Google Display Network Best Practices

This page is updated every two months with current best practices for Google Display Network campaigns. The Display Network reaches millions of sites and apps for very little, but without disciplined targeting, creative and exclusions it can quietly burn budget on low-quality placements. 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 Display Network 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. When to Use Display
  4. Targeting Options
  5. Optimised Targeting
  6. Responsive Display Ads
  7. Bidding Strategies
  8. Brand Safety & Exclusions
  9. Frequency & Budgets
  10. Display vs Demand Gen vs PMax
  11. Measurement & Optimisation
  12. Common Mistakes to Avoid
  13. What Changed Recently
  14. 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 Display Network (GDN) campaigns remain a powerful channel for both prospecting and remarketing in 2026, but the platform has shifted decisively toward automation, first-party data, and convergence with Demand Gen. Five principles govern best-practice management today.

  • First-party data is the strongest signal. Remarketing lists, Customer Match uploads, and CRM-derived audiences consistently outperform broad affinity or topic-only targeting for conversion efficiency. Build your audience stack from owned data outward, not from generic categories inward.[10][11]
  • Smart Bidding requires conversion volume to function correctly. Google recommends Target CPA or Target ROAS only once an ad group achieves at least 30 conversions per month. Below that threshold, use Maximise Conversions or enhanced CPC while data accumulates.[10][12]
  • Placement hygiene is non-negotiable. Mobile app inventory, parked domains, and error pages routinely inflate click counts without delivering meaningful conversions. Review the placement report weekly and exclude systematically from campaign launch.[2][3][5]
  • Creative completeness drives algorithmic performance. Responsive Display Ads (RDAs) with the full asset set — 15 images, 5 headlines, 5 descriptions, and at least one video — give Google’s system the combinations it needs to optimise. Sparse asset sets constrain delivery and slow learning.[1][8]
  • Display and Demand Gen are converging. Google launched a phased migration tool in June 2026, folding GDN inventory into Demand Gen campaigns. Standalone Display campaigns remain operable, but advertisers should audit their current setup against the GDN-in-Demand-Gen feature map now rather than waiting for a forced migration.[14][15]

2. Benchmarks and Numbers at a Glance

All figures below are vendor-published benchmarks from 2026 benchmark reports unless otherwise noted. They represent typical ranges across industries, not guarantees. USD figures are provided as published; apply a currency conversion appropriate to your reporting period when comparing against AUD account data. Where two sources report the same metric with different values, both are shown and the more conservative figure is recommended for target-setting.

Metric Typical range or threshold Applies when Source
GDN click-through rate (CTR) — overall 0.46% – 0.62% Vendor claim; cross-industry average across all Display placements. Use the lower bound (0.46%) for conservative planning. [63][66]
GDN conversion rate (CVR) — overall 0.57% – 0.84% Vendor claim; cross-industry average. Two sources conflict; use 0.57% as the conservative baseline for CPA modelling. [63][66]
GDN cost per acquisition (CPA) — overall USD $51.41 – USD $75.51 Vendor claim; cross-industry average. Sources conflict by USD $24.10; use USD $75.51 (~AUD $115) for target-setting until account data is available. [63][66]
GDN cost per click (CPC) — overall USD $0.43 – USD $0.63 Vendor claim; cross-industry average across all Display placements. [63][66]
GDN CPM — cross-industry average USD $3.12 Vendor claim; applies to awareness and CPM-optimised campaigns across all verticals. [65]
GDN CPM — Technology sector USD $3.20 Vendor claim; highest CPM vertical in the 2026 benchmark dataset. [65]
GDN CPM — Travel & Hospitality USD $2.74 Vendor claim; mid-range CPM vertical. [65]
GDN CPM — Nonprofits USD $1.62 Vendor claim; lowest CPM vertical in the 2026 benchmark dataset; useful for grant-funded awareness campaigns. [65]
Display CTR — Home Services 0.68% Vendor claim; highest CTR vertical in the 2026 industry-breakdown dataset. [70]
Display CTR — Healthcare 0.48% Vendor claim; below-average CTR; creative and audience relevance are critical levers in this vertical. [70]
Display CPC — Legal Services USD $0.72 Vendor claim; highest Display CPC by vertical in the 2026 industry dataset. [67]
Display CPC — Travel & Hospitality USD $0.31 Vendor claim; lowest Display CPC by vertical in the 2026 industry dataset; volume-heavy channel. [67]
Smart Bidding minimum conversion volume for Target CPA / Target ROAS 30 conversions per ad group per month Google recommendation; below this threshold, use Maximise Conversions or eCPC. [10]
Display audience segment minimum size 100 active users / visitors in the last 30 days Google policy; as of 2026, reduced threshold applies to all Display remarketing and Customer Match lists. [1][53]
Banner CTR — Gaming (highest vertical) 0.95% Vendor claim; banner-format CTR benchmark; significantly above cross-industry average of 0.46%–0.62%. [65]

3. When to Use the Display Network

Display is not the right channel for every objective or account stage. The key decision is whether your goal, your available data, and your creative resources align with what GDN can realistically deliver. Display excels at three distinct jobs: keeping your brand visible to people who have already shown interest (remarketing), reaching people who are actively researching a category but have not yet found you (prospecting with intent signals), and maintaining brand presence across a large audience at low CPM for awareness campaigns.[2][9]

Display is typically the wrong choice when search demand is high and capture is the priority, when your conversion volume is too low to support Smart Bidding, or when your creative team cannot supply enough quality assets to populate an RDA properly. In those cases, Search campaigns or a Performance Max campaign with strong asset groups will usually outperform a Display-only approach.[6][8]

When Display makes sense, it is most effective layered above or alongside Search rather than run in isolation. A user who has seen a Display ad and then searches for your brand or category is more likely to convert than a cold search visitor — but Display alone rarely closes the loop.[7][9]

