Introduction

Every time someone searches on Google, an auction decides which ads appear. Each auction is different: the person, their location, device, time of day, the exact words they used, their recent searches and many other signals affect how likely they are to become a customer. A bid that is right for one auction may be too high or too low for the next.

For many years, advertisers set bids manually, keyword by keyword, adjusting them for devices, locations and times of day. This was slow and could only use a few signals. Today, most advertisers use Smart Bidding, Google’s set of automated bid strategies that use machine learning to set a different bid for every auction, aiming at a goal such as more conversions, a target cost per acquisition or a target return on ad spend.

Smart Bidding can improve results significantly, but it is not magic. It depends on accurate conversion data, realistic targets, enough volume and sensible human oversight. Used carelessly, it can spend budgets on the wrong goals.

This article explains how Smart Bidding works, the main strategies and when to use each, what it needs to succeed, common mistakes and how our team at Dsigngo Group manages bidding for clients in Sri Lanka and the UAE.

What is Smart Bidding?

Smart Bidding is a group of automated bid strategies in Google Ads that optimise for conversions or conversion value in each auction, a process Google calls auction-time bidding. It uses machine learning trained on large amounts of data to predict how likely each click is to lead to a conversion, and how valuable that conversion might be, then sets a bid to match.

The signals Smart Bidding can use include:

• Device and operating system.
• Physical location and location intent.
• Time of day and day of the week.
• The actual search query.
• Language settings and browser.
• Remarketing lists and audience membership.
• Ad creative being shown.
• Website or app where an ad might appear, for some campaign types.
• Recent behaviour and interactions.

Many of these signals and their combinations are impossible to manage manually. That is the main advantage of Smart Bidding: it adjusts for each auction rather than for broad averages.

The main Smart Bidding strategies

Maximise conversions

This strategy aims to get the most conversions possible within your budget. It is useful when you want volume and do not yet have a clear cost-per-conversion target, or when starting a new campaign with conversion tracking in place.

Watch out for: it will try to spend the whole budget. If budgets are large compared with available conversions, cost per conversion can rise.

Maximise conversions with a target CPA

Adding a target cost per acquisition (CPA) tells Google the average amount you would like to pay for each conversion. Google then aims for as many conversions as possible at around that average cost. Some conversions will cost more and some less.

Best for: lead generation and sales where every conversion has a similar value, and you know what you can afford to pay.

Maximise conversion value

This strategy aims to get the highest total conversion value within your budget. It requires values to be tracked, such as order totals for online stores, or estimated values for different types of leads.

Best for: businesses where conversions have different values, such as online stores with products at different prices.

Maximise conversion value with a target ROAS

Adding a target return on ad spend (ROAS) tells Google the average conversion value you want for each unit of ad spend. For example, a target ROAS of 400% aims for four rupees of conversion value for every rupee spent.

Best for: eCommerce and businesses with reliable conversion values and clear profitability targets.

Other automated strategies

Google also offers Maximise clicks, which aims for the most clicks within budget, and Target impression share, which aims to show ads in a chosen position, such as the top of the page. These are not Smart Bidding strategies in the strict sense, because they do not optimise for conversions. They are useful for specific purposes, such as gathering initial data for a new campaign or ensuring visibility for brand searches.

Choosing the right strategy

Strategies can change as campaigns mature. A common path is to start with Maximise conversions, collect data for several weeks, then add a target CPA or target ROAS based on actual results.

What Smart Bidding needs to work well

Accurate conversion tracking

Smart Bidding optimises for the conversions you tell it to count. If you track page views or minor clicks as conversions, it will find more of those, not more customers. Track meaningful actions, such as completed enquiries, calls of a minimum length, purchases and qualified leads. Insights Article 3 explains tracking in detail.

Enough conversion data

Machine learning needs data. Campaigns with very few conversions each month give the algorithm little to learn from, which can make performance unstable. Combining similar campaigns, using portfolio bid strategies across campaigns, or optimising for a slightly earlier but meaningful action can help in smaller accounts.

Realistic targets

Targets should be based on actual performance and business economics. A target CPA far below what the campaign has achieved may cause it to stop showing ads. Adjust targets gradually, often by 10% to 20% at a time, and allow time for the system to adapt.

