--- title: "What I Got Wrong About the Google Ads Learning Phase" description: "I treated Learning like a seven-day timer and tried to speed it up with daily edits. Here’s why conversion quality, volume, and stability matter more than the badge." image: "https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6abb47900e14fd365c12a743_26d4e648-2f37-4af2-8083-0acca9c9d434.png" --- September 29, 2026 • min read # What I Got Wrong About the Google Ads Learning Phase ![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/everything-i-believed-about-the-google-a#) ![Cover image for: What I Got Wrong About the Google Ads Learning Phase](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6abb47900e14fd365c12a743_26d4e648-2f37-4af2-8083-0acca9c9d434.png) I treated the Google Ads Learning label like a microwave timer: launch on Monday, wait seven days, get clean CPA the next Monday. When it did not work, I made more changes to “feed” the system conversions. I was wrong on almost every count. The label is not a promise about next Monday. It reflects a bidding strategy working with the conversion data it has, while the account around it may still be changing. A thin campaign cannot collect a useful pattern just because I want one by Friday; a campaign whose target I keep moving cannot show me what the original target would have done. Those mistakes cost volume, and they can keep CPA messy for weeks. Here are the beliefs I would retire first. #### 1\. “Just feed it conversions and it exits faster” This sounded sensible when I started managing accounts. If the system wants data, give it data. Lower the bar, buy more clicks, loosen match, stack up that conversion count and wait for the badge to clear. It is an appealing plan because the number on the screen goes up. At least something looks productive. But **the conversion count is only useful if the action means something**. Google now frames Smart Bidding calibration as [up to around 50 conversion events or three conversion cycles after a change](https://searchengineland.com/google-ads-reframes-smart-bidding-learning-around-50-conversions-490382). That is not an invitation to replace the action you care about with the easiest one to collect. Fifty newsletter downloads do not tell bidding who signs a $20k contract. The click math makes the temptation clearer. At a 2% conversion rate, collecting 50 conversions takes about [2,500 clicks; at 1%, it takes 5,000; at 10%, only 500](https://medium.com/@areeba_1512/the-learning-phase-lie-google-ads-is-not-waiting-for-clicks-e547877ea01f). When volume is thin, an easier action can make the dashboard look healthier long before it makes the bidding signal better. You have paid for those clicks either way. Buying more of the wrong signal does not make the next bidding decision smarter. I would start with one conversion action worth bidding on and keep its definition stable long enough to see what it produces. If the account cannot generate enough of that action, I would look at whether the campaigns are spreading the available data too thin before I loosen the definition of success. **Build volume around the outcome you want, not around the badge you want gone.** #### 2\. “I’ll nudge the bids every day to help it along” A hot CPC makes sitting still feel irresponsible. A small target change feels like steering; a budget bump feels like responding to demand. Do it often enough and you can spend every morning correcting yesterday’s correction. I know the appeal of opening an account, seeing an ugly number and wanting to move a lever before anyone else sees it. The trouble is that Smart Bidding has to adjust when you introduce [a new strategy, change its settings or change what is in the strategy](https://support.google.com/google-ads/answer/6263057?hl=en&authuser=19&ref_topic=6294205). Practitioners also flag budget, target, targeting and creative changes, plus pause-and-restart cycles, as moves that can [send a campaign back into Learning](https://hawksem.com/blog/google-ads-learning-phase/). These edits are not equal, and there is no magic percentage below which every change is free. The point is simpler: if I keep changing the conditions, I cannot tell whether the strategy was settling or reacting to me. That is the expensive part of the habit. The account spends while the answer keeps moving farther away. A daily trim to a target may feel cautious, but it can leave me judging a series of short, interrupted stretches as though they were one clean test. Then I blame the bidding strategy for not producing a stable CPA under conditions I never held stable. I would batch necessary edits, record what changed and give a new bid strategy time to collect conversions before declaring a winner or a failure. A 7-to-14-day wait can be a useful pause before reacting; it is not enough evidence by itself. **Make a change to answer a question, not to quiet the discomfort of watching Learning.** #### 3\. “It clears in seven days, fourteen tops” I liked the countdown version because it gave me something tidy to tell a client: wait a week, maybe two, then we optimize. Some practitioner guides give a range of [one week to one month, depending on volume](https://hawksem.com/blog/google-ads-learning-phase/). The range sounds reasonable. The mistake is treating either end as a deadline the account owes me. The arithmetic does not care about my calendar. The same guide discusses [15 conversions in 30 days as a floor, with 30 to 50 per campaign per month for better results](https://hawksem.com/blog/google-ads-learning-phase/). Consider a B2B campaign getting eight demo conversions a month. Day 14 can arrive on schedule while the campaign still has very little conversion data. Calling that a two-week failure does not create a larger sample. One timeline guide warns that campaigns below 50 conversions can [stretch past 30 days, with CPCs 20–30% high and conversion rates 40–50% below projection](https://www.chatterbuzzmedia.com/blog/how-long-does-it-take-google-ads-to-work-and-how-does-it-work/). I would not paste those figures into a forecast for every account. I would use the warning for what it is worth: a slow conversion count can make the usual seven-day story a costly planning assumption. **Conversion cycles matter as much as conversion counts.