Worked example

Deciding whether Display is appropriate for a home services account

  • Setup: A Melbourne HVAC installation business spending AUD $8,000 per month across Search campaigns, generating 55 conversions (booked inspections) per month at an average CPA of AUD $145. The account owner is considering adding a Display remarketing campaign to re-engage site visitors who did not book.
  • Numbers: Site generates 4,200 sessions per month. Booking rate is 55 ÷ 4,200 = 1.3%. Estimated non-converting visitors per month: 4,200 − 55 = 4,145. With the revised GDN audience minimum of 100 active users in 30 days[1], a remarketing list of 4,145 qualifies comfortably. At benchmark Display CPC of USD $0.63 (~AUD $0.96)[63] and a conservative Display CVR of 0.57%[63], a AUD $800/month Display remarketing budget would generate approximately 833 clicks × 0.57% = ~5 additional conversions at an estimated CPA of AUD $160 — higher than Search but acceptable for warm audience recovery.
  • Decision: Launch a Display remarketing-only campaign with a AUD $800/month budget, targeting a 30-day site visitor list, with converters excluded. Do not add prospecting Display at this stage.
  • Why: The account has enough site traffic to exceed the 100-user minimum for remarketing lists, and the incremental CPA of ~AUD $160 is within 10% of the existing Search CPA, making warm-audience Display a justifiable extension rather than a speculative spend.[1][63]

4. Display Targeting Options

Google Display offers five main targeting mechanisms: audience segments, custom segments, topics, placements, and demographic filters. In 2026, best practice is to treat these as a layered stack, with first-party and high-intent signals at the foundation, and contextual controls such as topics and placements used primarily for reach management and brand safety rather than as the primary conversion driver.[7][10]

Audience segments

Remarketing lists and Customer Match are the highest-intent options and should anchor any performance-focused Display campaign. In-market segments are the next most useful for prospecting, placing ads in front of users who are actively researching a category. Affinity audiences are broad and are better suited to awareness goals than conversion efficiency.[4][10]

Custom segments

Custom segments allow you to build an audience defined by search terms users have entered on Google, URLs they have visited, or apps they use. Building a custom segment from 15–30 high-intent competitor URLs or category search terms gives you a prospecting audience with stronger purchase-intent signals than any pre-built Google category.[4][7]

Topics and placements

Topics provide a contextual layer — your ads appear on pages about a specified subject. Use them to add reach in awareness campaigns or to prevent ads from appearing in clearly irrelevant content categories. Placement targeting (selecting specific websites or apps) is most valuable for whitelist strategies and brand safety control, not for broad reach. Always review the placement report after the first two weeks and exclude any placement that has generated more than AUD $30 in spend with zero conversions.[2][5]

Targeting option Best use case in 2026 When to avoid
Remarketing lists Re-engaging non-converting site visitors; cart abandonment When list size is below 100 active users in 30 days[1]
Customer Match Existing customer upsell; CRM-driven suppression When data hygiene is poor or match rate is likely below 20%
In-market segments Category prospecting with purchase intent When CPAs are not being tracked; signal quality is hard to validate without conversion data
Custom segments (search terms / URLs) High-intent new customer acquisition When segments are built from broad, generic terms with no purchase intent
Affinity audiences Upper-funnel awareness; brand consideration Lead generation or direct response where CPA matters
Topics Contextual reach expansion; brand safety filtering As the sole targeting method in a conversion campaign
Managed placements Whitelist strategy; high-value publisher relationships Broad reach campaigns where manual list maintenance is not resourced
Demographic filters Refining delivery for age- or income-dependent products As a primary targeting layer without an audience segment underneath

Worked example

Building a custom segment for a B2B SaaS prospecting campaign

  • Setup: A Brisbane-based B2B SaaS business selling project management software, spending AUD $5,000 per month on Search. The account manager wants to add a Display prospecting campaign targeting users who are actively researching competitor tools.
  • Numbers: Benchmark Display CPC for B2B/SaaS is USD $0.52 (~AUD $0.80)[67]. A AUD $1,500/month Display budget ÷ AUD $0.80 CPC = approximately 1,875 clicks per month. At a conservative Display CVR of 0.57%[63], expected conversions = 1,875 × 0.57% = ~11 conversions at an estimated CPA of AUD $136. The custom segment is built from 20 competitor product URLs (e.g., asana.com/product, monday.com/pricing) plus 15 search terms including “project management software comparison” and “asana alternative”.
  • Decision: Create one custom segment combining the 20 competitor URLs and 15 search terms. Set campaign objective to Leads, bidding to Maximise Conversions (not Target CPA, as 11 projected conversions is below the 30/month threshold[10]), budget AUD $1,500/month.
  • Why: Custom segments built from competitor URLs and active research terms encode purchase intent that is materially stronger than any pre-built Google affinity or in-market category, and the 30-conversion-per-month Smart Bidding threshold is not yet reachable at this budget, making Maximise Conversions the appropriate interim strategy.[10][67]

5. Optimised Targeting and Audience Expansion

As of 2026, Optimised Targeting is enabled by default on new Display campaigns. When active, Google uses signals from your landing page, existing audience lists, and conversion data to find additional users beyond those you explicitly targeted. It is most effective when your conversion tracking is clean, your conversion volume is growing, and your goal is to scale prospecting reach efficiently.[6][10]

Optimised Targeting should be disabled in two specific scenarios: strict remarketing campaigns where you need delivery confined to your exact list, and brand-safety-sensitive campaigns where precision matters more than reach. In those cases, turn it off explicitly in campaign settings rather than leaving the default in place.[6]

The practical risk of Optimised Targeting is that it can expand into audiences that look superficially similar to your converters but do not share the same purchase intent. Monitor the audience insights report after the first 30 days and check whether expanded audiences are converting at or near the same rate as your seeded lists. If expanded audience CPA is more than 40% above your target, tighten the signal by improving landing page keyword relevance or uploading a larger Customer Match list.[6][15]