Time to learn

When a new strategy is applied or settings change significantly, Smart Bidding enters a learning period, during which performance can fluctuate. Avoid judging results or making further major changes during this time.

Conversion values and offline data

For value-based strategies, accurate values are essential. Businesses with offline sales benefit from importing qualified leads and sales, so Smart Bidding learns which clicks become real revenue, not just enquiries.

Good inputs beyond bidding

Bidding cannot fix poor keywords, weak ads or slow landing pages. Smart Bidding performs best when targeting, creative and landing pages are strong.

How Smart Bidding and match types work together

Smart Bidding and broad match keywords are designed to work together. Broad match lets Google show ads for a wider range of relevant searches, and Smart Bidding sets lower bids for searches less likely to convert and higher bids for those more likely to convert. This combination can find valuable searches that manual keyword lists would miss.

The risk is irrelevant traffic when tracking is weak or negative keywords are missing. We introduce broad match carefully, usually through experiments, and monitor search terms closely. Insights Article 2 explains match types in more detail.

Seasonality adjustments and data exclusions

Smart Bidding learns from recent data, but it cannot always anticipate short events. Google Ads provides two tools:

• Seasonality adjustments: tell Smart Bidding to expect a significant change in conversion rates during a short period, such as a flash sale or major festival promotion.
• Data exclusions: tell Smart Bidding to ignore data from periods when tracking was broken or unusual problems occurred, so faulty data does not distort future bidding.

Using these tools at the right times keeps bidding accurate around special events and technical problems.

Value-based bidding for lead generation

Value-based bidding is usually associated with online stores, but it can also help businesses that generate leads. Not every enquiry is worth the same. A request for a large corporate order is worth more than a question about a small repair; a booking for a full wedding package is worth more than a general price enquiry.

By assigning estimated values to different conversion types, or importing actual values when leads become sales, businesses can use Maximise conversion value or Target ROAS strategies to favour the most valuable enquiries. For example, a training company might value a corporate enquiry at ten times an individual course enquiry, based on historical sales. Google then bids more for searches and audiences likely to produce corporate enquiries.

Values do not need to be perfect. Reasonable estimates, reviewed every few months against real sales, are far better than treating every lead as equal. Conversion value rules can also adjust values automatically, for example increasing the value of conversions from a particular location or audience that historically spends more.

Smart Bidding in Performance Max and Demand Gen

Smart Bidding is not limited to Search campaigns. Performance Max uses conversion-based bidding across all Google channels, choosing where to show ads and how much to bid based on your goals. Demand Gen campaigns use similar strategies for YouTube, Discover and Gmail placements. The same rules apply: accurate conversion tracking, meaningful conversion actions, realistic targets and patience during learning. Because these campaigns cover many channels, reviewing channel and placement reports alongside bidding results is especially important.

Portfolio strategies and shared budgets

A portfolio bid strategy applies one Smart Bidding strategy across several campaigns, pooling their data so the algorithm learns faster. This is useful for smaller accounts where individual campaigns have few conversions. Shared budgets let several campaigns draw from one budget, so spending can shift towards campaigns with more opportunities. Both tools need careful monitoring, because one campaign can dominate spending if its results look better in the short term.

Reading bid strategy reports

Google Ads provides reports that show how a bid strategy is performing. The bid strategy status shows whether a strategy is learning, limited by budget, limited by targets or working normally. Bid strategy reports show targets compared with actual results, and the top signals that influenced performance, such as device, location, time or audience.

These reports help explain changes. For example, a status of “limited by budget” suggests that more budget could produce more conversions at a similar cost. A status showing that targets are limiting performance suggests the target CPA may be too strict. We review these reports as part of our weekly routine and explain any important changes in monthly reports.

A Smart Bidding checklist

Before switching to Smart Bidding, or when results disappoint, check:

1. Are we tracking real business outcomes, not minor actions?
2. Are primary and secondary conversions set correctly?
3. Do we have enough conversions per month for the chosen strategy?
4. Are targets based on actual performance and business economics?
5. Have we allowed time for the learning period?
6. Are budgets sufficient for the targets?
7. Are seasonality adjustments and data exclusions used where needed?
8. Are search terms and placements still being reviewed?
9. Are offline sales or qualified leads being imported where possible?
10. Is the sales team sharing feedback on lead quality?

