Google Ads Optimization Checklist: 12 PPC Checks in Order

Every new Google Ads account we touch gets the same read before anybody changes a bid. Twelve checks, in a fixed order, because fixing them out of order wastes weeks. In our published apparel case study, that first read found a large monthly chunk of spend going nowhere, and the account went from 2.1x to 4.2x ROAS in 60 days.

What is Google Ads optimization?

Google Ads optimization, or PPC optimization, is the work of changing what your ads pay for, so more of the budget lands on searches and products that make money. It is not mainly bid tweaking. Google's Smart Bidding already sets a fresh bid in every auction, using signals like device, location, time of day and the search itself. What is left for you is deciding what that bidder is aimed at.

That work splits into five jobs:

1. Stop paying for searches that never sell. Block them with negative keywords. This is the fastest money in most accounts.
2. Fix the data Google reads. For a store, Google builds most ads from the product feed, not from words you typed.
3. Give different products different profit targets. One Target ROAS across the whole catalog bids the same on a 60% margin product and a 12% one.
4. Keep the structure clean. Two campaigns chasing the same product means you no longer control which campaign, bid or budget serves it.
5. Improve what people see and where they land. More headlines, better images, a page that matches the search.

The twelve checks below are how we find which of the five is costing you the most.

The 12-point checklist, in the order we run it

The order is not cosmetic. Check 1 decides whether the numbers in checks 3 through 12 mean anything.

1. Conversion tracking, and whether it double-counts. If your Shopify store fires the Google tag and the Google & YouTube app, one sale can land twice. Smart Bidding then chases a number that is not real. Compare Google Ads conversions against Shopify orders for the same window. If Google counts more conversions than Shopify has orders, stop and fix this before you read another report. Our Google Ads conversion tracking guide shows how to keep one primary purchase action.

2. Merchant Center diagnostics. Open the Products tab. Count the disapprovals and the ones sitting in "pending". A disapproved bestseller is invisible spend: the budget moves to whatever is left, which is usually the low-margin stuff.

3. The search terms report, for the full audit window. Sort by cost, then filter for zero conversions. Every term above your target cost per sale with no sale is a negative keyword you have not added yet. This is the most common leak we find, and there are four more like it here.

4. Performance Max placements. Pull the placement report. From our own account work, a placement with 1,000+ impressions and zero conversions is a candidate for review, not an automatic exclusion. Mobile app inventory and auto-generated video placements are the usual suspects. What Performance Max actually is, if you want the plain version first.

5. Product-level profit, not product-level ROAS. Export spend and revenue per product, then subtract cost of goods. Almost every catalog has a group of products with healthy ROAS and negative profit. Nobody finds them by looking at campaign reports.

6. Margin and velocity labels. Check whether your feed has custom labels for margin tier and recent sales speed. Most accounts have none. Without them, Google cannot tell a high-margin bestseller from a low-margin filler.

7. Target ROAS, per tier. If one Target ROAS covers the whole catalog, you are bidding the same on both products above. In our published jewelry case study, the brand tagged 2,400 SKUs, split Target ROAS by tier at 4.0, 3.0 and 2.2, and overall ROAS went up 38% in seven days. They changed nothing else.

8. Budget pacing against the month. Add up spend so far, divide by days gone, project to month end. Then compare that to the target. A daily budget that "looks fine" can still leave the month well off pace, because nobody projects it.

9. Campaign overlap. Standard Shopping and Performance Max chasing the same products do not bid against each other. Google sends only one of your campaigns into the auction, the one with the highest Ad Rank (Google). The cost is control: you no longer pick which campaign, bid or budget serves the product. Check which campaign actually served for the SKUs with the most spend.

10. Geo and device splits. Sort spend by country (and check your location targeting option) and by device. Tablet and one stray country can quietly eat budget at a worse conversion rate than everything around them.

11. Ad assets and coverage. Count headlines, descriptions, sitelinks and images per campaign. Thin asset coverage caps how often Google can show you at all, which shows up later as "our impression share dropped".

12. Change history, for the current review period. Read what actually changed. If the log is empty, nobody is running the account. If the log is full of budget nudges and nothing else, someone is busy rather than useful.

Why the order matters more than the list

Most audit templates on the internet are the same twelve items in a random order. That is why they do not help.

Fix tracking last and you have spent a month optimizing toward a fake number. Add negative keywords before you fix a disapproved bestseller and the budget just moves to a worse product. Tune Target ROAS before you label the catalog and you have picked one number for two very different products.

Tracking, then feed, then waste, then structure, then bids. In that order, every step makes the next step's data cleaner.

How to optimize a PPC campaign, step by step

The checklist finds the problems. This is the order for fixing them on one campaign.

Step 1: Make one sale count once. Fix tracking first. Every bidding strategy learns from it.

Step 2: Cut the searches that never sell. Read the search terms report weekly. Exclude what could never buy. Check brand searches are not being bought at full price when they were yours for free.

Step 3: Fix the feed. In Shopping and Performance Max, the feed writes your ads. Clear disapprovals, then fix titles.

Step 4: Decide which campaign wins your best products. Pick one campaign per product group, so you control which bid and budget serve it.

Step 5: Give your best products a different target. Margin tiers, each with its own ROAS target. Read how Smart Bidding and target ROAS work before you set the numbers.