Worked example

Evaluating whether Optimised Targeting is helping or hurting a prospecting campaign

  • Setup: A Sydney online education provider spending AUD $3,000/month on a Display prospecting campaign targeting a custom segment of users who searched for “online short courses” and “professional development certificate”. Optimised Targeting is enabled. After 45 days, the campaign has spent AUD $4,500 across two calendar months.
  • Numbers: Custom segment audiences delivered 280 clicks at a 0.52% CTR and generated 3 conversions (enrolments) — CPA = AUD $4,500 ÷ 3 = AUD $1,500. Optimised Targeting expanded audiences delivered 1,420 clicks at a 0.48% CTR and generated 8 conversions — CPA = AUD $562. Benchmark Display CTR for Education is 0.52%[70]; both segments are near benchmark. The expanded audience CPA of AUD $562 is 62% lower than the seeded segment CPA of AUD $1,500.
  • Decision: Keep Optimised Targeting enabled. Upload a Customer Match list of the 11 converters to reinforce the signal. Do not disable expansion.
  • Why: Optimised Targeting is producing conversions at a CPA 62% below the seeded segment, which is well within the acceptable threshold; the decision rule is to disable expansion only when expanded audience CPA exceeds the target by more than 40%.[6][15]

6. Responsive Display Ads and Creative Specs

Responsive Display Ads are the standard format for GDN in 2026. Google assembles combinations of your uploaded images, logos, headlines, descriptions, and video assets to fit available placements automatically. The system can only optimise across the combinations it has, so the single highest-impact creative action is uploading the maximum number of high-quality, diverse assets.[1][8]

Asset specifications

Asset type Maximum allowed Character or size limit Key quality rule
Short headline 5 30 characters Each headline must make sense served in isolation
Long headline 1 90 characters Should function as a standalone value proposition
Description 5 90 characters Each description must be materially different, not a light rewrite
Images (landscape, square, portrait) 15 Minimum 600×314px (landscape); 300×300px (square) No overlaid text, logos, or buttons on the image itself
Logo 5 Square and landscape variants recommended High-quality; do not place over other artwork
Video Optional; include if available Hosted on YouTube; short and clear preferred Expand inventory and creative combinations

Ad Strength

Ad Strength is a creative completeness indicator, not a performance predictor. Aim for at least Good before scaling spend, and target Excellent as a pre-scale checkpoint. The fastest ways to move from Poor to Good are: adding the missing image formats, ensuring all 5 headline slots are filled with distinct copy, and uploading at least one additional description.[1][3]

Images must follow strict composition rules: blank space must not exceed 80% of the image, no collage compositions, no mirror effects, and no digital composite backgrounds. Products or services should be the clear focal point with a single, clean composition.[1]

Worked example

Rebuilding a sparse RDA asset set to reach Excellent Ad Strength

  • Setup: An Adelaide furniture retailer running a Display campaign with a single RDA containing 3 images (landscape only), 1 logo, 3 short headlines, and 2 descriptions. Ad Strength is rated Poor. Monthly spend is AUD $2,200. The account manager wants to improve creative performance before increasing budget to AUD $4,000/month in September 2026.
  • Numbers: Current asset count: 3 images, 1 logo, 3 headlines, 2 descriptions = 9 total assets. Target for Excellent Ad Strength: 15 images (mixing product, lifestyle, and room-setting shots), 2 logos (square and landscape), 5 headlines (30 characters each), 1 long headline (90 characters), 5 descriptions (90 characters each) = 28 total assets across 6 asset types. Gap to fill: 12 images, 1 logo, 2 headlines, 1 long headline, 3 descriptions — all achievable from existing product photography and copywriting in one 2-hour creative session.
  • Decision: Upload 12 additional product and lifestyle images, add a square logo variant, write 2 new distinct headlines and 1 long headline, and add 3 descriptions with different value propositions (free delivery, 5-year warranty, 60-day returns). Do not increase budget until Ad Strength reaches at least Good.
  • Why: Google’s guidance states that uploading the full asset set gives the system more combinations to test, which is the mechanism by which RDA performance improves; the budget increase is gated on Ad Strength as a proxy for creative readiness.[1][8]

7. Bidding Strategies for Display

Bidding strategy selection for Display campaigns in 2026 is determined primarily by conversion volume. Google’s Smart Bidding documentation sets 30 conversions per ad group per month as the threshold for Target CPA or Target ROAS; below that volume, Maximise Conversions is the appropriate automated strategy, and enhanced CPC or manual CPC are fall-back options when conversion tracking is incomplete.[10][12]

A key principle from Google’s Display bidding guidance is that targets should be set from recent performance data, not aspirational goals. Setting a Target CPA 30% below your observed CPA will cause the algorithm to restrict delivery to auctions it cannot win at that price, reducing conversion volume rather than improving efficiency.[10][13]

From 17 August 2026, Google’s bidding update confirmed that Display campaigns on Target CPA, Target ROAS, or Target CPC already run with more consistent targeting toward the bid target when campaigns are budget-limited — the same behaviour being introduced to Demand Gen at that date. This means overly tight budgets on Display are especially penalising: the algorithm holds to target rather than over-spending, so a budget that is too low will simply reduce volume without improving CPA.[13]

Bidding progression by conversion volume

Monthly conversions per ad group Recommended strategy Notes
Fewer than 15 Manual CPC or enhanced CPC (eCPC) Collect data without over-automating; review placement report weekly[12]
15–29 Maximise Conversions (no target) Let the algorithm spend the budget efficiently while volume builds[10]
30 or more Target CPA or Maximise Conversions with target Set target from a 30-day rolling average CPA; do not set target more than 10–15% below observed CPA at launch[10][13]
30 or more, with revenue values Target ROAS or Maximise Conversion Value Only viable when conversion values are consistent and tracked accurately[10]

