AdSense Auto Ads vs Manual Ad Units: What to Choose

AdSense gives you two ways to put ads on a page. With Auto ads you add one piece of code and Google decides where ads go. With manual ad units you create each unit yourself and paste its code exactly where you want it. Neither wins on every site. Auto ads trade control for convenience, manual units trade convenience for control, and Google's help pages do not claim that either one earns more or loads faster. The honest way to choose is to understand what each method decides for you, limit the ones you adopt, and run a short, fair test on matched pages. This guide walks through that process, including which settings exist, which questions the documentation leaves open and how to read the result.

What auto ads decide for you and what you keep control of

How Google describes auto ads

According to the AdSense help page on Auto ads (Opens in a new tab), you place one ad code on your site, Google scans it and automatically places ads where they are likely to perform well. The system looks at the page structure, notices existing Google ads on the page and places new ones based on layout, the amount of content and the ads already there. If you change the site, it detects the change, reanalyses the page and adjusts. The page lists the benefits as potentially higher revenue because new locations can be found, simple setup, automatic adjustment, responsiveness on every screen size and tailoring through settings. The wording is cautious throughout: "potentially", "likely to perform well". It never promises an increase.

How Google describes manual units

The help page on ad units (Opens in a new tab) defines an ad unit as one or more Google ads displayed as a result of one piece of AdSense ad code, and says you create and manage units on the Ads page of your account. It states that ad units give you control over where ads appear. The unit types are display ads, in-feed ads that blend with feed-style content, in-article ads meant to sit between paragraphs, multiplex ads in a grid and a search engine unit. For display units, the page says they can resize to fit the layout and device, or be set to fixed sizes. If you use another interface language, the type names and menus may look different from what is written here.

The real choice: where the decision lives

Control you keep with auto ads

The help page names several controls. You can exclude whole pages, exclude areas of pages from in-page auto ads, choose which formats you allow, adjust how anchor ads appear by changing the ad code, choose vignette frequency and prevent specific links from triggering vignettes. It also points to an optimization tips page with personalised suggestions. These are the levers you use instead of positioning ads by hand.

What the pages do not say

Neither page compares revenue or page speed between the two methods. The auto ads page also does not describe a separate setting for ad density or ad load. If your account shows such a control, read its own help text; do not assume it from this article. Treat any outside claim that one method always wins as a hypothesis for your own test.

A quick decision rule

• Little time for upkeep and templates that change often: start with auto ads and narrow them with exclusions.
• A few stable templates and a strong preference for a clean reading flow: start with manual units.
• Both apply: allow only the least intrusive auto formats, place one or two manual units where the content breaks naturally, and measure.

Side-by-side: control, speed, revenue, upkeep

Control

Auto ads give indirect control. You decide formats and exclusions, and Google decides the exact positions. Manual units give direct control, but you must repeat the work in every template. A common middle path is to use manual units for the main article template and auto ads for the long tail of rarely edited pages.

Speed

Both methods load Google ad code, and the documentation says nothing about which is lighter. What affects speed in practice is the number of ads, their formats, how much space is reserved before they load and how late they load. Measure the same page template with each method and compare layout shift and load behaviour. For the general trade-off between ads and Core Web Vitals, read ads and Core Web Vitals.

Why speed needs its own check

It is tempting to assume that the method with fewer lines of code on your side is also lighter for the reader, but the ad code that runs in the browser is loaded from Google either way, and what the visitor experiences depends on the number of ads, the creatives that win the auction and how your page reserves space. Measure the same template in a lab tool for a repeatable view, and in real-user data where your traffic is large enough. Look first at layout shift and at whether the main content appears before the first ad. If a method adds visible jumps, fix the space reservation before concluding that the method itself is the problem.

Revenue

The auto ads page says new locations may be found and revenue may be higher. The ad units page makes no revenue claim. Revenue depends on your audience, topic, content length, device mix and how many ads you accept. To compare, use a normalised figure such as revenue per thousand pageviews on matched pages. The definitions are in our article on RPM, eCPM and fill rate.

