TrailGear Co. sells running shoes, cycling gear, and hiking equipment online in the US. Their Google Ads account had three campaigns running with no real structure — keywords mixed together, no clear ad groups, and costs going up every month.

Here's how we rebuilt it using Claude and ADS MCP in one afternoon.

Step 1: Find the keyword gaps

First, we asked Claude to look at what people were already searching for on Google and check whether TrailGear Co. was running ads on those searches:

"Check Search Console top queries for the last 90 days and find which searches have over 500 monthly impressions but no active keywords in our Google Ads campaigns."

Claude checked both Search Console and Google Ads and returned a list. The biggest gaps:

  • "trail running shoes men" — 3,600 searches/month, no ads running
  • "waterproof hiking boots women" — 2,100 searches/month, no ads running
  • "clipless cycling shoes" — 1,400 searches/month, no ads running
  • "hydration vest running" — 900 searches/month, no ads running

People were already searching for these products — TrailGear Co. just wasn't paying to show up.

Step 2: Check how competitive each keyword is

"Get Keyword Planner data for these four terms and suggest 5–8 related keywords per term."

Claude returned monthly search volumes, how many other advertisers were bidding on each term, and how much each click typically costs. "Trail running shoes men" had a lot of competition — worth its own campaign. "Hydration vest running" was cheaper and less competitive — good to group with other accessories.

Step 3: Build the campaigns

"Create a Search campaign called 'Trail Running Shoes — US' with 4 ad groups: men's trail running shoes, women's trail running shoes, waterproof trail shoes, and trail running shoes reviews. Exact match and phrase match for each group. Write 3 ads per ad group. Daily budget $40, Max Conversions bidding."

Claude created the full structure — 4 ad groups, 15–20 keywords each, 12 ads — and showed a summary for review. One confirmation and the campaign was created in Google Ads, paused, ready to check before going live.

We did the same for the other three product areas. Four campaigns, 16 ad groups, all built in about 40 minutes.

Step 4: Set up bidding correctly

These were new campaigns with no sales data yet, so we used Max Conversions — this tells Google to find as many conversions as possible without worrying about cost per sale yet. Setting a cost target before you have data doesn't work — Google needs real conversions to learn from first.

After 3 weeks, we asked Claude:

"Check conversion data for the Trail Running Shoes campaign for the last 21 days and tell me if we have enough to set a Target CPA, and what to set it at."

Claude came back: 67 conversions, average cost per sale $18.40. Recommendation: set a target of $19 to start, then lower it slowly over time. That's what we did.

Why having separate ad groups matters

The old account had one campaign with 50 mixed keywords and no structure. Google couldn't tell the difference between someone ready to buy trail running shoes and someone just doing research. Ads showed up for the wrong searches, clicks were wasted, and costs kept rising.

When each ad group covers one specific type of search, Google gets better data and learns faster which searches lead to sales. The ads also match what people are looking for more closely, which means lower costs and better placement.

The rebuilt account had a 34% lower cost per sale in the first 60 days — not from any trick, just from giving Google cleaner data to work with.

All campaigns are created paused by ADS MCP — nothing goes live or spends until you enable it manually in Google Ads.