Google Ads Management Checklist: Why One List Never Fits Every Account [2026]

Why the checklist you downloaded doesn't help

Checklists on the internet exist to be downloaded. That's their entire job.

Take eCommerce - say, two skincare brands. Same products, face and body care, same audience. One runs on Shopify, the other on its own platform. Those are two completely different jobs: the product feed works differently, the buyer's path through the site looks different, measurement works differently. A list that fits the first one will do nothing for the second.

There's no three-hundred-question document that hands you a great checklist for one account and then for every other account you run. Every account needs its own list: what it has to have, and how it has to be set up technically.

The worst advice I've seen in other people's checklists isn't wrong. It's generic. And whatever has nothing to do with your business isn't relevant to you, no matter how good it sounds.

A checklist does make sense when it's built for one channel, one purpose - see the Performance Max audit checklist, or its interactive version, PMax Check. The difference is that checklist doesn't claim to work for Search, for Shopping, or for someone else's account - it works exactly for what it was built for.

What actually works: rhythm, not a list

The question "how many times a week should I touch the account" is framed wrong. There's no number that answers it.

Every account needs its own baseline. You know exactly what gets done on it every day, every other day, every week, every other week, and once or twice a month. Once that's set, you stop asking yourself what to do today.

I ran an account with a six-figure monthly budget, over forty Shopping campaigns and over forty Search campaigns, because we were targeting several markets at once. That account performed well. It didn't perform well because something happened overnight - it performed well because systematic, well-organized things got done every day and every week. The result was a consequence of the rhythm.

For years I kept that by hand. Open a project management tool, any one, click into a task or a plan, and you know what's waiting for you on that account that week. Today, a system I built delivers most of those checks for me. What's left for me has moved up a level - a decision, not a check.

How that rhythm breaks down into concrete weekly and monthly optimizations, week after week, is what I cover in the Google Ads optimization guide.

Daily: five minutes, three things

When I open an account in the morning, I don't open a report. I look at three things.

Most people do damage here by changing something that's measured in weeks based on yesterday's number.

The wider context - how this daily check fits into a full account audit - is covered in the Google Ads audit checklist.

The threshold that decides everything else: conversion count

This is the one point where I agree with Google's official position.

A new account needs to sit. Roughly ninety days before you take conclusions seriously.

After that, everything depends on one number - how many conversions the account produces per month.

The same move on those two accounts isn't the same move. That's why "how often should I change the bid strategy" is a question with no universal answer, and any checklist that gives you a number is lying.

How the choice of bidding strategy shifts depending on that stage of the account is what I break down in the Smart Bidding guide.

Search terms: the one data point that actually paints the account

For years, most of my attention on Search and Shopping campaigns has stayed on the search terms report. The reason is simple.

We don't know who the person behind the click is. We know rough location, we know demographics, we know whether they bought. But search terms is the only data point that tells us what that person actually wanted. It's the one that paints the whole account in different colors.

The order I look in:

The classic rule still holds: a search term that isn't in the account at all, has good-quality clicks and impressions, and hasn't converted in the last two weeks - goes in as a keyword so you can control it. And the reverse: a term that spends money and does nothing goes to negatives. How to keep that list running as a system rather than a one-time list from launch is what I wrote about in the negative keywords guide.

One more thing most people don't do: pull search terms for the last six months, then the last year, then two years. You're not looking for what's happening now. You're looking for what you missed.

And to be honest all the way through: even though a system handles this for me today, I still catch myself pulling everything by hand and checking it. I don't yet have the confidence to rely on automation here with my eyes closed.

What you don't touch

Recommendations from the Google Ads panel. I've ignored every single one of them for several years now. They're not technically wrong - they're written without a shred of insight into the account's context.

Here's a dumb but honest example. You have a branded Search campaign that has to exist, because it protects your brand from competitors. It runs on manual CPC and the click costs you twenty cents. The panel suggests switching to target ROAS. You switch, and the click now costs you three dollars. The system didn't get the math wrong. The system simply didn't know why that campaign exists.

That's the entire flaw of automated recommendations in one sentence.

The bid strategy, three days after a change. See the section on the conversion threshold above.

Anything measured in weeks that you're about to change based on a single day.

Quarterly: what doesn't make sense to review more often

I review ads and extensions on a longer horizon. We moved away from the old ad-writing formats a long time ago; today Responsive Search Ads is where all of that consolidated, so the evaluation also has to run on a bigger sample.

The question I ask there: is it worth keeping two genuinely different RSAs in an ad group, and what do I do with that test once it's done.

A few years ago I ran experiments far more often. Today I run one, at most three, over six months. Not because they stopped being worth it, but because most of the experiments I used to run didn't have enough data for the result to mean anything.

Where the money actually leaks

When an account falls apart, what's on the checklist is rarely to blame.

The pattern that repeats: conversions are measured wrong, and a bid strategy built on that measurement makes decisions every hour. Double counting, add-to-cart set as the primary conversion instead of purchase, a conversion that counts every time instead of once. The system dutifully optimizes toward a number that doesn't exist - and does it perfectly.

The second family of mistakes is even more mundane. A final URL that points to the wrong product. A sitelink pointing to a product you no longer sell. None of that shows up in a performance report - you only see it when you click your own ad.

Once you've been through this - and every senior has - you stop hunting for the mistake by list and start hunting for it by pattern.

The business owner with thirty minutes a week

I'll be uncomfortably honest here, because the alternative is selling you something that doesn't work.

An owner running their own account with thirty minutes a week can't do anything serious at all. That's not a discipline problem - it's a function of how much work the account actually requires.

The one exception: if you spend very little and get up to about a hundred clicks a week. Then in those thirty minutes you can check four things:

Anything beyond that needs either more time or someone doing it for you. I believe that will change - tools that do serious analysis fast already exist and are becoming more accessible. But today, for an account spending real money, there's no solution that fits into thirty minutes a week.

Frequently asked questions