Upkeep

Auto ads reduce template work but add a monitoring chore, because positions can move after Google reanalyses a page. Manual units are stable until you change the template, then they need hand edits. Budget for the chore that fits your team.

Summary table

Which page types lean which way

Limiting auto ads placement and density

Start by deciding where ads must never appear

Open your site's page types and write a list of exclusions before you switch anything on. Pages where people sign in, pay, submit a form or read a legal notice should be excluded. Then list areas inside article pages that must stay clean, such as a table of key data, a call-to-action block or an embedded video. The help page says you can exclude specific pages and specific areas of pages, so this list maps directly onto settings.

Format by format

Read the rules on intrusive interstitials and the Better Ads standards as well, and choose full-screen formats within those limits.

A manual density audit

Settings do not replace your own eyes. Open three typical pages on a phone and on a desktop and record:

1. How many ads are visible before the first paragraph is finished.
2. How many paragraphs of content separate any two ads.
3. Whether any layer covers navigation, a form field or a button.
4. Whether a dismiss control is easy to find and use on a touch screen.
5. Whether page text jumps when an ad loads.

Example scenario: a language-learning blog

Example scenario: A language-learning blog publishes vocabulary lists, grammar lessons and a members-only practice area. The owner enables auto ads for the lessons, excludes the practice area and the sign-in pages entirely, and turns off vignette ads because learners move through lessons by clicking next. After a week of daily screenshots the owner notices that an anchor ad overlaps the sticky lesson navigation on small phones, so the anchor format is switched off for lesson pages. No revenue figure is needed to make this decision, only a clear look at how readers use the page.

Density checklist

• Content begins before the first ad on every template.
• No page shows more ads than the reader can reasonably separate from the content.
• Task pages are excluded.
• Vignette frequency is low and task links are excluded.
• Anchor ads do not hide controls on small screens.
• Screenshots are taken weekly for the first month.

When to reduce auto ads

Reduce or stop an auto ads format when you see layout shift you cannot fix, reader complaints about a specific format, overlap with important interface elements, or a measured drop in engagement without a matching revenue gain. Turning off one format at a time makes the cause easy to identify.

Where to start with manual units

A three-slot starting layout

1. An in-article unit after the introduction of long pages.
2. A second in-article or display unit in the middle of very long pages only.
3. A display or multiplex unit after the content ends.

Short pages get the first or the last slot, not both. The point is a clear reason for each position, not a perfect count.

Choosing a unit type

The ad units page suggests choosing a more specialised unit when a plain display unit does not suit your site. Use in-article units inside text, in-feed units inside card lists and display or multiplex at the end. Name each unit after its template and slot, for example "lesson-after-intro", so reports tell you where income comes from.

Fixed or responsive

Display units can resize to fit the page or be fixed. Responsive units adapt to screen width but may shift content while loading. Fixed units are predictable but can overflow narrow screens. Reserve a minimum height for each slot and test on a small phone either way.

Making ads look like ads

Native-style units can be customised to match surrounding content, which is a design advantage and a trust risk. Keep clear spacing between content and ads and keep ad styling distinct from your editorial links and buttons.

Example scenario: a photography tutorial site

Example scenario: A photography site publishes long tutorials with many embedded images and short gear notes. The owner places one in-article unit after the introduction of each tutorial and one display unit after the final paragraph, and uses only the end-of-page unit on gear notes. Image-heavy tutorials get a fixed minimum height on each slot so the pictures do not jump when the ad loads. The mid-article slot is added later, and only to tutorials that are long enough that a reader can pass a natural break. The owner writes each template and slot name into a small table, which turns later reports into a quick lookup instead of guesswork.

Maintenance schedule

Common manual mistakes

• Reusing one unit code in several slots, which blurs reports.
• Adding units to one template but forgetting others.
• Leaving old ad code in place after adding new units, which duplicates ads.
• Putting a unit right above the headline so readers mistake it for content.
• Adding more units than the page needs because the first test earned a little.

