Google Ads in 2026 is an automated platform. Your job is setting the right inputs, not making manual bid decisions. Drop: broad match without reliable conversion data, manual CPC at scale, identical ad copy across ad groups, ignoring search term reports, and neglecting audience signals. Double down on: Smart Bidding with clean conversion tracking, Performance Max with controlled asset groups and exclusions, first-party audience data, creative testing, and profit-based bidding where margin variation matters. Avoid making constant structural changes — each significant edit can reset Smart Bidding’s learning phase and delay performance recovery.
The shift from manual to automated PPC management has been the defining change in paid search over the last three years. Google Ads in 2026 is a platform where the algorithm handles most tactical bid decisions — and an advertiser’s job has shifted from making those decisions to feeding the algorithm the right inputs and setting the right constraints.
Some tactics that worked well five years ago now actively harm performance. Others have become more important as automation has taken over. Here is a clear view of what to stop doing and what to prioritize.
5 Tactics to Drop
1. Broad Match Without Reliable Conversion Data
Broad match keywords have evolved. In 2026, broad match relies heavily on signals from your Smart Bidding data and landing page to determine what queries to match to. Without reliable conversion data, broad match has nothing useful to optimise against — and the result is often wasted spend on irrelevant or low-intent queries.
Broad match generally performs best when supported by reliable conversion tracking and sufficient conversion volume for the chosen Smart Bidding strategy. Many practitioners prefer to enable broad match only after the account has accumulated consistent conversion data, although the appropriate volume depends on campaign goals, bidding strategy, and vertical. Without those conditions, broad match tends to produce high impression volume and low conversion efficiency.
Before enabling broad match: ensure your conversion tracking fires on actual business outcomes and that you are running Target CPA or Target ROAS rather than volume-maximisation strategies without conversion data. (Google Ads Help: About broad match)
2. Manual CPC at Scale
Google has increasingly shifted advertisers toward automated bidding, and Manual CPC is no longer available or recommended for some campaign types and account configurations. Where it still exists, running it at scale means missing the real-time auction signals — device, user context, intent, audience membership — that Smart Bidding evaluates per auction. The result is typically less efficient spend compared to well-configured automated bidding with clean conversion data.
Smart Bidding evaluates numerous contextual signals at each auction — including device, location, time, language, audience characteristics, and other auction-time factors documented by Google — that manual bidding cannot process at scale. (Google Ads Help: About Smart Bidding)
The situations where Manual CPC may still apply are narrow: very low-volume campaigns with insufficient conversion data to exit the learning phase, or specific situations where temporary control over individual keyword bids is needed during a structural transition. For most accounts running at meaningful scale, automated bidding is the appropriate direction.
3. Identical Ad Copy Across Ad Groups
Running the same Responsive Search Ad assets across multiple ad groups with different keywords signals to the algorithm that your ads are generic, and it will favour competitors with more specific message match. Identical copy also limits what you learn from creative testing — you cannot tell which messages resonate with different audience segments if everyone sees the same creative.
Each ad group should have headlines directly relevant to that ad group’s theme. Pinning specific positions can be useful where legal, branding, or compliance requirements exist, or when testing specific messaging. However, excessive pinning may reduce Google’s ability to optimise headline and description combinations — Google’s own guidance generally recommends providing multiple unpinned options so the system can test what performs best. (Google Ads Help: About Responsive Search Ads)
4. Ignoring the Search Terms Report
Automation does not eliminate the need to review what queries your ads actually matched. The search terms report surfaces irrelevant matches, brand vs. non-brand bleed, and competitor terms that may not align with your campaign intent. Reviewing it regularly and adding negatives from it is still one of the highest-ROI manual tasks in Google Ads management. High-spend accounts may require weekly reviews; lower-volume campaigns may only need periodic reviews based on activity level and budget exposure.
Performance Max campaigns in particular have limited search term visibility, but what data is available in the Insights report should be reviewed regularly. PMax campaigns without negative keyword controls can match broadly, including to queries far outside intended targeting.
5. Neglecting Audience Signals
Running campaigns without audience observations or signals leaves Smart Bidding without additional context about what converting users look like. Adding relevant audience segments — Customer Match lists, in-market segments, website visitor lists — where they improve campaign understanding or Smart Bidding signals is worth doing, particularly as third-party cookie availability decreases.
Add audience segments in observation mode to campaigns where they are relevant. Smart Bidding can use these signals without restricting delivery to those audiences. Review audience performance reports to understand which segments convert at better or worse rates than average, and use those insights to inform targeting and bidding decisions over time.
5 Tactics to Double Down On
1. Smart Bidding With Clean Conversion Tracking
The quality of your conversion tracking is now the most important input in Google Ads performance. Smart Bidding optimises for whatever conversions you tell it to optimise for. If your primary conversion action is a page view, a button click, or an add-to-cart event rather than a completed purchase or submitted lead form, the algorithm will optimise for the wrong signal and deliver poor ROI.