Avoid changing targets more frequently than once per 1–2 conversion cycles. A conversion cycle is the typical lag between first ad impression and recorded conversion for your product. For most lead generation accounts, this is 3–7 days; for high-consideration purchases it can be 14–30 days. Making bid changes more frequently than the conversion cycle prevents the algorithm from distinguishing signal from noise.[13][14]

Worked example

Transitioning a Display campaign from Maximise Conversions to Target CPA

  • Setup: A Perth accounting software company running a Display prospecting campaign targeting a custom segment of “accounting software” and “small business bookkeeping” searchers. The campaign launched on 1 July 2026 on Maximise Conversions with a AUD $2,500/month budget. After 45 days (1 July – 14 August 2026), it has generated 38 conversions (free trial sign-ups) at an average CPA of AUD $65.79.
  • Numbers: 38 conversions ÷ 45 days = 0.84 conversions/day × 30 = ~25 conversions projected in next 30 days if volume holds. Wait: 38 conversions in 45 days already exceeds the 30/month threshold on a cumulative basis. 30-day rolling CPA from the most recent 30 days: 26 conversions at AUD $67.50 average. Safe Target CPA launch point = AUD $67.50 × 1.10 = AUD $74.25 (10% above observed, per conservative guidance[10][13]). Do not set target at AUD $55 (aspirational), as that is 18.5% below observed CPA.
  • Decision: On 15 August 2026, switch bidding from Maximise Conversions to Target CPA = AUD $74. Do not adjust the target again until 30 September 2026 (45-day review to allow 1–2 conversion cycles[13]).
  • Why: The 30-conversion-per-month threshold has been met and a 30-day rolling CPA benchmark is available; Google’s guidance states to set the target from recent performance and to allow 1–2 conversion cycles before re-evaluating.[10][13]

8. Placement Exclusions and Brand Safety

Placement exclusions are one of the most impactful and under-used optimisation levers available on GDN. Without active exclusion management, Display budgets routinely leak into mobile app inventory (where accidental clicks from children’s games and utility apps are common), parked domains, error pages, and low-quality content farms. None of these placements typically convert, but they consume budget and distort performance data.[3][5][8]

Google’s content suitability controls provide a first layer of protection by blocking categories such as tragedy and conflict, sexually suggestive content, or profanity. However, content suitability alone is not sufficient: it operates at a categorical level and misses placement-level quality problems within otherwise acceptable categories. Combine content suitability settings with regular placement-report audits for a defensible brand safety posture.[4][5]

Exclusion priorities at campaign launch

  • Mobile app categories: Exclude all mobile app placements unless your offer is specifically designed for an app-based audience. Add the placement exclusion mobileapp::* or use the campaign-level inventory filter to remove app inventory entirely.[3][8]
  • Parked domains and error pages: These placements represent no real audience. Exclude them via content suitability controls and confirm exclusion in the placement report.[8][11]
  • Converters from acquisition campaigns: Apply an audience exclusion for users who have already converted to prevent paying to reacquire existing customers in a new-customer prospecting campaign.[4][11]
  • Specific poor performers: After the first two weeks, pull the placement report filtered to zero conversions and more than AUD $30 spend. Exclude every qualifying placement in a single batch rather than one at a time.[2][5]

For brand-safety-sensitive advertisers — financial services, healthcare, legal — apply a placement allowlist strategy from day one. Restrict delivery to pre-approved premium publishers and accept the lower scale in exchange for guaranteed brand context. Widen the allowlist only after reviewing 30 days of placement-report data.[5]

Worked example

First placement exclusion audit on a newly launched Display campaign

  • Setup: A Gold Coast private health insurance comparison service launches a Display campaign on 1 October 2026 targeting an in-market segment for “health insurance” with a AUD $4,000/month budget. No placement exclusions are applied at launch beyond default content suitability settings. After 14 days, the campaign has spent AUD $1,867.
  • Numbers: Placement report pulled on 15 October 2026. Total placements served: 312. Placements with 0 conversions and spend ≥ AUD $30: 47. Of those 47, 29 are mobile app placements (gaming apps, utility apps, children’s education apps). Combined spend on zero-conversion placements: AUD $743 = 47 ÷ 47 average AUD $15.81 each, totalling AUD $743 of the AUD $1,867 spent = 39.8% of budget wasted on zero-conversion placements in 14 days. Projected monthly waste at this rate: AUD $743 ÷ 14 × 30 = AUD $1,591/month.
  • Decision: Exclude all 47 zero-conversion placements immediately. Add a campaign-level mobile app category exclusion. Apply content suitability exclusions for “Tragedy and conflict” and “Profanity and rough language”. Review placement report again on 1 November 2026.
  • Why: 39.8% of spend going to zero-conversion placements in 14 days exceeds any acceptable waste threshold; the recommended rule is to exclude any placement with zero conversions and spend exceeding AUD $30, applied in a weekly batch review.[2][5]

9. Frequency, Pacing and Budgets

Frequency management on Display in 2026 sits at an interesting tension point: Google’s Smart Bidding guide recommends removing frequency caps from conversion-optimised campaigns because caps can prevent the algorithm from participating in auctions where a high-value user is likely to convert.[10] However, external best-practice guidance for remarketing and awareness campaigns consistently recommends controlling frequency to prevent ad fatigue and brand damage.[6][15] The recommended resolution is to treat frequency as a brand health control rather than a bidding lever: leave frequency uncapped for conversion-optimised prospecting campaigns, but apply a cap of 3–5 impressions per user per day for remarketing campaigns where overexposure risk is real.

