Search Bidding & Smart Bidding Strategy in Google Ads
When I talk to clients and take coaching calls inevitably smart bidding comes up and my clients think that changing the bid strategy or having chose the wrong one is the issue and answer but once I start looking at the account more issues start to come to the surface.
This post today is going to be long.
Again, my clients come to me thinking they have chosen the wrong bid strategy.
I hear the same frustration: "Smart Bidding or Google is bringing in bad leads. "What bid strategy should I pick instead?
Then when I pull up their ad account, the algorithm is doing exactly what they told it to do.
Unfortunately what they (or their agency) told Google Ads to do (in the form of smart bidding) has nothing to do with revenue.
This long post today, breaks down:
how search bidding actually works
why your smart bidding strategies might be optimizing for the wrong thing
and what to do about it when you are frustrated with your ads.
Key Takeaways
Search bidding in Google Ads is now Smart Bidding for the vast majority of advertisers. Over 80% of Google advertisers use automated bidding. The algorithm will literally optimize for the conversion goal you give it-good or bad.
Low-intent "easy" conversions like page views, generic button clicks, and time on site train Smart Bidding to chase cheap, low-quality traffic. There are very specific cases where easy conversions make sense (early testing, brand-new accounts with zero data), but they should never be permanent primary goals.
The most effective search bidding setups optimize to a single, high-intent primary conversion which would be a sale or qualified lead with micro-conversions used only for diagnostics and as secondary conversions.
As a Google Ads coach who audits 5 to 10 accounts weekly, I see the same mistakes repeatedly: wrong conversion goals, messy ad group structure, and zero feedback loop from actual sales and leads back into the ad platform.
Fix the input (your conversion definition) and Smart Bidding fixes the output.
What Search Bidding Actually Does in Google Ads (Not What People Assume)
Search bidding is the process where advertisers compete for ad placement on search engine results pages. This has always been the bread and butter and the main ad type that people think about in Google.
Google runs a real time auction for every search query. This really is a live auction that occurs each time a search is executed.
Your bid, along with several factors (some of which Google won’t tell us), determines whether your search ads appear and where they appear on the page. (anyone remember the good old days of spots 1, 2 and 3 and the side bar???)
Here's what's actually happening under the hood:
Search bidding decides your maximum bid in each individual auction to balance traffic, conversions, and ad spend efficiency. These are the things you define and explicitly tell Google Ads.
Advertisers set a maximum amount they are willing to pay for an ad click under a PPC model, meaning payment occurs only when a user clicks the ad-advertisers typically pay cost-per-click.
Modern search bidding is driven by smart bidding strategies like Maximize Conversions, Target CPA, Maximize Conversion Value, and Target ROAS rather than pure manual CPC. But this is what is actually happening at auction time.
Automated bidding uses machine learning to optimize bids for conversions in real time based on auction signals-device, location, browser, operating system, search context, time of day, and other factors.
Ad Rank is determined by bid amount and quality score. Google uses a second-price auction system for clicks, meaning you often pay less than your max CPC. Auction algorithms also consider the expected impact of ad extensions, and higher-quality ads can outrace higher bids due to emphasis on ad relevance and user experience. Quality Score ranges from 1 to 10 based on ad relevance, landing page experience, and click through rate.
What advertisers think is happening: "Google will find me good leads."
What is really happening: "Google AI is maximizing your defined conversion metric, nothing more."
A simple numerical example:
You run a campaign with a $100 daily budget, Target ROAS of 400%, and an average conversion value of $200.
Smart Bidding backs into your bids: to hit 4× return, Google needs $800 in conversion value per $200 spent. It sets bids in each auction based on how likely that specific user is to convert at that value-if the probability is low, the bid drops; if high, it bids aggressively up to your maximum cost threshold.
Smart Bidding Follows Orders: The "Literal Robot" Problem I See on Coaching Calls
On nearly every coaching call, I hear some version of this: "Smart Bidding is bringing in low-quality leads" or "these conversions never turn into revenue."
When I dig in, the primary conversion is almost always something soft-"Contact page visit," "time on site over 60 seconds," or a generic form submit.
And I know exactly why it's set up that way: Performance Max forces a conversion action, and the advertiser didn't have developer resources to implement real tracking. So they grabbed whatever was easiest. Proper conversion tracking still requires a developer and some trial and error to get right-add a tag, check if it fires, debug, repeat.
Furthermore, and I think this is important to say, sometimes you set it up right and think it is wrong because it is a much harder goal to achieve so you don’t see any “conversions” happening and that is because your budget and settings are too restrictive or landing page is broken or something else and you never knew because you were obviously optimizing towards the wrong thing…
Smart Bidding is literal. If you tell it that "any form submit" or "pricing page view" is a conversion, it will aggressively hunt people most likely to do that cheap action-not to buy.
