary conversion actions

  • Setup: A Gold Coast dental clinic account spending AUD $7,500 a month on Search runs tCPA = AUD $65, targeting new patient bookings. The campaign records 95 conversions per month and appears to exceed the 30-conversion stability threshold[52], yet actual new-patient bookings tracked by the CRM average only 18 per month.
  • Numbers: 95 total conversions: 18 new-patient bookings (primary intent) + 52 contact-form views (micro-conversion incorrectly set as primary) + 25 click-to-call initiations under 30 seconds (low-quality signal, also set as primary). Smart Bidding is optimising to 95 conversions at tCPA = AUD $65, spending AUD $7,500 ÷ 95 = AUD $78.95 per conversion — but the true cost per booked patient is AUD $7,500 ÷ 18 = AUD $416.67. The 30-conversion threshold[52] is met in total but not for the economically meaningful conversion alone (18 bookings).
  • Decision: Demote contact-form views and sub-30-second calls to secondary conversion actions; set new-patient booking (confirmed appointment in CRM, imported via conversion import) as the sole primary action; reset tCPA to AUD $380 (15% below the current true CPA of AUD $416.67 to allow learning headroom); allow a 30-day re-learning period before evaluating performance.
  • Why: Mixing micro-conversions as primary actions inflates apparent volume and misleads Smart Bidding into optimising for the wrong goal, rendering the tCPA target economically meaningless.[3][13]

10. Account Hygiene and Maintenance

A structurally sound account degrades over time without active maintenance. Duplicate conversion actions, dormant campaigns consuming budget, non-serving keywords, and stale negative keyword lists all reduce the quality of signals entering Smart Bidding and make performance data harder to interpret.[2][4] The following maintenance cadences reflect 2026 best practice.

Weekly tasks

  • Review search term reports and add negatives for irrelevant queries; update account-level negative keyword lists.[2][6]
  • Check campaign learning statuses; identify campaigns re-entering learning after a structural or budget change.
  • Monitor PMax Search Insights report for cannibalisation of non-brand Search campaigns.[3][12]

Monthly tasks

  • Audit conversion actions: confirm each active campaign has exactly one verified primary conversion action; delete duplicate or dormant conversion goals.[2][4]
  • Review keyword performance: pause or remove keywords with more than 30 days of spend and zero conversions; flag ad groups exceeding 50 keywords for consolidation.[61]
  • Check campaign conversion volumes against the 15/30/50 thresholds; flag any campaign below 15 conversions for a consolidation decision.[1][52]
  • Review Quality Scores; investigate keywords with Quality Score below 5 for ad relevance and landing page issues.[62]

Quarterly tasks

  • Full structural audit: assess whether each campaign boundary still reflects a genuine business control point — budget, bid target, conversion action, or compliance — or is a legacy split that can be merged.[1][4]
  • Verify enhanced conversions and Consent Mode v2 implementation across all properties.[2]
  • Review naming conventions and labels for drift; enforce consistency across any newly created campaigns or ad groups.[11]
  • Assess impression share: campaigns below 60% impression share[62] should be evaluated for budget constraints before any structural changes are considered.

Worked example

Monthly conversion action audit for a multi-location retail account

  • Setup: A Melbourne multi-location sports equipment retailer with 8 Search campaigns conducts its monthly hygiene audit in October 2026 and finds 14 active conversion actions in the account, of which 6 are set as primary across various campaigns. Two of the 6 primary actions are legacy goals from a previous website platform (firing on a URL that no longer exists) and are recording 0 conversions despite being assigned as primary to 3 campaigns.
  • Numbers: 3 campaigns with a broken primary conversion action are running tCPA bidding against a conversion count of 0 — the algorithm has no valid signal. Those 3 campaigns spent a combined AUD $4,200 in September 2026 and recorded 0 attributed primary conversions (though actual sales occurred, they were tracked only by a secondary action). Effective CPA as reported: undefined (divide by 0). Actual new-customer transactions traceable to those campaigns via CRM import: 31 over the same period, implying a true CPA of AUD $4,200 ÷ 31 = AUD $135.48.
  • Decision: Delete 2 legacy conversion actions; assign CRM-imported purchase confirmation as the sole primary action across all 8 campaigns; reset tCPA on the 3 affected campaigns to AUD $150 (10.7% above the observed CRM CPA of AUD $135.48 to allow Smart Bidding headroom); observe for 30 days before tightening the target.
  • Why: A broken primary conversion action renders Smart Bidding non-functional regardless of budget or structural quality;[2][4] conversion action hygiene is the prerequisite for any bidding strategy to operate correctly.

