How restructuring an auction account by intent 6x'd bid completions
An Australian auction marketplace had 80+ campaigns split by state and division, spending hard on weekly auctions but not converting. I cut it to 32, restructured by buyer intent. Bids 6x'd.
Results Achieved
The Challenge
The account was one of Australia's largest online auction marketplaces, running weekly 7-day auctions across every category imaginable. Cars, industrial equipment, wine, boats, real estate, jewellery, everything.
On paper it looked serious. 80+ live campaigns, big monthly spend, sophisticated-looking segmentation. The structure split everything by division (Auto, Industrial, Wine, Boats) and then again by state (NSW, VIC, QLD and so on). Sensible in theory. In practice, most campaigns were too small to learn from, budgets were spread thin, and no single campaign had enough conversion data for smart bidding to work.
The weekly auction campaigns were the biggest problem. They were spending hard, driving impressions and clicks, but bid completions weren't showing up as conversions in Google Ads. The account was optimising against noise while the actual auction activity ran silent.
The Solution
Step 1: Rebuild the segmentation around intent, not geography
I stripped out the division-and-state grid and rebuilt the account around the two things a Google Ads campaign actually needs to know: what is this user trying to do?
- Buying/Bidding intent: users looking to buy from or bid on the auctions. Ecommerce campaigns, product-driven, optimised for Bid Complete.
- Selling intent: users looking to consign or sell through the platform. Lead-gen campaigns, form-driven, optimised for Sell Form submissions.
That single change did more than any keyword or bid tweak could have. Each campaign now had a clear job. The algorithm had a clear signal. The account had two coherent halves instead of 40 fragmented ones.
Step 2: Cut ruthlessly from 80+ campaigns to 32
Most campaigns weren't earning their spot. They existed because someone had once thought 'we probably need a Wine Display campaign for Victoria,' and it had never been killed. I consolidated aggressively. 32 campaigns instead of 80+. Each one big enough to actually feed the bidding algorithm.
Step 3: Deliberately deprioritise Buy Now to focus on the auctions
The account had a Buy Now conversion type running alongside the auctions. It was cannibalising attention and budget from the core business, which is the auction model. I made a strategic call to pull spend out of Buy Now and route it into the auction funnel. Buy Now conversions dropped 53%, deliberately, while Bid Completions grew 6x.
Full Case Study
The story in one chart
Same account. Fewer campaigns. Less spend. Six times the bids. What changed was that every campaign finally had a clear job.
The account when I inherited it
80+ live campaigns segmented by division and state. Weekly 7-day auctions running across every category. Big monthly spend, but the bid completions weren't showing up as conversions. The algorithm was optimising against clicks and page views. The real activity was running silent underneath.
The restructure
I did two things at the same time. Cut the campaign count from 80+ to 32, and re-segmented what was left around intent instead of geography. Buying and bidding on one side, selling on the other. Two clear jobs, two clear signals, no more fragmented budgets.
Spend down. Bids up.
Comparing my 30 months on the account (1 August 2023 to 31 January 2026) against the equivalent 30 months before I arrived (1 February 2021 to 31 July 2023), spend dropped by roughly $990k while bid completions grew by 1.25 million.
Cost per bid: down 88%
The efficiency gain came from three places at once. Fewer, larger campaigns for the algorithm to learn from. Intent-based segmentation feeding it cleaner signals. And a deliberate pull-back from Buy Now to focus the whole account on the core auction funnel.
Sell-side leads: up 11%, 31% cheaper
The sell side told the same story. Lead volume up 11%, cost per lead down 31%. The intent split cleaned up the signal on both sides of the marketplace.
The one number that went down on purpose
Buy Now dropped 53%. That was the plan. The platform is an auction business. Bidding is the core motion. Buy Now was a distraction that was pulling budget and attention away from where the real GMV lived. Cutting it wasn't a loss, it was a reallocation. Bids grew 6x. Trade I'd make every day of the week.
Key takeaways
1. Segment by intent, not by geography
Splitting campaigns by state or division feels sensible until you realise you've fragmented every campaign into something too small for the algorithm to learn from. Ask what the user is trying to do, not where they live.
2. Fewer, bigger campaigns beat more, smaller ones
80+ campaigns with small daily budgets is 80+ campaigns starved of the data they need. 32 campaigns with enough spend each to train smart bidding will beat that every time.
3. Focus on the conversion type that matches the business model
An auction platform lives and dies on bid completions. Buy Now was a distraction. Cutting it wasn't a loss, it was a reallocation.
4. Trust the intent framework
Once the account was split by buying intent and selling intent, the campaigns essentially ran themselves. The right creative, the right keywords, the right landing pages all fell out of the intent framework naturally.
The bottom line
An account that looked complex and sophisticated was actually being held back by that complexity. Cutting 80+ campaigns down to 32 and re-segmenting by intent 6x'd the primary conversion metric on 30% less spend.
The structure was the strategy.
Ready to Achieve Similar Results?
Let's discuss how I can help transform your Google Ads performance