Concrete B2B example: you track "PDF download" as your primary conversion. You get 200 downloads per month, but only 6 (3%) become sales-qualified leads. The rest are your competition or industry people reading your content. (I see this all the time). Smart Bidding optimizes toward the 194 junk downloads because that's what you defined as success.
Your ads looks great; your pipeline doesn't look so hot.
Smart Bidding has zero business context. It doesn't understand sales conversations, contract values, or churn unless you feed that back as conversion value or offline conversions.
In audits since 2024, I routinely see accounts with 10–20 different conversion actions all set as "primary," which confuses the algorithm about what actually matters.
This is a pattern I also unpack in depth when explaining why your Google Ads aren't working. Google's own documentation on conversion action sets emphasizes selecting which actions are included not dumping everything in.
The Easy-Conversion Trap: Why Most Search Bidding Fails Before It Starts
The "garbage in, garbage out" problem with search bidding is structural, not a bid-strategy toggle issue.
Google Ads nudges advertisers into enabling auto-created conversions-page views, Smart Goals, "session" events-just to increase reported conversion volume. The platform wants to show you results, even if those results are meaningless.
Many advertisers lack developer support to pass real revenue, qualified lead status, or offline deal data back into Google Ads. So they settle for basic button clicks or time-on-site metrics as conversions.
This all creates a false sense of campaign performance: Google Ads shows hundreds of "conversions," but your CRM shows no matching pipeline or sales.
Another example with numbers:
500 "lead" conversions in 30 days from a Maximize Conversions campaign, but only 5 real opportunities in the CRM. Smart Bidding is being trained on the 495 bad leads.
Switching from Maximize Conversions to Target CPA without fixing the conversion definition just optimizes harder for the wrong thing.
Budget considerations play a significant role, but no marketing budget can overcome a broken conversion signal and misunderstanding how Maximize Conversions can spike your budget only makes that pain worse.
Core Search Bid Strategies in Google Ads (And When They Actually Make Sense)
Different bidding strategies can prioritize different factors like traffic generation versus conversion efficiency. Campaign objectives should guide your selection. Here's what each bidding type actually does and which strategies fit your situation:
Manual CPC: The only fully manual bidding option available now. Full control at the keyword level and ad group level. Good for small accounts, validation phases, or when you don't trust tracking yet. It doesn't use Google's machine learning, so you lose auction-time optimization but gain more control.
Maximize Clicks: Sets bids to gain as many clicks as possible within your daily budget. Designed for more traffic and awareness, not sales. Use it for early testing or when you explicitly don't have conversion tracking. Think of it as a bidding strategy for as many clicks as your budget allows.
Maximize Conversions: Tries to get as many conversions as possible within budget. Automated strategies like this depend heavily on historical conversion data for effectiveness-the maximize conversions strategy needs at least 30 conversions monthly for optimal performance. Use when you have a clean, single primary conversion and are discovering realistic CPA levels.
Target CPA: Set your CPA target based on what you can afford per lead or sale. Target CPA works best with stable conversion rates. Start with a target based on recent account averages, not a wishful number. Bid based on actual data.
Maximize Conversion Value: Optimizes for total conversion value, useful for ecommerce where every order differs. Great when you don't want to lock in a fixed ROAS target yet.
Target ROAS: The target ROAS bid strategy is ideal for mature ecommerce or lead gen with reliable value mapping.
Note: Understanding how Maximize Conversions vs. Maximize Conversion Value affect budget pacing and when to use Target Impression Share vs. Maximize Conversions keeps those strategies aligned with your goals.
Target Impression Share: Aims for ad visibility at a specific rank or percentage of impressions. Use for brand defense or competitive advantage plays, not conversion efficiency.
For each strategy, allow at least one full conversion cycle (1–2 weeks minimum) as a learning period before evaluating results. Bidding strategies can align with different objectives including traffic, conversions, and visibility-pick the one that matches yours.
Search Bidding Structure: Campaigns, Ad Groups, and Keyword-Level Control
Smart Bidding is applied at the campaign level (or portfolio level). Individual keywords and ad groups mainly influence which auctions you enter and how relevant your ads are. Ad rank depends on this structure.
Under manual bidding, you can set different max CPC bids at the ad group and keyword level. Manual bidding allows adjustments at the ad group or keyword level. Keyword-level bids override ad group bids to prioritize top performers on individual keywords.
Tightly themed ad groups-small sets of closely related queries-help Smart Bidding produce better results. Organize keywords into themed ad groups for better performance: you get a better good quality score, more stable CPCs, and clearer signal on which queries drive real conversions. Use Google Keyword Planner for relevant keyword research to build these groups.