** A purchase that closes in two days gives you a much earlier read than a demo that closes in 60. For the latter, the proposed [30-, 60- and 90-day windows using Conversion Value by Time](https://www.searchenginejournal.com/how-to-test-a-new-bid-strategy-in-google-ads/571237/) are more useful than staring at last Tuesday. A week-two report cannot contain a 60-day outcome that has not happened yet. Performance Max can make the calendar story still less useful. One agency breakdown argues that it can take [roughly 45 days just to see what was built](https://sol8.com/google-ads-learning-phase/) when learning is spread across Search, Shopping, YouTube, Discover, Display and Local. I would treat that as a reason to plan a longer observation window, not as a new universal finish line. Before I trust a tCPA read, I want roughly 30 conversions; for tROAS, roughly 50. I also want the conversion cycle to have had time to show up in the results. If five campaigns are each short of data, splitting the same opportunity five ways will not make the answer arrive faster. **Judge the data you have, not the number of sunsets since launch.** #### 4\. “Screw it, I’ll just build a fresh campaign” A campaign stuck in Limited Learning makes a rebuild look satisfying. Clean names, clean ad groups, no old decisions to explain and, briefly, no stubborn status badge. It feels like control. It can also be the most elaborate way to recreate the same shortage of data. The history matters. Calibration continues after the Learning label clears, and [a new tCPA or tROAS target does not discard prior knowledge](https://ppc.land/learning-phase/). That does not mean every restructure is harmless: merging similar campaigns is different from switching conversion actions or adopting AI Max. It means “start over” should not be my first response to an account that needs more repeated signals, not a nicer-looking campaign list. Picture a Shopping account with 4,000 SKUs spread across equal budgets. If each product gets one conversion, there is little repetition for the system to work with. The same agency breakdown calls out [stale or thin history and low spend](https://sol8.com/google-ads-learning-phase/) as obstacles to learning. Rebuilding the same spread with fresh campaign names does not fix the distribution. It buys another pass through it. I would look for a smaller surface area first: concentrate budget on a handful of products, let each collect several conversions, then expand. That keeps the work pointed at the shortage rather than at the cosmetic relief of a reset. **Keep useful history; fix the structure that starves it.** ![Cartoon of a marketer pressing a restart button on a Google Ads console as a CPA graph rises](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6abb47900e14fd365c12a747_b67d6c3b-dfc7-4935-9b8a-b24a6036b70b.png) #### 5\. “If Learning drags on, Google is broken” Blaming the platform is tempting when Limited Learning lingers and CPCs look rude. Sometimes the more useful suspect is sitting on the landing page. Imagine a Friday change that keeps the same URL but swaps the headline, adds two form fields and slows the load. If the page converted at 4% before and 2% after, the bidding strategy is now working with a history that describes a different page. Waiting for the badge to disappear will not put the old conversion rate back. This is also where the labels need a closer read. *Learning* means the strategy is calibrating. *Limited Learning* points me toward constraints such as too little conversion data or an account spread too thin for the bidding strategy to learn well. A separate status, [*Eligible (limited)*, can concern policy-related serving restrictions](https://support.google.com/google-ads/answer/2684542?hl=en-GB&ref_topic=24937); it is not the same bidding diagnosis. Treating every “limited” label as a cue to wait wastes time before I have even identified the problem. For lead-gen or B2B accounts with very few events, the useful moves may be structural. That can mean [consolidating under portfolio bidding or considering an earlier-funnel conversion action when volume is under about 15 conversions in 30 days](https://ppc.land/learning-phase/). It can also mean checking whether budget, tracking, the offer or the page is the constraint. An earlier action is a tradeoff, not a license to count anything with a pulse: it needs to tell me something about the customers I want. For a Search campaign spending, say, $20k a month on tCPA, I would make the first month a short decision sequence rather than a daily bid-adjustment ritual: 1. **Week 1:** Check broken tracking and disapproved ads. Keep the initial target in place while the campaign starts collecting data. 2. **Weeks 2–3:** Avoid structural changes unless something is clearly wrong. Batch a necessary budget or copy change instead of making a new edit every morning, and note when it happened. 