Gareth Puckey Say hello

Marketing automation

For anyone running paid media, marketing automation means getting software to do the routine checks, changes and reports on your ad accounts, so that wasted spend and mistakes are caught before they cost you sales. In practice, PPC automation comes down to a handful of tools: Google Ads scripts, automated rules, shared exclusion lists, scheduled reports and CRM syncs that feed sales back into the platforms. This page explains what each one does, why it matters for performance, where AI agents are taking it, and what to automate first (platform details correct on 30 October 2026).

By Gareth Puckey Google Ads and GA4 certified Updated 17 sources

What it is

The platforms run plenty of automation of their own, which I come back to below. PPC automation, as I use the term here, is the layer you set up and control yourself: the jobs that watch your accounts, make changes you have agreed in advance, and send the right numbers to the right people on a schedule.

The term “marketing automation” is also used for email and lead nurture software. That is a separate subject, and this hub doesn’t cover it. Here it means automation for ad accounts and reporting.

There are five main building blocks.

Google Ads scripts

Google Ads scripts are JavaScript programs that run inside Google Ads, written in a browser-based editor. They can manage campaigns, ad groups, keywords and ads, change bids, budgets and status, and run reports using Google Ads Query Language. They can be scheduled to run hourly, daily, weekly or monthly, and can read and write Google Sheets, send email and call outside services (Google for Developers).

From a manager account, one script can work across child accounts, up to 50 at a time in parallel. A standard run gets up to 30 minutes, or up to 60 minutes for a manager script using executeInParallel with a callback. Scripts run on Google’s infrastructure at no extra cost, and you can test them in preview mode before they touch anything (same source).

Automated rules

Automated rules are the no-code option. In Google Ads, a rule can change ad status, budgets and bids when conditions you choose are met, or simply send an email when they occur (Google Ads Help).

Meta Ads Manager has the same idea. Its rules check campaigns, ad sets and ads, then update them or notify you, and they apply at ad account level, so they keep working whoever on the team builds the ads (Meta Business Help Centre).

Shared exclusion lists

A shared list holds negative keywords or placement exclusions in one place and applies them to many campaigns or accounts. A Google Ads manager account can hold three placement exclusion lists of up to 250,000 exclusions each, applied at account level to the accounts beneath it. It can also hold 20 negative keyword lists of up to 5,000 keywords each (Google Ads Help).

The automation here is keeping the list current. My placement blocker reviews where Performance Max and display ads ran and blocks junk apps, sites and videos on one shared list, Monday to Friday.

Scheduled reports

Reports can be built once and then sent automatically. Google Ads lets you schedule a saved report from the report editor to be emailed to people with access to the account, daily or weekly, as a CSV or other file format (Google Ads Help). For raw data, Google points analysts to the BigQuery Data Transfer Service, which loads Google Ads data into BigQuery on a schedule without code (Google for Developers).

In my own work, monthly GA4 reports for 36 websites are drafted automatically, then checked by a person before sending.

CRM syncs

A CRM sync sends what happened after the click (a qualified lead, a sale) back to the ad platform and into your reports. Google’s Data Manager lists direct connections for offline conversion imports from CRMs including HubSpot, Salesforce and Zoho CRM (Google Ads Data Manager Help).

CRM data is connected in my setup, so that revenue can be traced back to marketing. Without that link, the other automations on this page can only judge campaigns by leads.

For worked examples with the limits of each, see my write-up of five Google Ads automations that pay for themselves.

Why businesses need it

Automation earns its place through performance: lower cost per sale, less money spent on traffic that was never going to buy, and faster fixes when something breaks. Time saved is a side effect.

Catching mistakes before they run for days

Ad accounts can spend quickly when something is set wrongly. On a given day, a Google Ads campaign can spend up to twice its average daily budget, and up to 30.4 times that average over a month (Google Ads Help). A budget or bid typed wrongly on a Monday can do real damage before a weekly review.

A daily check of mine flags manual CPC bids above a £5 threshold. Its job is to catch a bid entered as £50 instead of £5.00 the next morning, before it has run for a week.

Catching waste that hides in reports

Some waste only shows up if you go looking for it. Search terms that spend without converting sit in a report that is easy to leave unopened for weeks. Junk app and site placements in Performance Max and display campaigns are just as easy to miss.

My weekly search term report lists terms that spent without converting, with a suggested negative for each. It also checks for keywords blocked by conflicting negatives, which is a quieter kind of loss: a keyword you want to show for that can’t.

Keeping up across accounts

My automations check 31 Google Ads accounts every weekday. Checking that many by hand every morning isn’t realistic, and the dull checks are the first to slip in a busy week.

Watching what the platform changes for you

Google can also change your account automatically if auto-applied recommendations are switched on. The types it can apply include adding broad match keywords, expanding to Google search partners, turning on Display Expansion and adjusting CPA or ROAS targets (Google Ads Help). Each of those can be right for one account and wrong for another, so you need checks that tell you what changed and what happened afterwards.

Judging by sales

My view is that automation is only as good as the number it optimises towards. An automated rule that cuts the campaign with the highest cost per lead may be cutting the campaign with the lowest cost per sale. A CRM sync is what lets your checks and reports see the difference.

Where it’s heading (2026 onwards)

Scripts and rules follow instructions a person wrote. The change in 2026 is AI agents that suggest changes and, increasingly, make them.