11. Common Mistakes to Avoid

The following structural mistakes are consistently identified across Google’s own guidance and 2026 industry sources. Each represents a pattern that either fragments data, misleads automation, or creates operational complexity without a business benefit.[1][3][6][7][9][10]

Mistake 1: Maintaining SKAG or match-type campaign structures

Single-keyword ad groups and separate campaigns per match type were designed for manual bidding. Under Smart Bidding, they fragment conversion data without providing additional control. Google explicitly recommends moving away from both.[1][6]

Mistake 2: Launching Smart Bidding before meeting the conversion threshold

Applying tCPA or tROAS to a campaign generating fewer than 15 conversions per month produces unstable bids and can result in significant overspend or near-zero impression share while the algorithm attempts to learn from insufficient data.[1][52] Use Maximise Conversions or Maximise Clicks until the threshold is met.

Mistake 3: Mixing primary and secondary conversion actions

Assigning micro-conversions (page views, scroll depth, add-to-cart without purchase) as primary conversion actions inflates apparent conversion volume and misleads Smart Bidding into optimising for the wrong outcome.[3][9][13]

Mistake 4: Creating structural splits for reporting purposes

A campaign created solely to separate data for a report — without a different budget, bid target, or conversion action — is a fragmentation cost with no structural benefit. Use labels and naming conventions for reporting segmentation.[11][12]

Mistake 5: Ignoring PMax and Search cannibalisation

Running PMax alongside non-brand Search without brand exclusions and regular review of the Search Insights report allows PMax to claim credit for brand queries and inflate its apparent CPA performance, while deflating the non-brand Search campaign’s volume.[3][12]

Mistake 6: Making multiple structural changes simultaneously

Changing campaign structure, bid strategy, and budget in the same week forces multiple simultaneous re-learning events and makes it impossible to isolate which change drove a performance shift. Change one variable at a time and allow a minimum of 2–4 weeks between major structural interventions.[9]

Mistake 7: Neglecting negative keyword maintenance

Broad match with Smart Bidding can serve on a wider range of queries than phrase or exact match alone. Without regular search-term audits and updated negative keyword lists, budget leaks to irrelevant queries that dilute conversion data and inflate CPA.[2][6]

Worked example

Identifying and correcting a simultaneous multi-change structural mistake

  • Setup: An e-commerce garden supplies account spending AUD $15,000 a month consolidates from 18 campaigns to 4 in March 2026, simultaneously switching from Manual CPC to tROAS = 300%, adding a new PMax campaign, and increasing the total budget from AUD $15,000 to AUD $20,000 — all in the same week. By week 3, ROAS has dropped from 310% to 190% and the account manager cannot determine which change caused the decline.
  • Numbers: 4 simultaneous changes: (1) campaign structure, (2) bid strategy (Manual CPC → tROAS 300%), (3) new PMax campaign added, (4) budget +AUD $5,000 (+33%). The new tROAS campaigns need 50 conversions per 30-day window to stabilise[52]; after consolidation, the 4 campaigns average 22 conversions each in week 1 — below the 30-conversion tCPA floor[52][63] let alone the 50-conversion tROAS floor. The account is simultaneously in 4 re-learning events with no control group.
  • Decision: Revert to Maximise Conversions (no target) on all 4 consolidated campaigns for 30 days to build conversion history; pause the new PMax campaign until Search campaigns each reach 30 conversions per month[52]; hold budget at AUD $15,000; schedule tROAS implementation for 30 days after the conversion threshold is consistently met.
  • Why: Making structural, bidding, and budget changes simultaneously prevents identification of root causes and compounds re-learning instability; Google recommends a staged approach with guardrails and sufficient data before adding automation layers.[9]

12. What Changed Recently (Last 30 Days)

The most significant development in Google Ads account structure guidance in the 30 days to August 2026 is not a new structural framework, but a stronger and more explicit push from Google’s own documentation toward broad match + Smart Bidding + Responsive Search Ads as the default operating model for Search.[1][8] The practical implications for account structure are as follows.