3. **Week 4:** Make one decision at a time. If volume is thin, consider consolidating campaigns or reassessing the conversion action. If volume is there, judge the bid strategy against the conversions and the sales cycle, not the date on the launch note. Manual CPC avoids a Smart Bidding learning phase, which can sound like an escape hatch. It also gives up the prediction work you wanted Smart Bidding to do. I keep a longer reference on [how long the learning phase lasts and which edits reset it](https://groas.com/post/google-ads-learning-phase-how-long-it-la) for the moments when a proposed fix looks suspiciously like another reset. A long sales cycle changes how I test, too. For low-volume accounts, a sequential test can be more useful than a native Experiment that splits data between two arms; I would then judge the result using [Conversion Value by Time over the actual conversion cycle](https://www.searchenginejournal.com/how-to-test-a-new-bid-strategy-in-google-ads/571237/). The cost of skipping that step is practical: I can call a 60-day close a week-two failure and change course before the outcome arrives. **Check the setup before you blame the system, and check the cycle before you blame the result.** #### 6\. “So I should leave it completely alone” This is the belief I still go back and forth on. I tell beginners to keep their hands off the bid targets early on, and I mean it. But *total freeze* never sat right with me. Fixing a broken conversion tag, adding a junk search term as a negative or stopping an obvious source of waste does not serve the same purpose as moving a tCPA target or doubling the budget on a Friday. One kind of attention protects the signal. The other changes what the strategy is trying to do. There is still a judgment call in the middle. An edit that looks small to me may change the data the strategy sees; refusing to touch anything can mean paying for a problem I already understand. I break down the data needs in my notes on [how Smart Bidding strategies handle their learning periods](https://groas.com/post/google-ads-smart-bidding-strategies-2026-learning-period-guide), but I do not have a perfect line for every account. **Protect the signal; do not keep steering it.** That is a rule I can use, not a promise that every call will be easy. It is also why I do not like managing a learning phase on a weekly check-in rhythm. A person who looks every Tuesday can overreact to three bad days or miss three days of waste. The better operating model is continuous attention to search terms, budget pacing and tracking, with targets held steady until the conversion count earns a decision. That is the model groas runs: the machine watches across the week while a human strategist owns the direction and the outcome. Stability plus attention beats stability plus neglect, and it beats attention plus constant steering. If you take one thing from the beliefs I got wrong, take this: the phase does not test your patience. It tests whether you can leave the target alone long enough for the math to settle. ## Related Posts [![Cover image for: Stop Asking When Google Ads Will Finish Learning. Stop Resetting the Conditions.](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6abb4bda3748f4695a8d3309_5fc122e6-dd3d-4a83-87da-ca6f1fd63488.png) ##### Stop Asking When Google Ads Will Finish Learning. Stop Resetting the Conditions. Budget swings, tCPA cuts, and Friday asset dumps can keep Smart Bidding recalibrating. Before you blame the learning phase, check your change history. September 29, 2026 • min read Written by](https://groas.com/post/stop-asking-how-long-the-learning-phase) [![Cover image for: ROAS vs. CPA: The Margin Mistakes Beginners Make in Google Ads](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6abb49c21470bda4cd649a58_cb8a62fb-d905-43b9-a16e-399a855282fa.png) ##### ROAS vs. CPA: The Margin Mistakes Beginners Make in Google Ads ROAS and CPA are different bets about your margins, not interchangeable bidding settings. Here’s how the wrong target can starve volume or buy unprofitable conversions. September 29, 2026 • min read Written by](https://groas.com/post/what-i-got-wrong-about-roas-and-cpa-when) [![Cover image for: Target CPA vs. Target ROAS: Which Goal Protects Profit?](https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6abb499dcae7506325addb77_7abaaade-6db5-4e1b-9fd4-069892192d8e.png) ##### Target CPA vs. Target ROAS: Which Goal Protects Profit? Target CPA controls acquisition cost; Target ROAS bids for conversion value. Here is the strongest case for each—and the default I would choose when the data cannot settle it. September 29, 2026 • min read Written by](https://groas.com/post/roas-or-cpa-the-bid-goal-argument-argued) ![](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/everything-i-believed-about-the-google-a#) [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/everything-i-believed-about-the-google-a#article", "headline": "What I Got Wrong About the Google Ads Learning Phase", "description": "I treated Learning like a seven-day timer and tried to speed it up with daily edits. Here’s why conversion quality, volume, and stability matter more than the badge.", "url": "https://www.groas.com/post/everything-i-believed-about-the-google-a", "image": { "@type": "ImageObject", "url": "https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6abb47900e14fd365c12a743_26d4e648-2f37-4af2-8083-0acca9c9d434.png" }, "thumbnailUrl": "https://cdn.prod.website-files.com/6823bbd57170ea42b357cf81/6abb47900e14fd365c12a743_26d4e648-2f37-4af2-8083-0acca9c9d434.png", "datePublished": "2026-09-29T05:07:29.525Z", "dateModified": "2026-09-29T05:07:29.525Z", "inLanguage": "en-US", "isAccessibleForFree": true, "articleSection": "", "keywords": "Google Ads, ", "mainEntityOfPage": { "@type": "WebPage", "@id": "https://www.groas.com/post/everything-i-believed-about-the-google-a" }, "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": "What I Got Wrong About the Google Ads Learning Phase", "item": "https://www.groas.com/post/everything-i-believed-about-the-google-a" } ] } ] } ```