A two-week test on matched page groups

Pick what you will measure

Choose three signals before you start: revenue per thousand pageviews, layout shift or loading behaviour on the same template, and a reader behaviour signal such as pages per session. Behaviour metrics are indirect hints, not proof of cause.

Step by step

1. Select two groups of pages that share a template, have similar traffic and come from similar sources.
2. Run auto ads on group A and manual units on group B. Exclude group B from auto ads so the methods do not mix.
3. Change nothing else on either group for two full weeks, so both weekdays and weekends appear.
4. Record the same numbers on the same day each week.
5. If time allows, swap the methods in a second round to cancel out any difference between the groups.
6. Compare the results side by side and decide.

What can spoil the test

• A site redesign or heading change in the middle of the test.
• One group receiving mainly search traffic and the other mainly social traffic.
• Too few pages, so one viral article drives the result.
• A holiday or sales event that shifts ad demand for everyone.

Handling seasonal and traffic swings

Ad demand moves with the calendar, and so does your audience. If the two weeks you choose contain a holiday, a school break or a large traffic spike from a single link, note it and consider extending the test by a week. Compare the groups against each other on the same days rather than against last month, because the groups share the same calendar. When your site is small, a single busy day can dominate the totals; look at daily numbers as well as the sum, and discount an outlier day only if you can name its cause.

Reading the result

If income is similar, choose the method that costs less upkeep. If one method earns clearly more but slows pages or hurts reader behaviour, weigh both sides, because short-term income gained by tiring readers can cost more later. A two-week test is directional evidence, not a verdict. Repeat it after major template or traffic changes, and keep a one-page note of dates, groups, numbers and the decision.

Result decision table

Test log checklist

• Both groups use the same template and similar traffic.
• No other change was made for two weeks.
• Revenue and speed were measured the same way on both groups.
• Unexpected events were written down.
• A person checked the result on a phone before the winner was chosen.

Related guides: Google Ad Manager for Small Publishers: Setup Guide, Choosing Affiliate Programs for a Blog: Decision Guide, Paywall or Ads? Choose a Model and Mark Gated Content.

Conclusion

Auto ads let Google choose placements and adapt to your edits, while manual ad units let you choose every position at the cost of template work. The help pages describe those mechanics, not a winner: they promise no extra income and make no claim about speed. So decide with three steps. Exclude pages where people complete tasks, audit density yourself on a phone, and compare the two methods on matched page groups for two weeks. Start manual units with no more than three slots, name each one by template and position, and limit any full-screen format. Menu names, formats and settings can change, so confirm them on Google's current help pages before you apply them.

If you want to monetise by offering your own placements to advertisers, you can look at the ad spaces on Akua Adforce. Whichever method you pick, keep your seller records current with the ads.txt guide.

Frequently Asked Questions

What is the difference between auto ads and manual ad units?

With auto ads you add one piece of code and Google places ads based on page layout and content. With manual units you create each unit in AdSense and paste its code where you want the ad. The difference is who decides the position.

Do auto ads earn more than manual units?

Google's help page says auto ads may find new ad locations and revenue could be higher, but it does not promise an increase, and the ad units page makes no revenue comparison. Measure both methods on matched pages in your own account.

Do auto ads slow down a site?

The help pages do not state an effect on page speed for either method. Speed depends on how many ads you show, which formats, and whether space is reserved. Compare layout shift and loading on the same template.

Can I stop auto ads from appearing on some pages?

Yes. The help page says you can exclude specific pages and specific areas of pages from auto ads. Excluding sign-in, checkout and form pages is a sensible default.

Can I use auto ads and manual units together?

The help page notes that auto ads detect existing Google ads on the page. It does not describe the revenue effect of combining them, so test that mix as its own variant.

How long should the comparison test run?

Two full weeks on matched page groups is a practical minimum because it includes weekdays and weekends. Change nothing else during that time and repeat the test after big template or traffic changes.