Google’s ABCs of Account Structure — updated emphasis

Google’s “ABCs of Account Structure” guidance now explicitly instructs practitioners to move single-keyword ad groups into themed ad groups, consolidate all match types for a given theme into one ad group, and simplify further by using broad match where appropriate.[1] This is a direct signal that SKAG-style structures are no longer considered aligned with Google’s recommended approach, even for accounts that previously received this advice from Google representatives.

Reduced segmentation emphasis

Google’s updated guidance reiterates that automated bidding does not require — and is not improved by — splitting campaigns by keyword match type, geography, or device.[8] Any account currently maintaining these splits should treat this as an active recommendation to consolidate, subject to the business control-point test described in Section 4.

RSAs as the only editable Search ad format

Responsive Search Ads remain the only search ad type that can be created or edited in standard Search campaigns.[3] This reinforces the shift toward automation at the creative level as well as the targeting and bidding level. Ad group structure should now account for the fact that creative testing happens within RSAs via asset rotation, not by creating multiple Expanded Text Ad variants.

Broad match with Smart Bidding — stronger recommendation

Google continues to strengthen its recommendation to pair broad match keywords with Smart Bidding so the AI can find additional relevant queries to meet conversion goals.[8] This recommendation is most safely acted upon in campaigns that already meet the 30-conversion stability threshold[52][63] and have clean negative keyword lists in place. Introducing broad match into a campaign below the conversion threshold, or without updated negatives, carries a meaningful budget risk.

Worked example

Safely introducing broad match into an established Search campaign in Q3 2026

  • Setup: A Canberra accounting services account spending AUD $8,000 a month runs a non-brand Search campaign with 42 conversions per month (above the 30-conversion tCPA stability threshold[52][63]) and tCPA = AUD $160. All keywords are currently exact match. The account manager wants to act on Google’s updated broad match + Smart Bidding recommendation from July 2026.[8]
  • Numbers: Current: 42 conversions/month at AUD $160 tCPA = AUD $6,720 conversion spend. Impression share = 68%, within the 60%–80% benchmark range.[62] Proposed: add broad match versions of the top 10 exact-match keywords (by conversion volume) alongside existing exact match terms within the same themed ad groups; retain tCPA = AUD $160; apply a negative keyword list of 85 irrelevant terms compiled from the previous 90-day search-term report. Budget remains AUD $8,000/month. Review at 30 days: if CPA rises above AUD $192 (20% above tCPA target), pause broad match additions and investigate search-term quality.
  • Decision: Add broad match variants to the top 10 exact-match keywords in the existing ad groups; set a CPA review trigger at AUD $192 (tCPA × 1.20); schedule a search-term audit for 7 days after implementation to update negatives.
  • Why: Google’s July 2026 updated guidance recommends pairing broad match with Smart Bidding to expand reach[8], but this is only safe to implement in campaigns that already meet the 30-conversion stability threshold[52][63] and have an active negative keyword list — both conditions are met here.