Budget pacing should be set to Standard delivery (not Accelerated, which is deprecated for most campaign types). Standard delivery spreads budget evenly across the day, which is important for Display because auction composition and user intent vary significantly by time of day. Evaluate pacing over a 7-day window, not day-to-day, because conversion lag — the gap between impression and recorded conversion — means same-day data is routinely misleading for Display campaigns.[10][12][14]

Budget sizing for Smart Bidding campaigns should be set comfortably above the average daily spend. A budget that is consistently capping delivery (showing “Limited by budget” status) restricts the algorithm’s ability to participate in auctions and learn. If your campaign shows budget-limited status for more than 3 consecutive days, increase the daily budget by 20–30% and monitor CPA stability over the next 14 days before increasing further.[13]

Worked example

Setting budget and frequency controls for a remarketing campaign before a seasonal peak

  • Setup: A Canberra travel agency running a Display remarketing campaign targeting users who visited the “Bali packages” page in the last 30 days but did not convert. Campaign launches on 1 November 2026, ahead of a December 2026 holiday booking peak. Current list size: 1,850 active users (above the 100-user minimum[1]). Monthly budget: AUD $1,200.
  • Numbers: AUD $1,200 ÷ 30 days = AUD $40/day average. At benchmark Display CPC for Travel of USD $0.31 (~AUD $0.47)[67], AUD $40/day = approximately 85 clicks/day. List size of 1,850 users. 85 clicks/day across 1,850 users = 0.046 clicks per user per day. At a frequency of 5 impressions per user per day (target cap for remarketing), total daily impressions served = 1,850 × 5 = 9,250 impressions/day. At benchmark Travel CTR of 0.62%[70], 9,250 × 0.62% = 57 expected clicks/day — below the 85 clicks/day the budget could support, so the frequency cap is the binding constraint, not the budget. The budget is appropriately sized.
  • Decision: Set frequency cap at 5 impressions per user per day. Keep daily budget at AUD $40 (AUD $1,200/month). Set campaign end date to 31 December 2026 to capture the full holiday booking window. Do not increase budget until the placement report at day 14 confirms clean inventory.
  • Why: For remarketing campaigns, frequency caps protect brand health and prevent overexposure; the arithmetic confirms the budget is not the binding constraint at the chosen frequency, so increasing budget without first running the placement audit would increase waste rather than performance.[6][10][67]

10. Display vs Demand Gen vs Performance Max

Google’s June 2026 migration tool formalises what has been a gradual convergence: GDN inventory is being integrated into Demand Gen as “Google Display Network (GDN) in Demand Gen”, with overlapping feature sets across remarketing, custom segments, Customer Match, contextual targeting, placements, topics, and exclusions.[14][15] Standalone Display campaigns remain available and fully functional as of August 2026, but the strategic direction from Google is consolidation.

Choosing between Display, Demand Gen, and Performance Max requires clarity on three variables: creative control requirements, audience targeting precision needs, and where in the funnel the campaign is operating.

Dimension Standalone Display Demand Gen (with GDN inventory) Performance Max
Primary funnel role Remarketing; contextual prospecting Mid-funnel awareness + consideration; visual channel reach Full-funnel conversion; cross-channel automation
Audience control High — direct list, segment, and placement control High — same audience features as Display, plus YouTube and Discover[14] Low — audience signals only; Google controls final targeting
Creative control Moderate — RDA asset-level control Moderate — similar asset-level control; Google assembles combinations Low — assets provided, Google assembles all combinations across all formats
Placement transparency High — full placement report available Moderate — placement reporting available but across more inventory types Low — limited placement visibility
Minimum conversion volume for Smart Bidding 30 conversions/ad group/month[10] Similar threshold; Google guidance applies across campaign types 50 conversions/month recommended for PMax asset group
Migration status (August 2026) Available; migration tool launched June 2026[14] Actively developed; GDN inventory now included[15] Stable; separate product path

The practical recommendation for August 2026 is: keep well-performing standalone Display campaigns running and do not migrate prematurely. Review the GDN-in-Demand-Gen migration tool, map your current audience segments and creative assets to the Demand Gen equivalent, and run a parallel test before committing. Do not consolidate into Performance Max if placement transparency and audience control are priorities for your brand safety requirements.[14][15]

Worked example

Deciding whether to migrate a Display remarketing campaign to Demand Gen in Q4 2026

  • Setup: A Sydney financial planning firm running a standalone Display remarketing campaign targeting a 30-day site visitor list of 2,200 active users. Campaign spends AUD $1,800/month and generates 22 conversions (consultation bookings) per month at an average CPA of AUD $81.82. The account manager is evaluating whether to migrate to GDN-in-Demand-Gen using Google’s June 2026 migration tool ahead of a Q4 2026 budget increase to AUD $3,500/month.
  • Numbers: Current CPA: AUD $1,800 ÷ 22 = AUD $81.82. Remarketing list: 2,200 users (well above 100-user minimum[1]). Demand Gen adds YouTube and Discover inventory — broader reach but lower placement transparency. If the firm’s brand safety policy requires placement-level control (financial services, regulated content), the reduced transparency of Demand Gen’s cross-channel reporting is a material concern. At AUD $3,500/month and current CPA of AUD $81.82, projected conversions on standalone Display = 3,500 ÷ 81.82 = ~43 conversions/month — above the 30-conversion Smart Bidding threshold[10], so Target CPA can be applied.
  • Decision: Do not migrate in Q4 2026. Increase standalone Display budget to AUD $3,500/month and switch bidding to Target CPA = AUD $85 (4% above observed CPA, conservative launch margin). Revisit Demand Gen migration in Q1 2027 after reviewing placement transparency improvements.
  • Why: The firm’s regulated financial services brand safety requirements demand placement-level reporting that standalone Display provides more reliably than Demand Gen in its current form; migrating before placement transparency is confirmed risks brand safety and makes the migration’s CPA impact unmeasurable.[4][14]

11. Measurement and Optimisation

Measurement for Display campaigns requires accounting for two structural differences from Search: conversion lag is longer (Display impressions often precede conversion by days or weeks), and view-through conversions are counted differently from click-through conversions. Treat view-through conversions as directional signals rather than hard performance currency unless your attribution model has been validated against actual business outcomes.[10][14]