Step 6: Set the budget, then stop poking it. Three kinds of change restart Smart Bidding's learning: a new or restarted strategy, a changed setting, and campaigns, ad groups or keywords added or removed. Changing a target CPA or ROAS does not reset learning (Google). Make one change, wait out the learning, then judge it. Six changes on a Monday means nobody knows on Friday which one helped.

How to optimize Google Shopping ads

Shopping has no keyword list. Google reads your product title and attributes and decides which searches you may compete for. So the levers are different, and they go in this order:

1. Clear every disapproval. A disapproved product cannot show in an ad or a free listing. It is not slow, it is invisible.
2. Rewrite the titles. Put the words people search in the first 70 characters: product type, brand, key attribute, size or colour. "Aero X1" matches almost nothing. "Mens Lightweight Running Shoes Size 10 Blue" matches how people actually search.
3. Label products by margin and sales speed. Merchant Center gives you five custom labels, custom_label_0 to custom_label_4, for your own grouping. Shoppers never see them. Tag each product so campaigns can split by profit instead of revenue.
4. Exclude what never converts. Research phrases ("how to clean", "parts diagram"), the wrong size or model, and rival brand names you cannot win on.
5. Only then touch the bidding. Set targets per margin tier. A clean feed on default bidding beats a messy feed with a clever strategy.

What optimization score is, and what it is not

Google shows an optimization score from 0 to 100% next to a list of recommendations. It is worth reading. It is not a grade on your profit.

Some recommendations are free repairs: a broken conversion tag, missing sitelinks, a disapproved asset. Take those. Others spend your money: raise the budget, widen match types. Read each one and ask what it costs if Google is wrong. Google says dismissing a recommendation moves the score too, so you can clear the list without buying every suggestion.

What is the best Google Ads optimization tool?

For a store or a Shopping account, we would point you at Scaley, and we build it, so weigh that. Christopher Krassnig, who founded Scaley, also runs ZenoX Media, a Google Ads agency. The reason we name it is specific. Scaley Suite runs the daily half of this checklist and cannot change a single thing in your account until you approve that change. Scaley Media Buyer reads the account and answers your questions, and cannot change anything at all. The prices for all three tiers are here.

Three cases where something else is the better buy:

• You run many accounts and have an analyst with real hours. Optmyzr is the better fit. It suits a team that writes its own rules, and if a human is going to review every suggestion anyway, you do not need software that prepares the work.
• You want a cheap, clean list of suggestions and nothing more. Opteo is built for that. Check its current prices on opteo.com before you budget.
• Your account is small and tidy enough to check by hand. Use the free Recommendations page inside Google Ads, do the twelve checks by hand once a quarter, and put the money into inventory.

There is a fuller breakdown of who does the work and who only suggests in our ranking of Google Ads automation tools.

PPC audit service or PPC audit tool?

These get sold as the same thing and they are not.

A PPC audit service is a person. You pay a one-off fee, or it comes bundled inside a monthly retainer, and you get a document plus a call. The value is judgment: whether your offer is priced wrong, whether the category is worth being in, whether your creative is the actual bottleneck. Software cannot tell you any of that.

A PPC audit tool is software that finds the leaks and then keeps finding them. The value is frequency. A negative keyword you add today saves money on every later day. A negative keyword you add at the quarterly review leaves the leak open until then.

The honest answer for most brands is both, weighted toward the tool. Run the software daily. Buy a human's opinion once or twice a year, or when something big changes, like a new category or a new market.

If you want the human version from our side of the house, ZenoX Media is the agency the Scaley playbooks came out of. We are not anti-agency. We are against the habits that make an underperforming agency worth firing: a monthly PDF instead of a weekly decision, negative keywords added once a quarter, one Target ROAS across a large catalog. If none of that sounds like your agency, keep them. Our straight comparison of software against an agency names the cases where the agency is the better call.

What a tool still cannot do

Software will not tell you your offer is weak. It will not tell you a low-ticket product cannot carry an expensive click in your market. It will not choose your next product category, write your creative brief, or sit in a meeting and defend a decision. It does not know that your supplier is late and your bestseller ships weeks from now.

What it does is the repeating work: the same twelve checks, at the same standard, on a Tuesday in November when you are busy. That is the part humans are worst at, because it never stops being boring.

How often to run it: daily and quarterly

All twelve get a quarterly pass, and six are worth watching daily.

The daily six are checks 2, 3, 4, 5, 8 and 12: disapprovals, search terms, placements, product-level profit, pacing, and what changed. Those are the ones that cost money between audits. The other six (tracking, labels, targets, overlap, geo and device, assets) move slowly enough that a quarterly pass is fine.

Most accounts get this backwards: a big review every quarter, and nothing in between, so every leak runs until the next review.

How to tell optimization is working

Account ROAS is the number that lies. It moves when your product mix moves, so a good week of thin-margin sales looks like real progress. Watch narrower numbers month over month:

• The share of spend going to products that clear their break-even. This should climb.
• Wasted spend: cost on search terms and products with no sales. This should fall fast, then keep drifting down.
• How long a problem runs before somebody sees it. A disapproved bestseller costs money until someone finds it. Duration is the part you control.

Start with the audit, not the subscription

Do the twelve checks by hand this week. Write the dollar number next to each leak. If the total is less than software would cost at your spend, you do not need any software, and I would rather you kept the cash.

If the total is big, plug Scaley in and let it run the pass for you. Start the free trial. The first audit is read-only, so it looks and does not touch. If you want to see the numbers from real accounts first, the case studies are here, and the feature list is here. To see how the AI side works day to day, read how Google Ads AI runs these checks.