References

  1. [1] https://www.groas.com/post/google-ads-account-structure-in-2026-the-framework-that-actuall… www.groas.com
  2. [2] https://digitelia.co/blog/complete-guide-google-ads-account-structure-2026/ digitelia.co
  3. [3] https://ad-times.com/how-to-build-a-bulletproof-google-ads-account-structure-in-2026/ ad-times.com
  4. [4] https://www.get-ryze.ai/blog/structure-google-ads-account-with-ai-guide www.get-ryze.ai
  5. [5] https://twominutereports.com/blog/google-ads-best-practices twominutereports.com
  6. [6] https://support.google.com/google-ads/answer/14752782?hl=en support.google.com
  7. [7] https://services.google.com/fh/files/events/account_structure_best_practices_for_modern_se… services.google.com
  8. [8] https://www.groas.com/post/google-ads-best-practices-2026-rules-that-actually-matter www.groas.com
  9. [9] https://emarketed.com/paid-ads/running-google-ads-2026-best-practices-when-to-use-ai-featu… emarketed.com
  10. [10] https://support.google.com/google-ads/answer/6167145?hl=en support.google.com
  11. [11] https://leadsbridge.com/blog/google-ads-campaign-structure/ leadsbridge.com
  12. [12] https://aysa.ai/google-ads-campaign-structure-smart-bidding-pmax-profitability/ aysa.ai
  13. [13] https://support.google.com/google-ads/answer/7065882?hl=en support.google.com
  14. [14] https://business.google.com/uk/resources/articles/account-structures-for-search-growth-wit… business.google.com
  15. [15] https://www.digitalapplied.com/blog/ai-google-ads-bidding-automation-pmax-2026 www.digitalapplied.com
  16. [16] https://www.definedigitalacademy.com/blog/how-google-ads-will-work-in-2026 www.definedigitalacademy.com
  17. [17] https://directiveconsulting.com/blog/the-b2b-marketers-guide-to-google-ads-best-practices-… directiveconsulting.com
  18. [18] https://business.google.com/en-all/accelerate/podcasts/ads-decoded-s1e4/ business.google.com
  19. [19] https://support.google.com/google-ads/thread/413560562/google-ads-best-practices-for-2026?… support.google.com
  20. [20] https://sheafmediagroup.com/google-ads-account-structure-best-practices/ sheafmediagroup.com
  21. [21] https://support.google.com/google-ads/thread/413560562 support.google.com
  22. [22] https://twosquares.co.uk/blog/google-ads-strategies-2026 twosquares.co.uk
  23. [23] https://www.youtube.com/watch?v=3xKUU0PsHn0 www.youtube.com
  24. [24] https://growthleadersnews.com/google-ads-2026-audience-segmentation-for-20-less-cpa/ growthleadersnews.com
  25. [25] https://paidmediastudio.com/google-ads-2026-master-audience-segmentation/ paidmediastudio.com
  26. [26] https://www.youtube.com/watch?v=DQKF8O1ZcPA www.youtube.com
  27. [27] https://paidmediastudio.com/google-ads-2026-segmentation-for-20-more-conversions/ paidmediastudio.com
  28. [28] https://www.youtube.com/watch?v=nAnZY1iDvzg www.youtube.com
  29. [29] https://www.reddit.com/r/googleads/comments/1qaoyec/its_2026_whats_one_piece_of_google_ads… www.reddit.com
  30. [30] https://www.youtube.com/watch?v=wrgRbZVLrH0&vl=en-US www.youtube.com
  31. [31] https://www.factors.ai/blog/google-ads-strategy www.factors.ai
  32. [32] https://improvado.io/blog/google-ads-management improvado.io
  33. [33] https://www.youtube.com/watch?v=1_G9YN9U73s www.youtube.com
  34. [34] https://www.linkedin.com/posts/blasonde_google-ads-2026-complete-guide-for-advertisers-act… www.linkedin.com
  35. [35] https://ppchero.com/advanced-google-ads-techniques-to-master-in-2026/ ppchero.com
  36. [36] https://pressvisibility.com/google-ads-4-precision-tactics-for-2026-marketing/ pressvisibility.com
  37. [37] https://roa-marketing.com/blog/google-ads-account-structure-best-practices-2026/ roa-marketing.com