The core optimisation loop for a Display campaign should run on a 14-day cycle: pull the placement report, exclude zero-conversion placements above AUD $30 spend, review audience performance segments, check Ad Strength ratings, and assess whether conversion volume justifies a bidding strategy change. Do not evaluate CPA daily — Display conversion lag means same-day CPA is almost always artificially inflated.[10][13]

Key optimisation actions by campaign age

Campaign age Priority actions What to avoid
Days 1–14 Verify conversion tracking is firing; pull first placement report; exclude mobile apps and zero-conversion placements; confirm audience list sizes exceed 100 users[1] Changing bids or targets; pausing ad groups based on 7-day data
Days 15–30 Second placement exclusion batch; review audience segment performance; check Ad Strength and add missing assets; assess conversion volume vs 30/month threshold[10] Switching bidding strategy before 30 conversions are achieved
Days 31–60 Evaluate bidding strategy upgrade; set or refine Target CPA from 30-day rolling average; add negative audience exclusions for non-converters; refresh creative assets Setting Target CPA more than 15% below observed CPA[13]
Ongoing (monthly) Placement report audit; audience performance review; creative rotation check; budget vs delivery status check; competitive CPM review against benchmarks[65] Making more than one significant change per 1–2 conversion cycles[13]

Worked example

Running the 30-day optimisation review on a Display prospecting campaign

  • Setup: A Brisbane recruitment agency running a Display prospecting campaign targeting an in-market segment for “job seekers” and a custom segment built from HR software URLs. Campaign launched 1 August 2026, budget AUD $2,000/month, Maximise Conversions bidding. Review date: 1 September 2026.
  • Numbers: August 2026 results: 31 days, AUD $2,000 spent, 34 conversions (CV submissions), CPA = AUD $2,000 ÷ 34 = AUD $58.82. Placement report: 18 placements with 0 conversions and spend ≥ AUD $30 (total wasted spend: AUD $312). Ad Strength: currently Good (missing 4 image assets and 1 description). Conversion volume: 34 — exceeds the 30/month Smart Bidding threshold[10]. 30-day average CPA: AUD $58.82. Target CPA launch point: AUD $58.82 × 1.10 = AUD $64.70, rounded to AUD $65.
  • Decision: (1) Exclude 18 zero-conversion placements. (2) Add 4 images and 1 description to push Ad Strength to Excellent. (3) Switch bidding to Target CPA = AUD $65. (4) Next review: 15 October 2026 (45 days, covering 1–2 conversion cycles).
  • Why: All three upgrade conditions are met — 34 conversions exceeds the 30/month threshold, a 30-day CPA baseline is available for target-setting, and the 10% conservative margin above observed CPA is the recommended launch buffer to avoid over-restricting delivery on day one.[10][13]

12. Common Mistakes to Avoid

The following mistakes appear repeatedly in Display campaign audits and are responsible for the majority of wasted spend and poor performance outcomes. Each has a specific, correctable cause.

  • Running Display with no placement exclusions. Without exclusions, 30–50% of Display budgets typically flow to mobile app inventory and low-quality placements with no conversion potential. Apply app exclusions and run the first placement report within 14 days of launch.[3][5][8]
  • Setting Target CPA below the observed CPA too early. Setting a target more than 15% below the current CPA before the campaign has 30 conversions forces the algorithm into an impossible constraint and causes delivery to drop sharply. Use Maximise Conversions until the threshold is met.[10][13]
  • Using affinity audiences as the sole targeting method in a conversion campaign. Affinity audiences are designed for awareness and reach; they do not carry purchase-intent signals. Conversion campaigns need remarketing, Customer Match, in-market, or custom segments as the primary layer.[4][10]
  • Sparse RDA asset sets. Uploading 3 images and 2 headlines gives the algorithm fewer than 6 combinations to test. A full asset set of 15 images, 5 headlines, and 5 descriptions generates hundreds of combinations, which is what the system needs to find high-performing permutations.[1][8]
  • Evaluating CPA on a daily basis. Display conversion lag means that conversions attributed to today’s impressions may not appear for 3–14 days. Daily CPA comparisons trigger premature bid changes that destabilise the algorithm. Use 14-day rolling windows as the minimum evaluation period.[10][14]
  • Leaving Optimised Targeting enabled on strict remarketing campaigns. When Optimised Targeting is on, Google can serve ads to users outside your remarketing list. For campaigns where staying within a defined audience is a requirement, disable Optimised Targeting explicitly.[6]
  • Not excluding recent converters from acquisition campaigns. Without a converter exclusion, acquisition budgets are spent re-advertising to people who have already completed the desired action. Apply a 30-day post-conversion audience exclusion as a default.[4][11]
  • Making multiple simultaneous changes. Changing the bid strategy, target, and creative in the same week makes it impossible to attribute which change drove the resulting performance movement. Change one variable per review cycle.[13]
  • Ignoring the Demand Gen migration. Google launched the GDN-in-Demand-Gen migration tool in June 2026. Advertisers who do not engage with the migration path risk being unprepared when Google sets a firm transition date. Review the feature mapping now.[14][15]

13. What Changed Recently (Last 30 Days)

The most significant confirmed change in the last 30 days affecting Display campaigns is a bidding behaviour update effective 17 August 2026. Google confirmed that Demand Gen campaigns on Target CPA, Target ROAS, or Target CPC that are budget-limited will now perform more consistently toward the bid target — explicitly noting that Display campaigns already run this way. The practical implication is that budget-limited Display campaigns on Target CPA or Target ROAS have been behaving conservatively (holding to target rather than over-spending) for some time, which means a budget flagged as “Limited” is not merely a pacing issue but a signal that the algorithm is artificially restricting volume. Increase budgets on budget-limited Target CPA campaigns before expecting CPA improvements.[13]

