Every large Google Ads account has a list of jobs nobody enjoys. Check whether each campaign is on pace for the month. Shift budget from the ones that are behind to the ones that are earning. Sort a few thousand new search terms into sensible groups. Build the same set of ads forty times with one word changed.
These jobs matter. They are also dull, which is why they get done on a Friday afternoon, or the week after they were needed, or by whoever drew the short straw. The account doesn’t fail because of it. It just drifts, a little further each month, and the drift costs money that nobody can point to.
Rules are for machines
Most of that list is rule-following. If a campaign is ahead of pace, slow it. If it is behind and still earning well, give it more. If a search term contains these words, it belongs in that group. A person applying those rules by hand is doing a machine’s job, slowly, and with a person’s habit of skipping it when something more interesting turns up.
Google lets you run your own scripts inside an ad account. Most accounts never use them, or use a free one copied from a blog that nobody on the team can read. That is a missed opportunity. A script written around how your account is supposed to run will do the dull work every day, in the same way, without being asked.
What a good one does
The largest one I’ve built balances a national account by itself. It keeps monthly spend on pace, and it moves budget towards the campaigns with the best return on the next dollar, not just the best return so far.
The useful part is what it does before it raises a bid. It checks why the campaign is being held back. If the bid is the limit, it raises the bid. If the limit is the quality of the ad or the landing page, more money won’t fix that, so the campaign goes on a fix list for a person to look at. A lazy rule would have outbid the problem and called it optimisation.
Others are simpler. One sorts thousands of keywords into groups. Another produces bulk ad builds ready to import.
Where it goes wrong
A script is not a way to stop paying attention. I’d be wary of any of these:
- A script nobody can explain. If the person who wrote it has left and nobody knows what it changes, it is a risk with a schedule.
- Rules that were never written down. The script is only the last step. The useful work is agreeing what the account should do in each situation.
- No record of what it did. Every change should be logged somewhere a person will read.
- No limits. A script should have a ceiling on how much it can move in a day, so a bad input can’t empty a budget.
What I’d do
Start with the job your team likes least and does most often. Write down the rule you’d want a careful new starter to follow. That document is worth having even if no script is ever written, because it usually shows that three people have been following three different rules.
Then have the rule built, with limits, a log and plain notes on how to change it. The script should belong to you, sit in your account and be readable by the next person.
What you get back is not a smaller team. It is the same people spending their week on the questions a script can’t answer, such as whether the account is chasing the right thing in the first place.