--- title: "Google Ads Best Practices for 2026: What Still Works and What to Retire" description: "A practical guide to the Google Ads habits that still help CPA and ROAS, the old controls that fight automation, and what to check before changing bids." image: "https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6ab8a5a6ad3114f8d83beb50_03b523dc-a966-408b-aed7-03584a59cdf3.png" --- September 27, 2026 • min read # Google Ads Best Practices for 2026: What Still Works and What to Retire ![Young man with curly hair wearing a black shirt outdoors against green foliage background.](https://cdn.prod.website-files.com/6821efca072e48f6f495a47e/68562d390107b3921a6e3d68_1743932904108.jpg) **Alexander Perleman**, Head Of Product @ groas Ex-Goldman Sachs and Stanford Computer Science alex@groas.ai [LinkedIn](https://groas.com/post/google-ads-best-practices-for-2026-the-o#) ![Cover image for: Google Ads Best Practices for 2026: What Still Works and What to Retire](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6ab8a5a6ad3114f8d83beb50_03b523dc-a966-408b-aed7-03584a59cdf3.png) Most **Google Ads best-practices** lists will cost you money in 2026. I spent years building single-keyword ad groups by hand and babysitting manual bids at 1am, and I still see that 2019 checklist reprinted as though nothing changed. Google blocked creation of campaign-level broad match and legacy auto-created assets on Aug 3, 2026, and is [auto-migrating Search structures to AI Max through September](https://searchengineland.com/google-sets-ai-max-migration-timeline-for-search-campaigns-485006). What you need now is a way to separate practices that feed automation from controls that fight it. Four still help move CPA; five deserve retirement. The mistake is tightening the old controls while starving Smart Bidding of the conversion data it actually bids on. #### Why does the old checklist fail now? It was written for a different control surface. In 2019, you won by isolating variables: one keyword per ad group, tight exact match, a bid for every hour of the day. In 2026, that isolation can starve the system setting the bid. Google says [Search Term Matching will not work with manual CPC because AI Max relies on automated-bidding signals](https://support.google.com/google-ads/answer/15913066), and recommends Maximize Conversions or Maximize Conversion Value for full capability. Keep the old levers and you get a thinner version of the new machine. The matching changed underneath the advice, too. With AI Max on, a standard ad group containing at least one broad-match keyword can gain keywordless recall, while [negative keywords are still honored](https://support.google.com/google-ads/answer/15913066). A keyword is no longer a neat wall around every query the way an old account diagram suggests. I used to tell clients tight control meant lower CPA. I was wrong when that control split useful conversion data into tiny piles. **Control the inputs the system learns from**, not every movement it makes. #### What should I keep doing before I touch a bid? ##### 1\. Is the primary conversion telling the truth? **Conversion tracking accuracy comes first because every bid depends on it.** Smart Bidding no longer needs a 30-conversion bank to launch. Google's liaison says you can launch straight on target CPA or ROAS if tracking is solid, with [30 conversions, or 50 for tROAS, as the evaluation window](https://www.searchenginejournal.com/google-ads-bidding-strategies-where-to-spend-your-time-in-2026/581479/) for reading results. I take that as a warning, not permission to call any event a conversion. If your primary action counts pageviews, duplicate submits, and sales calls that never got qualified, you teach the model to buy more noise. Fix imports, dedupe events, and make the paid event the primary action. Then give the result its evaluation window before declaring a bid strategy broken. A faster bid change cannot rescue a false success signal. ##### 2\. Which search terms are actually wasting money? **Keep negative-keyword discipline; drop indiscriminate negating.** In one Optmyzr poll, [77% of PPC pros said removing all negatives would worsen performance](https://www.optmyzr.com/blog/negative-keywords/). This is not the year to delete your lists and trust the machine. But a negative is a precise exclusion, not a general instruction to pursue better customers. Negatives do not get close-variant expansion: blocking broad `cheap` still allows `affordable luxury hotel` through, and [phrase and exact negatives are narrower still](https://www.optmyzr.com/blog/negative-keywords/). Say you spend $20k a month and suspect the [10% to 30% waste range described for SMB accounts without hygiene](https://www.cjco.com.au/article/negative-keywords-google-ads-small-business/). Do not turn that suspicion into a giant exclusion list. Pull Search Terms, sort by cost, and group proven waste by meaning. Keep the queries that might convert long enough to find out. I would rather block a clear waste cluster than [over-negate broad match](https://groas.com/post/negative-keywords-killing-google-ads-performance-over-negating-broad-match) until Smart Bidding has nothing useful left to learn from. ##### 3\. Does the landing page deliver the offer in the ad? **Offer-to-page match matters twice: for conversion and for matching.