The second confirmed change is that the minimum audience size for Display targeting was reduced to 100 active users or visitors in the last 30 days. This makes small remarketing lists — previously ineligible at the old threshold — usable for Display targeting, which is particularly relevant for smaller advertisers and niche B2B accounts with limited site traffic.[1][53]

The third confirmed structural change, which began in June 2026 with Google’s migration tool launch, is the formal integration of GDN inventory into Demand Gen. Google’s migration documentation maps existing Display features — including remarketing, custom segments, Customer Match, contextual targeting, placements, topics, and exclusions — directly to their Demand Gen equivalents. No firm end date for standalone Display campaigns has been announced, but the direction is consolidation.[14][15]

One area where no confirmed change was found in the research: there is no new Display-specific brand safety policy announced in the last 30 days. Existing content suitability controls and placement exclusion workflows remain the correct approach.[4][5]

Worked example

Acting on the August 2026 bidding update for a budget-limited Target CPA Display campaign

  • Setup: A Hobart aged care services provider running a Display prospecting campaign on Target CPA = AUD $90, targeting an in-market segment for “aged care” and a custom segment from relevant search terms. Daily budget: AUD $60 (AUD $1,800/month). Campaign status shows “Limited by budget” for 8 of the last 14 days. August spend to date (1–17 August 2026): AUD $720, generating 8 conversions at an observed CPA of AUD $90.00 — exactly on target.
  • Numbers: Budget-limited status on 8 of 14 days = 57% of days the algorithm is capping delivery. At AUD $60/day actual and a benchmark Display CPC for the healthcare segment of USD $0.49 (~AUD $0.75)[67], the campaign is delivering approximately 80 clicks/day when not limited. When limited, it likely stops at AUD $60 = 80 clicks exactly. Increasing daily budget by 25% to AUD $75/day (AUD $2,250/month) would allow the algorithm to compete in an additional 15–20 auctions per day. At the current CVR implied by 8 conversions from an estimated 960 clicks (8 ÷ 960 = 0.83%), 15–20 additional clicks/day × 0.83% = 0.12–0.17 additional conversions/day = ~4–5 additional conversions/month at the same AUD $90 CPA target.
  • Decision: Increase daily budget from AUD $60 to AUD $75 on 18 August 2026. Do not change Target CPA = AUD $90. Review delivery status and CPA on 1 September 2026.
  • Why: The August 2026 bidding update confirms that budget-limited Display campaigns on Target CPA hold to target rather than over-spending, so increasing budget is the correct lever to increase volume when CPA is already on target; changing the CPA target would be incorrect because observed CPA already meets the goal.[13]