Review conversion tracking regularly and whenever significant website, CRM, or business-process changes occur. Ensure primary conversion actions represent actual business outcomes. Use secondary conversions — newsletter signups, video views — in observation mode only, not as Smart Bidding targets. The investment in clean conversion tracking compounds: every improvement in tracking quality improves Smart Bidding performance on an ongoing basis. See our full Smart Bidding strategy guide for setup details.
2. Performance Max With Controlled Asset Groups
Performance Max provides access to Google’s full inventory — Search, Shopping, Display, YouTube, Gmail, Maps, and Discover — within one campaign. For advertisers with strong creative assets and reliable conversion tracking, PMax can be a useful channel for reaching additional inventory beyond Search.
The most common PMax mistake is creating a single asset group with a broad mix of assets and no exclusions. Many advertisers find that themed asset groups aligned to specific products or audience segments, negative keyword lists applied at the campaign level, and brand exclusions — where separate branded Search campaigns already exist — improve campaign organisation, reporting clarity, and spend efficiency. (Google Ads Help: About Performance Max campaigns)
| Google Ads Tactic | Status in 2026 | Action |
|---|---|---|
| Broad match without conversion data | Drop | Enable only once conversion tracking is reliable and Smart Bidding is active |
| Manual CPC at scale | Drop | Migrate to Maximize Conversions or Target CPA with clean conversion data |
| Identical RSA copy across ad groups | Drop | Write theme-specific headlines per ad group; limit pinning to compliance needs |
| Skipping search term reports | Drop | Review regularly based on spend and volume; add negatives as needed |
| No audience signals | Drop | Add relevant Customer Match + in-market lists in observation mode |
| Smart Bidding + clean conversion tracking | Double down | Review conversion actions regularly; primary actions must represent real business outcomes |
| Performance Max with controlled asset groups | Double down | Use themed asset groups + brand exclusions where branded Search campaigns exist |
| First-party audience data (Customer Match) | Double down | Upload CRM lists regularly; keep lists fresh for audience signal quality |
| Creative testing at scale | Double down | Review Asset Report periodically; replace low-rated assets with genuinely different messaging |
| Profit-based bidding (Target ROAS by margin) | Double down | Pass product margin data into conversion value via feed, API, or offline import |
3. First-Party Audience Data
As third-party cookie deprecation continues and browser privacy changes restrict cross-site tracking, first-party data — customer email lists, CRM segments, loyalty program members — becomes the most durable audience signal in Google Ads.
Customer Match allows you to upload hashed email lists to Google Ads and use them as audience segments for Smart Bidding signal enhancement, audience exclusions, and bid adjustments. Customer Match provides additional first-party audience signals that can improve audience understanding and campaign strategies when sufficient eligible data is available — eligibility requirements apply, and list size affects how much signal the system can use. (Google Ads Help: About Customer Match)
4. Creative Testing at Scale
As bidding has automated, creative is one of the remaining areas where manual input directly drives performance. RSA performance depends heavily on headline and description quality, relevance to user intent, and message differentiation from competitors.
Effective creative testing means reviewing asset performance ratings periodically, replacing consistently underperforming assets with genuinely different messaging — not minor wording variations — and testing different value propositions against each other. Avoid replacing assets too frequently before they have accumulated enough impression data to produce a reliable rating.
5. Profit-Based Bidding
Many Google Ads accounts optimise for revenue — Target ROAS — without accounting for margin differences between products. Optimising for revenue can direct the algorithm toward high-revenue, low-margin products while underserving high-margin products that generate more actual profit per sale.
Profit-based bidding passes product margin or profit values through as conversion values rather than revenue. The algorithm then optimises for profit rather than revenue. This requires more sophisticated conversion value setup — typically offline conversion imports, enhanced conversions, a product feed with margin data, or CRM and API integration that passes margin values back to Google Ads. Without one of these mechanisms, Target ROAS does not automatically account for margin. For e-commerce accounts with meaningful margin variation between products, this setup can produce significantly better business outcomes than revenue-based ROAS targets.
Avoid Constant Campaign Changes: The Learning Phase
Smart Bidding learns from conversion data over time. When a campaign enters or re-enters the learning phase — typically triggered by significant changes to bid strategy, budget, targeting, or creative — performance can be temporarily unstable while the algorithm adjusts. This is expected, but repeatedly triggering the learning phase by making continuous structural edits can prevent Smart Bidding from ever reaching stable performance.
Changes that can reset or prolong the learning phase include:
- Switching bid strategies (e.g. from Maximize Conversions to Target CPA)
- Making large budget increases or decreases in a short period
- Significantly restructuring ad groups, keywords, or targeting
- Pausing and re-enabling campaigns after extended inactivity
- Changing the primary conversion action a campaign is optimising for
Large structural changes should be made thoughtfully rather than continuously. Where possible, test changes in a controlled way — using campaign experiments or making one change at a time — and allow sufficient time for performance to stabilise before evaluating results or making further edits. Google Ads’ status column will indicate when a campaign is in learning mode.