** AI Max reads page content to find queries, so a vague page can match worse as well as convert worse. One [message-match example cites a 212% lift](https://foundrycro.com/blog/google-ads-landing-page-best-practices-2026/); pages rated Above average for landing experience and ad relevance correlate with [36% lower CPCs](https://foundrycro.com/blog/google-ads-landing-page-best-practices-2026/). Those are reasons to inspect the page, not promises that changing a headline will reproduce either number. The baseline is unforgiving: a one-second load converts at [three times the rate of a five-second load](https://foundrycro.com/blog/google-ads-landing-page-best-practices-2026/), and mobile accounts for [82.9% of landing traffic](https://foundrycro.com/blog/google-ads-landing-page-best-practices-2026/). If your ad promises fixed pricing and the page says “call for a quote,” no bid strategy fixes that mismatch. Read the ad, open the page on your phone, and see whether the promise survives the click. ![Conversion tracking dial connected to bidding dials on a control-room wall](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6ab8a5a7ad3114f8d83beb54_730bae8c-c0bf-4df1-8730-6945fd0c9b61.png) ##### 4\. Is each intent getting enough signal to learn? **Feed the model clean signal instead of slicing it thin.** Consolidate related intent into fewer ad groups, give Smart Bidding real value differences through conversion values or qualified stages, and let broad match with AI Max handle query discovery you once did by hand. The cited benchmarks make the case for testing that approach: [AI Max delivered 14% more conversions at similar CPA versus standard Search](https://foundrycro.com/blog/google-ads-landing-page-best-practices-2026/), while a migration from exact and phrase to AI Max drove [27% more conversions](https://foundrycro.com/blog/google-ads-landing-page-best-practices-2026/). Neither result excuses bad tracking. The practical setup is less glamorous than the pitch deck: one learning stream per intent, one true primary conversion, and copy and pages that describe the same offer in the same words. If you cannot explain what the model should value, adding more keyword compartments will not explain it for you. Those four practices work because they improve what automation sees. The next five felt like discipline when I learned this job. Under Smart Bidding and broad match with AI Max, they can become ways to hide useful information from the system. For a longer teardown, see [8 Google Ads best practices that no longer work in 2026](https://groas.com/post/google-ads-best-practices-that-no-longer-work-2026). #### Myth 1: “Build single-keyword ad groups for control” People believe this because it used to work. One keyword, one ad, one landing page made it easier to see what triggered what. The trouble is that the walls are no longer as solid as the account structure suggests. With AI Max on, a standard ad group containing broad match can gain [keywordless recall](https://support.google.com/google-ads/answer/15913066), and exact match itself covers same-intent rewrites. Fourteen SKAGs can leave the same intent scattered across fourteen places while splitting 30 conversions into slivers. **Group by intent, not by string.** Give emergency intent shared copy and a page that answers it. Keep a separation when it reflects a real difference in offer or customer intent, not just because the keywords look different in a spreadsheet. #### Myth 2: “Nudge manual bids to beat the auction” People believe this because it once felt like skill. I lived in those bid-adjustment columns. But a CPC tweak cannot supply the automated-bidding signals AI Max uses; Google explicitly says [Search Term Matching will not work with manual CPC](https://support.google.com/google-ads/answer/15913066). You can keep the familiar lever and lose part of the matching you wanted to control. **Set the target and fix the inputs instead.** That means solid tracking, meaningful values, and enough budget to let the strategy learn. Leave the bid alone long enough to read what happened. Otherwise you are adjusting the gauge while arguing with the engine. #### Myth 3: “Hoard exact match to keep control” People believe it because exact used to mean exact. It has not meant that for years: same meaning or intent can include rewrites and reorders. Meanwhile, [broad match with AI Max can reach queries beyond the keyword](https://support.google.com/google-ads/answer/15913066). Hoarding 400 exacts does not restore the old boundary. It can split useful history into small piles. **Keep exact for a handful of proven money terms and brand; use broad for discovery.** Put the control where it still has a job: negatives for proven waste, budgets, conversion values, and pages. Skip an exact-match expansion that merely gives you more rows to maintain. #### Myth 4: “Optimize on a fixed schedule” People believe it because agencies sold it as process. Pause losers every Monday, check search terms every Friday, push bids on the first of the month. A calendar makes the work look orderly. It does not make an auction hold still for you. **Optimize on evidence, not Tuesdays.