Don't mix radically different intents (e.g., "buy [product] now" with "what is [product]") in the same ad group. Bidding on the same keyword with different intent muddies performance data and keyword-level optimization. High intent keywords convert better for ads-keep them separated from research queries.
Negative keywords prevent ads from irrelevant searches. A simple structure example: one campaign with Target CPA, three ad groups by product category, with specific keywords, match types, and negatives to control search query matching. Your ad copy should match the intent of each ad group tightly.
Data Requirements for Smart Bidding: It's About Quality, Not Just Volume
Smart Bidding requires at least 15 conversions in 30 days for effectiveness with value-based strategies. The "30 conversions per month" threshold is for evaluation stability, not a hard technical requirement. You can start Smart Bidding earlier if the conversion action is high quality and you have enough data to avoid noise, especially when you're layering in value-based bidding for small businesses rather than just raw conversion counts.
For long B2B sales cycles, bid to the lowest-funnel conversion with reliable volume-like a booked demo or Sales Qualified Lead-instead of final closed-won revenue if deals take 90+ days.
Concrete example: if you only get 15 purchases per month but 80 add-to-cart events where 40% become purchases, you might initially target add-to-cart with value scaling, then graduate to purchase conversion bidding as conversion data accumulates.
Data recency matters: Smart Bidding weighs the last 30–60 days of historical data more heavily. Recent tracking breaks or major offer changes can temporarily destabilize conversion performance. This is smart bidding exploration at work-it's testing new signals while relying on recent patterns.
Allow at least one full conversion cycle (7–30 days depending on funnel length) before judging a new bid strategy or major Target CPA / Target ROAS change.
Designing High-Intent Conversion Goals for Search Bidding
A primary conversion is the one action that most closely represents business value. There should usually be only one per campaign. Ads shown to potential customers should drive toward this single goal.
High-intent conversions for ecommerce: completed purchase with transaction value, subscription signup with first payment amount, or high-value pre-order with deposit.
High-intent conversions for B2B: form submissions that pass a qualification rule, booked calendar calls, or CRM-qualified opportunities pushed back via offline conversion import.
Avoid using generic micro-conversions (homepage visit, scroll depth, time on site, simple menu clicks) as primary goals for Smart Bidding. They correlate poorly with actual revenue and produce a lower cost per "conversion" that means nothing to your business.
Use conversion value to differentiate lead quality: demo request = value $50, eBook download = value $5. Then use maximize conversion value or Target ROAS instead of plain conversion count. This lets Google AI distinguish between a tire-kicker and a buyer.
Align your primary conversion with business math: set your CPA target tied to average customer lifetime value, or your ROAS target from actual gross margin. Your bid amount should reflect what you can truly afford per acquisition, ideally using profit-driven metrics like POAS instead of only ROAS.
The Strategic Role of Micro-Conversions in Search Bidding
Micro-conversions (newsletter signups, video views, product page visits) are useful as diagnostic and early-stage signals-not main goals.
Track micro-conversions as "secondary" conversions in Google Ads so they appear in reports but do not drive Smart Bidding decisions. This gives you visibility without corrupting your paid search campaigns.
Scenarios where temporarily bidding to a micro-conversion makes sense: early product launches, brand-new domains with zero credibility, or when validating search intent before committing deeper tracking resources. These are the very specific cases where easy conversions serve a purpose.
A progression model: start by observing micro-conversions with Manual CPC or Maximize Clicks → move to Maximize Conversions on a mid-funnel action → shift to Target CPA or Target ROAS on a bottom-funnel conversion once data accumulates.
This guide breaks the process into manageable stages and reflects some of the counterintuitive truths about Google Ads optimization that most advertisers only learn after painful experience.
Set a planned transition threshold (e.g., after 50–100 quality leads) so you don't get stuck permanently optimizing to soft goals.
Bid Adjustments, Signals, and What You Still Control Under Smart Bidding
Smart Bidding already evaluates dozens of signals at auction-time using machine learning. It adjusts bids for each individual auction-device, location, time of day, audience lists, browser, language, ad position likelihood.
Bid adjustments can be set based on device, location, and demographics, but some traditional adjustments are limited or ignored by certain Smart Bidding strategies because the algorithm manages them internally. Bid adjustments can increase visibility for high-converting segments when used under manual strategies.
Levers you still control: location targeting to exclude non-service areas, ad schedule limits where you can't serve customers, and audience exclusions for clearly irrelevant segments.
Avoid stacking too many manual bid adjustments on top of Smart Bidding-it restricts the algorithm's flexibility and causes under-delivery. I see this pattern regularly in audits.
Use observation audiences (remarketing lists, in-market audiences) with Smart Bidding to get reporting visibility without forcing adjustments.