  38. [38] https://www.kampaio.com/blog/google-ads-account-restructure www.kampaio.com
  39. [39] https://www.clicktrends.com/blog/google-ads-campaign-consolidation www.clicktrends.com
  40. [40] https://www.linkedin.com/posts/bilaldalgun_googleads-ppc-campaignstructure-activity-739512… www.linkedin.com
  41. [41] https://business.google.com/uk/resources/articles/how-to-streamline-multi-account-manageme… business.google.com
  42. [42] https://developers.google.com/merchant/storebuilder/online/sections/4_3_2_consolidated_bil… developers.google.com
  43. [43] https://www.prosemedia.com/blog/google-ads-account-structure-2026-templates www.prosemedia.com
  44. [44] https://www.linkedin.com/posts/devonclarkppc_your-account-looks-organised-that-might-activ… www.linkedin.com
  45. [45] https://www.wordstream.com/blog/google-ads-account-structure www.wordstream.com
  46. [46] https://www.ppcmastery.com/newsletter/tpe-94-why-consolidation-is-the-key-to-success-with-… www.ppcmastery.com
  47. [47] https://www.youtube.com/watch?v=eZvEkVOsFgA www.youtube.com
  48. [48] https://www.linkedin.com/posts/muhammad-talha-876430182_googleads-ppc-digitalmarketing-act… www.linkedin.com
  49. [49] https://leadsbridge.com/blog/google-ads-best-practices/ leadsbridge.com
  50. [50] https://support.google.com/google-ads/answer/14099860 support.google.com
  51. [51] https://www.marinsoftware.com/learn/google-ads-account-structure-best-practices www.marinsoftware.com
  52. [52] https://www.groas.com/post/google-ads-best-practices-for-2026-the-a www.groas.com
  53. [53] https://www.keywordme.io/blog/google-ads-account-structure-best-practices www.keywordme.io
  54. [54] https://www.reddit.com/r/PPC/comments/1t5lsk7/google_ads_account_structure_best_practice/ www.reddit.com
  55. [55] https://www.youtube.com/watch?v=hYEujorNIsk&vl=en www.youtube.com
  56. [56] https://cypressnorth.com/resources/guides/google-ads-account-structure-infographic-account… cypressnorth.com
  57. [57] https://support.google.com/google-ads/answer/6154846?hl=en support.google.com
  58. [58] https://www.youtube.com/watch?v=yEVPXHl3hkE www.youtube.com
  59. [59] https://www.linkedin.com/posts/jacksonblackledge_4-advanced-google-ads-strategies-to-massi… www.linkedin.com
  60. [60] https://www.definedigitalacademy.com/blog/first-60-days-of-a-new-google-ads-campaign www.definedigitalacademy.com
  61. [61] https://www.leoads.ai/answers/google-ads/how-to-structure-google-ads-account/ www.leoads.ai
  62. [62] https://www.aol.com/articles/2026-google-ads-benchmarks-account-163006000.html www.aol.com
  63. [63] https://hocdigitalsolutions.co.uk/blog/google-ads-account-structure-2026/ hocdigitalsolutions.co.uk
  64. [64] https://marketingagent.blog/2026/04/21/winning-google-ads-campaign-structures-for-dtc-ecom… marketingagent.blog
  65. [65] https://www.digitalapplied.com/blog/google-ads-benchmarks-2026-cpc-ctr-cvr-industry www.digitalapplied.com
  66. [66] https://www.digitalapplied.com/blog/google-ads-audit-checklist-100-items www.digitalapplied.com
  67. [67] https://www.youtube.com/watch?v=ABnQ4Dc93sQ www.youtube.com
  68. [68] https://www.reddit.com/r/PPC/comments/1sk4mfn/whats_your_google_ads_campaign_structure_for… www.reddit.com
  69. [69] https://agrowth.io/blogs/google-ads/google-ads-benchmarks agrowth.io

Don’t want to do this yourself?

Get a complete, ready-to-launch Google Ads campaign — keywords validated against real search data, copy written to the best practices on this page, and a proper negative-keyword list — built for your business and delivered within 24 hours.

Build my campaign — $129

One-off price · Human-reviewed · No subscription · No access to your Google Ads account

Share the Post:

Related Posts