Should you run Google Ads yourself?

by Harry Watts · 27 August 2026 · updated 16 September 2026 · 7 min read

Yes, you can. Google spent twenty years making sure of it. Whether you should is a question about hours and habits, not intelligence, and the people selling management have an obvious reason to answer it for you.

I sell management, so read this knowing that. I also built my setup service specifically for people who want to run their own account, which only makes commercial sense if DIY is sometimes the right answer. Here's when it is.

The platform wants you to self-serve — on its terms

Google Ads is deliberately easy to start: guided campaign creation, AI-written ads, automatic recommendations, a credit voucher to get you going. The onboarding is excellent. The catch is what's switched on when you arrive.

New campaigns default to location targeting that includes people merely interested in your area, the Display Network pre-ticked, broad match encouraged at every turn, and auto-apply recommendations quietly enabled, which lets the platform change your own campaigns in whatever direction its algorithm favours. Every one of those defaults tilts toward Google's revenue. None of them is what a careful beginner would choose deliberately, and together they're a large share of the waste I find in audits.

So the honest framing isn't "DIY is risky". It's: DIY on defaults is expensive; DIY on a proper setup is fine.

What running it yourself actually takes

Two different jobs, usually conflated:

The build. Done once, and it's the hard 20%: campaign structure, keyword and match-type choices, ad copy, and above all conversion tracking that counts the right thing. Get the build wrong and the weekly work becomes bailing a leaky boat.

The running. The part people underestimate not in difficulty but in regularity:

• Weekly, 30–60 minutes, non-negotiable: the search terms report. Read what your money bought; exclude what should never have matched. Skip this for a month and waste compounds quietly.
• Weekly, ten minutes: spend pacing and anything anomalous.
• Monthly, an hour or two: budgets, bids, ad tests, and one honest look at cost per customer rather than cost per click.

Call it two to four hours a week for the first couple of months, tapering to one or two. The learning curve is mostly vocabulary (the acronyms guide exists for exactly this), plus the patience to judge the account at 60–90 days rather than week two.

When DIY is the right call

• Small budgets. On £400 a month of ad spend, a typical management fee doubles your cost of customer acquisition before a single click improves. The arithmetic simply doesn't clear.
• Simple offer, tight geography. One service, one area, one campaign: the account a competent owner can genuinely run well.
• You have the hours and some appetite. Not enthusiasm for marketing, just tolerance for a weekly routine and reading a report.
• You're still proving demand. Early on, the fast feedback of running it yourself teaches you things about your market no report conveys.

When it stops being the right call

• The arithmetic flips. A specialist typically recovers 15–30% of spend in an account that hasn't had expert attention. Once fee < recovered waste + the value of your evenings, paying is the cheaper option.
• Complexity jumps. Shopping feeds, Performance Max pressure, multiple locations, real budget growth. Each multiplies the ways to be quietly wrong.
• The plateau you can't explain. When performance stalls and you've run out of hypotheses, more of your hours won't fix it.
• You've stopped doing the weekly work. No judgement: it's the most common ending. But an unattended account is the most expensive kind, and admitting it beats funding it.

If it's the last one and an agency is already involved, check what you're paying for before assuming the answer is more outsourcing.

The middle route nobody advertises

The choice isn't binary. The version I see work most often for small accounts: pay for the build, run it yourself. A specialist does the one-off hard part (structure, tracking, the defaults corrected, a short handover on the weekly routine), and you do the part that mostly needs consistency rather than expertise.

That's precisely what my account setup service is: a proper build with tracking included, handed over, no retainer required. The companion move for accounts you've already built is a one-off audit: a specialist marks your homework once, you carry on. Both cost a fraction of a year's management, and neither locks you in.

One warning for this route: percentage-of-spend pricing punishes exactly the small accounts most suited to it. If you do buy help at DIY scale, buy it flat-fee and one-off.

If you do run it yourself

• Turn auto-apply recommendations off on day one
• Location options to presence only; untick the Display Network on Search campaigns
• Only real outcomes set as primary conversions: no page views, no button clicks
• The weekly search-terms half hour, in the diary, treated as unmissable
• Judge at 60–90 days against cost per customer; resist re-engineering weekly

Quick answers

Want the build done properly, then handed over

That's the account setup service: structure, tracking and corrected defaults, then it's yours, with no retainer attached. Already built it yourself and want it checked? That's the audit. A free intro call will tell you which, and if the honest answer is "keep doing it yourself", that's the answer you'll get.