  1. Platform automation keeps changing

    What the platforms automate for you changes too. From 26 January 2026, Google’s ‘Add responsive search ads’ recommendation stopped suggesting or applying new responsive search ads automatically, and was removed from the auto-apply settings page (Google Ads Help). Settings you agreed to a year ago may not do the same thing today.

  2. Google: Ask Advisor

    On 20 May 2026 Google announced Ask Advisor, an AI agent that works across Google Ads, Google Analytics and Merchant Center. It is in beta for English language accounts, with more features promised in the coming months (Google). Google’s product page says it can create Performance Max campaigns and apply bid and budget recommendations in one click (Google Business).

    The Google Ads Help Centre says you can use it to “implement the suggested changes with your approval”. Its examples include pausing or renaming a campaign, changing a daily budget, changing location targeting, appealing disapproved ads and updating final URLs. It isn’t available in manager (MCC) accounts (Google Ads Help). Those account actions were reported as new in September 2026 (PPC News Feed).

  3. Outside agents connecting to ad accounts

    AI tools outside the platforms are being given access too, through the Model Context Protocol (MCP).

    • Google Ads MCP server. It lets an AI agent query Google Ads data in plain language. In its current release it is read-only: it can’t change bids, pause campaigns or create assets (Google for Developers).
    • Meta’s ads MCP server. On 16 July 2026 Meta opened it to any developer with a Meta app. Its tools can create, edit and delete campaigns, ad sets and ads. Rules control what an agent may do on each ad account, for example reading performance data without being allowed to change budgets. Meta’s own example is starting an account in a read-only or approval-required setting and widening permissions as confidence grows (Meta for Developers).
  4. Ongoing

    Why approval steps and audit trails matter more

    Google Ads change history lists changes from the past two years, shows who made each one, and labels changes made through the API with “Google Ads API” or the tool’s name. Google also notes that not every account-level setting change is listed (Google Ads Help).

    My view is that once agents can change budgets and pause campaigns, three things matter more than they did: a clear rule about what an agent may do without asking, a named person who approves everything else, and a log of what changed, why and who approved it, kept outside the platform as well as inside it. I also expect scripts and rules to stay useful, because the checks that catch a person’s mistake will catch an agent’s mistake too.

Common mistakes

Letting automation make changes nobody reviews

Auto-applied recommendations, rules that change bids and scripts that write to accounts all act without asking unless you build the step in. I’d have scripts suggest changes into a Sheet for a person to approve. My monthly bid adjustments (device, age, gender, location, schedule) are set from 180 days of results, with a dry run first.

Alerting on normal behaviour

A “spend above daily budget” alert will fire on ordinary days, because campaigns can spend up to twice the daily average (Google Ads Help). Alerts that fire too often get ignored, including the one that matters. Pace spend against the month instead.

Skipping the test run

Preview mode exists for a reason. If a script times out, Google Ads still applies any changes it made before it was stopped (Google for Developers), so a half-finished run can leave an account in a state nobody planned.

Building without checking the limits

An account-level negative keyword list holds 1,000 keywords (Google Ads Help), and a manager account gets only three placement exclusion lists (Google Ads Help). An automation that adds every suggestion will hit those ceilings quickly. Plan how you will split lists before you build.

Relying on fixes that no longer work

Old tricks get carried from account to account. Google Ads no longer supports the adsenseformobileapps.com placement for app inventory, although it still shows up in placement reports, and individual apps can be excluded instead (Google Ads Help). An exclusion list built around the old placement won’t keep ads out of apps. Check that each rule still does what it was built to do.

No named owner

If nobody’s name is on an automation, nobody notices when it breaks or starts making bad calls. The same rule applies to AI tools, which I cover in using AI in a marketing workflow without losing control.

How to get started

  1. List what you check by hand. Write down the checks you do each week and the mistakes that have cost money in the last year. Pick the one that cost the most.
  2. Build it as a flag first. Start with something that changes nothing: a scheduled report, an automated rule set to email only, or a script that writes findings to a Sheet.
  3. Test it before you trust it. Run scripts in preview mode, dry-run any bulk change, and compare the output against what you would have found by hand.
  4. Put a name and a log on it. Decide who approves changes and where they are recorded, and look at change history each week to confirm only expected changes went through.
  5. Connect your sales data. Sync your CRM so that checks and reports judge campaigns by sales as well as leads.
  6. Only then let it act, in small steps. Give it narrow rules and limits, review what it caught and what it got wrong each month, and give any AI agent read-only access before anything else.

Guides on Marketing automation

Sources

17 sources, all opened and checked
  1. Google Ads scripts product overview, Google for Developers
  2. Set up automated rules, Google Ads Help
  3. About automated rules in Meta Ads Manager, Meta Business Help Centre
  4. About your Google Ads account limits, Google Ads Help
  5. Create and manage reports, Google Ads Help
  6. Supported data sources, Google Ads Data Manager Help
  7. About overdelivery and your average daily budget, Google Ads Help
  8. About applying recommendations automatically, Google Ads Help
  9. Meet Ask Advisor, your new AI-powered collaborator, Google
  10. Ask Advisor in Google Ads, Google Business
  11. Ask Advisor in Google Ads (beta), Google Ads Help
  12. Google Ask Advisor can pause campaigns, change budgets and appeal ads, PPC News Feed
  13. Google Ads MCP server: developer integration guide, Google for Developers
  14. Meta’s ads MCP server is now available for developers, Meta for Developers
  15. About change history, Google Ads Help
  16. About account-level negative keywords, Google Ads Help
  17. Show your ads in mobile apps, Google Ads Help