** I check cost-sorted search terms when spend warrants it, change targets after the [30-conversion window, or 50 for tROAS](https://www.searchenginejournal.com/google-ads-bidding-strategies-where-to-spend-your-time-in-2026/581479/), gives me something to read, and avoid a manual reset for a 2% tweak. [Smart Bidding Exploration's cited extra 10% in conversions](https://www.searchenginejournal.com/google-ads-bidding-strategies-where-to-spend-your-time-in-2026/581479/) is a better reason to leave room for discovery than a standing appointment is to cut a bid. #### Myth 5: “Chase CTR as the goal” This is the one I find hardest to kill, because I taught it myself for years. High CTR feels like winning, and it sits right there in the interface. But **CTR measures clicks, not customers**. Broad match with AI Max can lower CTR while lifting total conversions; the cited comparison found [14% more conversions at similar CPA](https://foundrycro.com/blog/google-ads-landing-page-best-practices-2026/). If you judge that trade by CTR alone, you call a useful result a failure. Treat CTR and relevance as diagnostics. Above-average landing experience and ad relevance correlate with [36% lower CPCs](https://foundrycro.com/blog/google-ads-landing-page-best-practices-2026/), which is worth watching. The target is still cost per qualified lead, ROAS, and closed revenue. A 12% CTR that buys free-trial hunters at $400 a pop is not a win. A 4% CTR that buys buyers is. #### How do I tell a live practice from a fossil? The test I use is simple: **does this give automation better information, or just give me another knob?** Feeding it means truer primary conversions, values that reflect real differences, negatives that block proven waste, and pages that deliver the ad's promise. Fighting it means splitting 40 conversions across 12 ad groups, capping bids so tightly the model cannot explore, or pausing after two bad days and calling the reset an optimization. That distinction matters even with exclusions. Performance Max campaign negatives went from [100 to 10,000 per campaign in 2025](https://www.cjco.com.au/article/negative-keywords-google-ads-small-business/): control still has a place at the exclusion layer. Yet a study of more than 7,000 PMax campaigns found only a [0.24% conversion-rate difference with versus without account exclusions](https://www.optmyzr.com/blog/negative-keywords/). Merely having a large negative list is not the skill. Finding cost-weighted waste without starving discovery is. Before you adopt the next “best practice,” ask what it changes. Does it make a conversion more truthful, an intent clearer, or a page more relevant? Keep it. Does it mainly let you intervene more often without improving any of those things? Skip it. #### What should I check before changing bids? Run this **pre-flight checklist in order**. It takes an afternoon and heads off the expensive mistakes I still see in accounts spending $20k to $50k a month: 1. **Check the primary conversion.** Make sure it is the money event, deduped and importing correctly. Make everything else secondary. 2. **Attach real values.** If a qualified SQL is worth 10 times a raw form fill, the bidding signal should reflect that difference. 3. **Pull 30 days of Search Terms and sort by cost.** Negate proven waste in clusters. Leave the maybes for the model to sort. 4. **Check the page for each top ad group.** Repeat the offer in the same words and make sure the page loads in under two seconds on mobile. 5. **Consolidate around intent.** Give each cluster a chance to clear the 30-conversion read window, then leave targets alone long enough to evaluate them. Do that and you have fed the machine. Skip it and every bid change afterward is guesswork with better branding. Your first thing today is the cost-sorted search-term pull. Not a rebuild, not a new campaign type, not another exact-match list. Open the last 30 days, sort by cost descending, and find the top three waste clusters you can actually defend excluding. Kill those. You will have done more for the account than another hour spent fighting the automation for a knob. ## Related Posts [![Cover image for: The Negative Keyword Lists I Paste Before a New Account Spends a Dollar](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6ab8a5f79083c676d267e619_3e6ecb9f-9b60-4ca1-aa24-915c1721d6b8.png) ##### The Negative Keyword Lists I Paste Before a New Account Spends a Dollar Four paste-ready negative keyword lists for new campaigns, plus the match-type checks that keep useful buyer queries from getting blocked. September 27, 2026 • min read Written by](https://groas.com/post/the-negative-keyword-starter-lists-i-act) [![Cover image for: Discovery to Demand Gen to AI Max: The Google Ads Timeline Behind Your Broken Playbook](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6ab8a5da48ded8acd14adabf_afc8d58b-7619-4c10-bbba-5c4b17964bb0.png) ##### Discovery to Demand Gen to AI Max: The Google Ads Timeline Behind Your Broken Playbook From Performance Max in 2022 to AI Max in 2026, follow the Google Ads migrations that changed how advertisers