Testing a new bid strategy safely

Switching bid strategies affects the whole campaign, so it is worth testing the change rather than simply switching. Google Ads campaign experiments can split traffic between the original strategy and the new one, running them side by side for several weeks. This shows whether the new strategy genuinely improves results, rather than reflecting a seasonal change or a competitor’s actions. For smaller accounts where experiments would take too long, we make the change during a stable period, record the date clearly and compare results over comparable weeks, avoiding major holidays and promotions.

Common Smart Bidding mistakes

• Optimising for the wrong conversions. Counting page views, scrolls or button hovers as conversions.
• Setting unrealistic targets. Demanding a very low CPA or very high ROAS from the start.
• Changing too much, too often. Frequent strategy, budget and target changes keep campaigns in learning mode.
• Budgets that limit learning. Very tight budgets can prevent Smart Bidding from finding opportunities.
• Ignoring lead quality. Optimising for form fills without checking whether leads become customers.
• Judging results too early. Making decisions based on a few days of data.
• Leaving it unsupervised. Smart Bidding still needs regular review of search terms, placements, budgets and results.

Human oversight still matters

Smart Bidding handles the mechanics of bidding, but people still provide the strategy. Our team decides what to track and how to value conversions, sets and adjusts targets based on business economics, manages budgets across campaigns and seasons, reviews search terms and placements, tests strategies through experiments, and connects campaign results with the client’s sales outcomes. The best results come from combining machine learning with human judgement and business knowledge.

Illustrative example

The following example is illustrative and does not describe a real client.

A Sri Lankan company selling solar power systems runs Search campaigns using manual bids. It tracks all form submissions as conversions, including simple brochure downloads, and its sales team complains that many leads are not serious.

The team first separates conversions: completed quotation requests and calls longer than a minute become primary conversions, while brochure downloads become secondary. After several weeks of collecting data, campaigns switch to Maximise conversions. Once performance stabilises, a target CPA is introduced based on the actual cost per qualified quotation request. The company also starts recording which quotations become installations and imports them as offline conversions. Over time, bidding shifts towards searches and locations that produce actual installations, and the sales team reports better lead quality.

Frequently asked questions

Is Smart Bidding better than manual bidding?

For most accounts with good conversion tracking and enough data, Smart Bidding performs better because it adjusts for each auction using many signals. Manual bidding may still suit very small or unusual campaigns.

How long does Smart Bidding take to learn?

The learning period varies with conversion volume and conversion delay, often from several days to a few weeks. We avoid major changes during this time.

Can Smart Bidding overspend?

It stays within your average daily budget, though daily spending can vary. Strategies without targets aim to spend the full budget, so budgets should be set carefully.

What if our business gets very few conversions?

We may combine campaigns, use portfolio strategies, track a meaningful earlier action or use a simpler strategy until more data is available.

Do we lose control with Smart Bidding?

You still control goals, budgets, targets, keywords, ads, audiences, locations and schedules. Smart Bidding manages individual auction bids within those settings.

Conclusion

Smart Bidding uses machine learning to set the right bid for each auction, using far more signals than any manual approach. When it is fed accurate conversion data, realistic targets and strong campaigns, it can deliver more conversions or more value from the same budget. When it is fed poor data or left unsupervised, it optimises for the wrong things.

Our team sets up tracking, chooses and adjusts bidding strategies, tests changes carefully and connects results to real business outcomes.

Want to know whether your bidding strategy is working? Ask for a free Google Ads review.

Review My Bidding

Sources

• Google Ads Help, About Smart Bidding: https://support.google.com/google-ads/answer/7065882
• Google Ads Help, About Target CPA bidding: https://support.google.com/google-ads/answer/6268632
• Google Ads Help, About Target ROAS bidding: https://support.google.com/google-ads/answer/6268637
• Google Ads Help, About Maximise conversions bidding: https://support.google.com/google-ads/answer/7381968
• Google Ads Help, About seasonality adjustments: https://support.google.com/google-ads/answer/10369906
• Google Ads Help, About data exclusions: https://support.google.com/google-ads/answer/10370105
• Google Ads Help, About bid strategy statuses and learning periods: https://support.google.com/google-ads/answer/7066642