Diagnose Before You Optimise
Before making structural changes to an underperforming campaign, use Google Ads’ built-in diagnostic tools to understand what is actually happening. Many accounts over-optimise at the campaign level when the issue is upstream — in conversion tracking, landing pages, or audience quality.
- Recommendations tab: Review suggested changes, but evaluate each one against your campaign goals rather than accepting automatically. Some recommendations increase spend or change structure in ways that may not serve your account’s specific objectives.
- Change History: Review recent changes before troubleshooting. Performance drops often follow a structural change made days or weeks earlier that triggered a learning phase or altered targeting.
- Insights: Shows which search categories, audiences, and products are driving performance. Useful for understanding where the algorithm is allocating budget and whether it aligns with business priorities.
- Auction Insights: Shows how your impression share compares to competitors. Useful for diagnosing whether a CPM or click volume drop is account-specific or market-wide.
- Diagnostics: Surfaces delivery issues — disapproved ads, policy flags, billing problems, or targeting that is too narrow — that may be limiting delivery independently of bid strategy.
The Landing Page Factor
Most underperforming Google Ads campaigns have a landing page problem, not an ad or bidding problem. Smart Bidding can drive relevant clicks efficiently, but if your landing page cannot convert those clicks, no amount of bid optimisation will fix the core issue.
Landing page optimisation for paid search includes:
- Message match: The landing page headline should reflect the promise made in the ad. Mismatches between ad copy and landing page increase bounce rate and reduce conversion rate.
- Loading speed: Slow pages lose paid traffic before it converts. Mobile page speed is particularly important for Search and Performance Max campaigns where mobile traffic is significant.
- Trust signals: Reviews, security badges, client logos, or guarantee statements reduce friction for first-time visitors arriving from paid search.
- Form friction: Longer forms reduce submission rates. For lead generation campaigns, test shorter forms or multi-step flows to identify the right balance between lead quality and volume.
- Mobile usability: Paid traffic on mobile converts poorly when landing pages are not optimised for mobile viewport, tap targets, and thumb navigation.
- Single conversion goal: Landing pages with multiple competing CTAs reduce conversion rate. Each paid search landing page should have one clear primary action.
Full guidance is in our PPC landing page optimisation guide. The Quality Score guide explains how landing page experience feeds into how Google evaluates your ad in each auction — which affects both the CPCs you pay and which auctions you compete in.
Frequently Asked Questions
Should I still use Manual CPC in Google Ads?
Manual CPC is no longer available for new Search campaigns in most account configurations and continues to be deprecated across campaign formats. For accounts already using it, the cases where it still makes sense are narrow: campaigns with too little conversion data to support Smart Bidding, or brief transitional periods when tighter bid control is needed. For most accounts running at meaningful scale with reliable conversion tracking, an automated bidding strategy is more appropriate than manual CPC.
Is Broad Match safe to use now?
Broad match has improved significantly with Smart Bidding signal integration, but it is not universally appropriate. It performs best in accounts with reliable conversion tracking and consistent conversion volume on a relevant Smart Bidding strategy. Without those conditions, broad match often produces high impression volume and low conversion efficiency. Enable it alongside Target CPA or Target ROAS, monitor search terms closely, and maintain a strong negative keyword list. It is a tool that rewards well-instrumented accounts more than poorly tracked ones.
How many conversions does Smart Bidding need to work?
Google does not publish a single universal conversion minimum. Different Smart Bidding strategies and campaign types have different data requirements. Maximize Conversions can begin learning with fewer conversions than Target CPA, which in turn requires less data than Target ROAS set to a tight target. The general principle is that more reliable conversion data — in terms of tracking accuracy, consistency, and relevance to actual business outcomes — matters more than hitting a specific number. Accounts with sparse data typically perform better starting with a looser strategy (Maximize Conversions) before moving to a more constrained one (Target CPA or Target ROAS).
Should every account use Performance Max?
No. Performance Max is appropriate for accounts with strong conversion tracking, sufficient creative assets across multiple formats, and a business goal that benefits from full-funnel reach across Google’s inventory. For accounts with limited creative, unreliable conversion tracking, or a need for tight search-term control, a well-structured Search campaign is a more predictable starting point. Most accounts with reliable Search performance can test PMax alongside Search rather than replacing it.
Is Performance Max replacing Search campaigns?
No, not directly. Google has stated that Search campaigns remain a distinct and supported campaign type. Performance Max and Search serve different purposes: Search gives advertisers keyword-level control and search query visibility; Performance Max provides access to broader inventory with less manual control. Many advertisers run both simultaneously — Search for core branded and high-intent terms, PMax for broader reach and additional inventory. The two campaign types can run in the same account and Google applies its own logic to determine which serves a given query.
Sources
TL;DR Google Ads in 2026 is an automated platform. Your job is setting the right inputs, not making manual bid decisions. Drop: broad match without…