Diagnosing Bad Search Bidding Outcomes: A Coach's Audit Checklist
When clients ask "why is Smart Bidding not working?", here's my exact audit process:
Tracking first: Verify which conversion actions are marked "primary." Confirm definitions match real outcomes in your CRM or ecommerce platform for the same date range.
Conversion quality: Compare Google Ads conversions to qualified leads or sales. If you show 500 conversions but CRM shows 5 deals, Smart Bidding is optimizing to the wrong thing.
Bid strategy alignment: Ensure it matches data availability. No Target ROAS on an account with 3 sales last month. Automated bidding requires at least 15 conversions in 30 days to function, and your Auction Insights data often reveals when competitors' strategies, not your bids alone, are the issue, which is where mastering Auction Insights for competitive edge becomes critical.
Search terms review: Look for low-intent traffic patterns-job seekers, how-to queries, student research. If ads compete based on bids and quality during auctions, but your queries are informational, you're paying for the wrong search query matches.
Structure check: Mixed-intent ad groups, too many match types, or 1–2 broad keywords dominating ad spend and skewing signals.
Common fixes: Redefine primary conversion, narrow keyword intent, reset overly aggressive Target CPA/ROAS, and allow a new learning period.
The most conversions come from clean structure and honest conversion definitions exactly the kind of foundational issues a structured Google Ads audit for four core mistakes will surface, and that Google's newer Business Agent for Leads is trying to make more visible through richer lead signals.
From Theory to Action: Setting Up a High-Intent Smart Bidding Campaign
Step 1: Define business math. If lead-to-sale rate is 10% and average sale is $2,000, you can spend $200 per lead to break even. Set your initial CPA target at or near that, then adjust. Your maximum cost per acquisition must be grounded in reality.
Step 2: Choose a single primary conversion that reflects real value. Ensure accurate tracking via Google Tag, GA4, or CRM-based offline conversions.
Step 3: Build clean structure-tightly themed ad groups, keyword-level separation of high-intent vs research terms, negative keywords to block persistent bad queries.
Step 4: Select an appropriate automated bidding strategy. Start with Maximize Conversions, then move to Target CPA. Or start with Maximize Conversion Value, then progress to Target ROAS. Match your current conversion rates and data volume.
Step 5: Set realistic initial targets from the last 30–60 days of performance. Make gradual changes-no more than 15–20% target shifts per adjustment-to avoid volatility.
Step 6: Monitor the first 2–4 weeks focusing on lead quality and revenue, not just the dashboard. Coordinate with sales to tag good vs bad leads and feed that back as value adjustments.
Step 7: Use experiments in Google Ads to A/B test bid strategies or target levels instead of sweeping changes. This protects your core campaign while testing for competitive advantage.
FAQ
Should I turn on Smart Bidding if my account only gets a few conversions per month?
Smart Bidding can technically run with low volume, but it will be noisy and unstable. Start with Manual CPC or Maximize Clicks while you fix tracking and narrow keyword intent. Once you consistently get at least 15–30 real conversions per month, experiment with Maximize Conversions on a carefully chosen high-intent goal. High-quality tracking matters more than hitting an arbitrary conversion count before switching.
Is it ever OK to use page views or simple button clicks as conversions for bidding?
Using page views as primary conversions is risky because they're too easy and low-intent, causing the algorithm to chase cheap, unqualified traffic. Track them as secondary conversions for diagnostics or early testing, but plan to transition to a stronger goal-like add-to-cart, form submit, or booked call-as soon as possible. Document a clear threshold (e.g., once you have 50–100 strong conversions) for when you'll update.
What if my Smart Bidding campaigns suddenly stop performing after doing well?
Check for recent changes: new landing pages, altered conversion tracking, CPA/ROAS target changes, or new keywords that shifted traffic quality. Review the change history in Google Ads for the last 30 days and compare conversion numbers in your CRM. Give the system time to re-learn after major edits-at least one full conversion cycle-but if results don't stabilize, consider reverting changes or running a controlled experiment.
Can I use different bid strategies for different ad groups within the same campaign?
Google Ads sets one bidding strategy per campaign unless you use portfolio bid strategies, so you cannot assign different Smart Bidding types to individual ad groups. Split materially different goals into separate campaigns so each can have its own bid strategy. Mixing brand, competitor, and generic intent in a single campaign produces unstable results and should be separated.
How often should I change my Target CPA or Target ROAS?
Adjust no more than once every 1–2 weeks, and only after at least one full conversion cycle of data since the last change. Make gradual moves-around 10–20% at a time-rather than big jumps that shock the algorithm. Use experiments to test more aggressive targets while preserving stability in your main campaigns. Evaluating results too quickly or making massive shifts is one of the most common mistakes I see on coaching calls.