control creative, placements, budgets, and search queries. September 27, 2026 • min read Written by](https://groas.com/post/discovery-to-demand-gen-to-ai-max-the-go) [![Cover image for: The Google Ads Account That Kept Resetting Its Own Learning Phase](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6ab8a5a80a3d07ba11d85bce_b3eb3ecb-e0d1-4b74-9e5e-779fb580f0f1.png) ##### The Google Ads Account That Kept Resetting Its Own Learning Phase A composite B2B SaaS postmortem traces four months of unstable Google Ads bidding to four human decisions: an offer swap, a page launch, a budget jump, and a conversion cutover. September 27, 2026 • min read Written by](https://groas.com/post/the-account-that-never-left-the-learning) ![](https://cdn.prod.website-files.com/6821efca072e48f6f495a47e/6823c03839b2b618430d5ae6_Light-p-500.png) [![White stylized owl eyes with green background icon.](https://cdn.prod.website-files.com/6821efca072e48f6f495a47e/6a68a036a5dd5627409d3579_Groas%20Icon.png)](https://groas.com/post/google-ads-best-practices-for-2026-the-o#) [contact](mailto:scale@groas.ai) Explore [Home](https://groas.com/)[Philosophy](https://groas.com/our-philosophy)[Blog](https://groas.com/blog) What We Do [Paid Search](https://groas.com/paid-search)[Earned Search](https://groas.com/earned-search) Who We Serve [Businesses](https://groas.com/for-businesses)[Agencies](https://groas.com/for-agencies) MCP [ChatGPT](https://chatgpt.com/plugins/plugin_asdk_app_6a9c2d7108f08191beb5b3d736be8655?q=groas)[Claude](https://claude.ai/directory/mcp-groas-ai) Legal [Privacy Policy](https://groas.com/legal/privacy-policy)[Terms of Service](https://groas.com/legal/terms) Get Started [Apply](https://groas.typeform.com/to/xC1bQNUT) © 2026 groas 🇺🇸 [![](https://cdn.prod.website-files.com/62434fa732124a0fb112aab4/62434fa732124a389912aad8_linkedin%20small.svg)](https://www.linkedin.com/company/groas/) ```json { "@context": "https://schema.org", "@graph": [ { "@type": "", "@id": "https://www.groas.com/post/google-ads-best-practices-for-2026-the-o#article", "headline": "Google Ads Best Practices for 2026: What Still Works and What to Retire", "description": "A practical guide to the Google Ads habits that still help CPA and ROAS, the old controls that fight automation, and what to check before changing bids.", "url": "https://www.groas.com/post/google-ads-best-practices-for-2026-the-o", "image": { "@type": "ImageObject", "url": "https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6ab8a5a6ad3114f8d83beb50_03b523dc-a966-408b-aed7-03584a59cdf3.png" }, "thumbnailUrl": "https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6ab8a5a6ad3114f8d83beb50_03b523dc-a966-408b-aed7-03584a59cdf3.png", "datePublished": "2026-09-27T05:12:07.842Z", "dateModified": "2026-09-27T05:12:07.842Z", "inLanguage": "en-US", "isAccessibleForFree": true, "articleSection": "", "keywords": "Google Ads, ", "mainEntityOfPage": { "@type": "WebPage", "@id": "https://www.groas.com/post/google-ads-best-practices-for-2026-the-o" }, "author": { "@id": "https://www.groas.com/author/#person" }, "publisher": { "@id": "https://www.groas.com#organization" }, "about": [ { "@type": "Thing", "name": "Google Ads" }, { "@type": "Thing", "name": "Pay-per-click advertising" }, { "@type": "Thing", "name": "Performance marketing" } ], "mentions": [ { "@type": "Organization", "name": "groas", "url": "https://www.groas.com" }, { "@type": "Thing", "name": "Google Ads" } ], "speakable": { "@type": "SpeakableSpecification", "cssSelector": ["h1"] } }, { "@type": "Person", "@id": "https://www.groas.com/author/#person", "name": "", "description": "", "jobTitle": "", "email": "", "url": "https://www.groas.com/author/", "image": { "@type": "ImageObject", "url": "" }, "sameAs": [ "" ], "knowsAbout": [ "Google Ads", "Performance Max", "AI Max", "Pay-per-click advertising", "Conversion tracking", "Bid management", "Search advertising" ], "worksFor": { "@id": "https://www.groas.com#organization" } }, { "@type": "Organization", "@id": "https://www.groas.com#organization", "name": "groas", "alternateName": "groas.com", "url": "https://www.groas.com", "logo": { "@type": "ImageObject", "url": "https://cdn.prod.website-files.com/6821efca072e48f6f495a47e/6834a970e03a7016560e3515_Logo%20Design%20256x256.png" }, "description": "Dedicated strategists run your account and a proprietary engine trained on $500B in profitable ad spend optimises the execution underneath them", "sameAs": [ "https://www.linkedin.com/company/groas" ], "knowsAbout": [ "Google Ads", "Performance Max", "AI Max for Search", "Pay-per-click advertising", "Autonomous campaign management", "AI advertising agents" ], "areaServed": "Worldwide" }, { "@type": "BreadcrumbList", "itemListElement": [ { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://www.groas.com" }, { "@type": "ListItem", "position": 2, "name": "Blog", "item": "https://www.groas.com/blog" }, { "@type": "ListItem", "position": 3, "name": "", "item": "https://www.groas.com/category/" }, { "@type": "ListItem", "position": 4, "name": "Google Ads Best Practices for 2026: What Still Works and What to Retire", "item": "https://www.groas.com/post/google-ads-best-practices-for-2026-the-o" } ] } ] } ```