References

  1. [1] https://support.google.com/google-ads/answer/9823397?hl=en support.google.com
  2. [2] https://digitelia.co/blog/google-ads-display-network-when-to-use/ digitelia.co
  3. [3] https://www.seoauthori.com/en/blog/google-display-network-guide www.seoauthori.com
  4. [4] https://www.adship.ai/blog/google-ads-display-network-guide www.adship.ai
  5. [5] https://www.webtonic.io/blog/google-display-network www.webtonic.io
  6. [6] https://benly.ai/learn/google-ads/google-display-network-guide benly.ai
  7. [7] https://www.thinkwithgoogle.com/_qs/documents/963/google-display-network-playbook-for-perf… www.thinkwithgoogle.com
  8. [8] https://aimers.io/blog/google-display-ads-best-practices aimers.io
  9. [9] https://www.sitecentre.com.au/blog/guide-to-google-display-network www.sitecentre.com.au
  10. [10] https://business.google.com/us/resources/articles/reach-larger-new-audiences/ business.google.com
  11. [11] https://www.groas.com/post/google-display-remarketing-2026-complete-guide-audiences-creati… www.groas.com
  12. [12] https://www.channable.com/blog/google-display-ads-best-practices www.channable.com
  13. [13] https://www.reddit.com/r/googleads/comments/1qaoyec/its_2026_whats_one_piece_of_google_ads… www.reddit.com
  14. [14] https://aimers.io/blog/display-ads-best-practices aimers.io
  15. [15] https://support.google.com/google-ads/thread/413560562/google-ads-best-practices-for-2026?… support.google.com
  16. [16] https://www.reddit.com/r/googleads/comments/1pdhpd8/what_are_your_google_ad_tips_for_2026/ www.reddit.com
  17. [17] https://www.northcountrygrowth.com/blog/google-ads-display-network-targeting/ www.northcountrygrowth.com
  18. [18] https://twosquares.co.uk/blog/google-display-ads twosquares.co.uk
  19. [19] https://leadsbridge.com/blog/google-ads-best-practices/ leadsbridge.com
  20. [20] https://www.youtube.com/watch?v=JU61PKzAbYw www.youtube.com
  21. [21] https://growthmindedmarketing.com/blog/google-display-ads-complete-guide/ growthmindedmarketing.com
  22. [22] https://adspyder.io/blog/google-display-ads-best-practices-examples/ adspyder.io
  23. [23] https://leadsbridge.com/blog/google-ads-campaign-structure/ leadsbridge.com
  24. [24] https://megadigital.ai/en/blog/google-display-ads/ megadigital.ai
  25. [25] https://business.google.com/us/accelerate/resources/articles/optimize-your-display-campaig… business.google.com
  26. [26] https://christopholivierconsulting.com/google-display-ads-2026/ christopholivierconsulting.com
  27. [27] https://support.google.com/google-ads/thread/413560562/google-ads-best-practices-for-2026?… support.google.com
  28. [28] https://www.linkedin.com/posts/digital-pundit-ahmedabad_display-in-2026-google-ads-activit… www.linkedin.com
  29. [29] https://www.youtube.com/watch?v=ArwfTSVvZUM www.youtube.com
  30. [30] https://twominutereports.com/blog/google-ads-best-practices twominutereports.com
  31. [31] https://www.searchenginejournal.com/google-ads-bidding-strategies-where-to-spend-your-time… www.searchenginejournal.com
  32. [32] https://www.youtube.com/watch?v=NRws5-cA_8w www.youtube.com
  33. [33] https://support.google.com/google-ads/thread/413560562 support.google.com
  34. [34] https://improvado.io/blog/display-ads-best-practices improvado.io
  35. [35] https://www.youtube.com/watch?v=NQCLCLR3NQU www.youtube.com
  36. [36] https://support.google.com/google-ads/answer/2472725?hl=en support.google.com
  37. [37] https://services.google.com/fh/files/misc/gda_smart_bidding_guide.pdf services.google.com
  38. [38] https://services.google.com/fh/files/misc/display_automated_bidding_guide.pdf services.google.com
  39. [39] https://support.google.com/google-ads/answer/17061251?hl=en support.google.com
  40. [40] https://business.google.com/en-all/accelerate/podcasts/ads-decoded-s1e4/ business.google.com
  41. [41] https://www.youtube.com/watch?v=NoNWbKzVBnU www.youtube.com
  42. [42] https://www.youtube.com/watch?v=BmCaZI3JA3s&vl=en www.youtube.com
  43. [43] https://support.google.com/google-ads/faq/10286469?hl=en support.google.com
  44. [44] https://www.wizardcreativelabs.com/post/the-2026-google-ads-guide-campaign-types-bidding-a… www.wizardcreativelabs.com
  45. [45] https://www.seroundtable.com/google-ads-100-customer-lists-39472.html www.seroundtable.com
  46. [46] https://support.google.com/google-ads/answer/2404190?hl=en support.google.com
  47. [47] https://support.google.com/google-ads/answer/2454137?hl=en support.google.com
  48. [48] https://www.searchenginejournal.com/google-display-network-settings-maximize-ad-spend/2806… www.searchenginejournal.com
  49. [49] https://aimers.io/blog/google-display-ads-targeting-options aimers.io
  50. [50] https://www.linkedin.com/posts/maulikpanchal_googleads-ppcupdate-marketingnews-activity-74… www.linkedin.com
  51. [51] https://support.google.com/displayvideo/announcements/9210754?hl=en support.google.com
  52. [52] https://improvado.io/blog/google-display-and-video-360-optimization improvado.io
  53. [53] https://ppc.land/google-slashes-audience-targeting-thresholds-to-100-users-across-all-netw… ppc.land
  54. [54] https://ads-developers.googleblog.com/2024/09/targeting-and-pacing-changes-in-display.html ads-developers.googleblog.com
  55. [55] https://www.optmyzr.com/blog/google-ads-august-2026-bidding-update/ www.optmyzr.com
  56. [56] https://support.google.com/google-ads/answer/17051545?hl=en-AU support.google.com
  57. [57] https://blog.google/products/ads-commerce/boosting-your-reach-and-performance-with-google-… blog.google
  58. [58] https://www.reddit.com/r/googleads/comments/1v38fz9/no_more_google_display_ads_display_ads… www.reddit.com
  59. [59] https://www.searchenginejournal.com/google-is-retiring-standalone-display-campaigns-in-fav… www.searchenginejournal.com
  60. [60] https://support.google.com/google-ads/thread/161834268/how-to-stop-serving-ads-to-the-goog… support.google.com
  61. [61] https://lineardesign.com/blog/display-targeting/ lineardesign.com
  62. [62] https://www.benchmarketing.org/benchmarks/google-ads/display www.benchmarketing.org
  63. [63] https://theadspend.com/blog/google-display-overall theadspend.com
  64. [64] https://rudys.ai/google-ads-benchmarks/ rudys.ai
  65. [65] https://www.digitalapplied.com/blog/display-advertising-benchmarks-2026-data-points www.digitalapplied.com
  66. [66] https://visionary-marketing.co.uk/blog/google-ads-benchmarks-2026 visionary-marketing.co.uk
  67. [67] https://www.digitalapplied.com/blog/google-ads-benchmarks-2026-cpc-ctr-cvr-industry www.digitalapplied.com
  68. [68] https://www.storegrowers.com/google-ads-benchmarks/ www.storegrowers.com
  69. [69] https://www.whitelabelagency.co/post/google-ads-benchmarks-in-2026-ctr-cpc-conversion-rate… www.whitelabelagency.co
  70. [70] https://ppcchief.com/blog/google-ads-benchmarks-2026 ppcchief.com
  71. [71] https://www.triplewhale.com/blog/google-ads-benchmarks www.triplewhale.com
  72. [72] https://www.useproactiveai.com/blog/google-ads-benchmarks-by-industry-2026-updated-data/ www.useproactiveai.com
  73. [73] https://localiq.com/blog/search-advertising-benchmarks/ localiq.com
  74. [74] https://metricnexus.ai/blog/google-ads-benchmarks-2026 metricnexus.ai
  75. [75] https://www.wordstream.com/blog/ws/2016/02/29/google-adwords-industry-benchmarks www.wordstream.com
  76. [76] https://www.get-ryze.ai/blog/google-ads-cost-benchmarks-by-industry-2026 www.get-ryze.ai
  77. [77] https://www.cristantadigitalmarketing.ca/blog/what-google-ads-benchmarks-tell-you-in-2026 www.cristantadigitalmarketing.ca
  78. [78] https://www.dollarpocket.com/google-ads-benchmarks-2026-cpc-ctr-cvr-cpl-by-industry/ www.dollarpocket.com
  79. [79] https://www.uproas.io/blog/google-ads-benchmarks www.uproas.io
  80. [80] https://terrahq.com/en/blog/google-ads-benchmarks-2025/ terrahq.com
  81. [81] https://www.promodo.com/blog/ppc-benchmarks www.promodo.com

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