Smart Shopping campaigns were a Google Ads campaign type that used machine learning to automatically place product ads across Google’s network, including Search, Shopping, YouTube, Gmail, and the Display Network. Google retired Smart Shopping campaigns in mid-2022, automatically upgrading them to Performance Max campaigns between July and September of that year. You can no longer create new Smart Shopping campaigns, but the core automation and logic behind them lives on in Performance Max.
What Smart Shopping Campaigns Did
A standard Shopping campaign required advertisers to manually set bids, choose placements, and decide how much to spend on each product or product group. Smart Shopping removed most of that manual work. You provided a product feed from Google Merchant Center, set a daily budget, and chose a goal like maximizing conversion value or hitting a target return on ad spend (ROAS). Google’s algorithms handled everything else: which products to show, where to show them, how much to bid, and which audiences to target.
The system pulled product images, titles, prices, and descriptions directly from your Merchant Center feed, then assembled ads in different formats depending on the placement. A Shopping ad on the Search results page looked like a product listing with an image and price. On YouTube or the Display Network, the system generated responsive display ads using the same product data plus any additional creative assets you uploaded, like logos or marketing images.
How the Machine Learning Worked
Smart Shopping campaigns relied on automated bidding that analyzed multiple signals in real time to decide how much to bid on each auction. These signals included the user’s device, location, time of day, search query, browsing history, and the likelihood they would convert. The system weighed all of these factors simultaneously, something no human advertiser could do manually at scale.
When you chose “maximize conversion value” as your bidding strategy, the algorithm spent your full budget while prioritizing the clicks most likely to result in high-value purchases. If you set a target ROAS (say, 400%), the system aimed to generate $4 in revenue for every $1 spent on ads. It would pull back on auctions where the expected return was too low and bid more aggressively where signals pointed to a high-value conversion.
New campaigns went through a learning period, typically lasting one to two weeks, during which the algorithm gathered conversion data and calibrated its bidding. Performance was often inconsistent during this phase. The more conversion data the campaign accumulated, the more accurately it could predict which impressions were worth bidding on.
What You Needed to Run One
Smart Shopping campaigns required a few pieces of infrastructure before you could launch:
- Google Merchant Center account: This is where your product feed lives. The feed includes product titles, descriptions, images, prices, availability, and other attributes. Google required product data to be updated at least every 30 days, though most successful advertisers updated daily.
- Google Ads account linked to Merchant Center: The two accounts had to be connected so the campaign could pull product data into ads automatically.
- Conversion tracking: The algorithm needed conversion data to optimize bids. At minimum, you needed purchase tracking installed on your website. Revenue values attached to each conversion gave the system better signals for maximizing conversion value.
- Shopping ads policy compliance: Google enforced a separate set of policies for Shopping ads beyond standard Google Ads rules. Products, promotions, and your website all had to meet these requirements, or individual products (or your entire feed) could be disapproved.
Why Google Replaced Them
Performance Max builds on the same foundation as Smart Shopping but expands it. Smart Shopping ran across Shopping, Search, YouTube, Gmail, and Display. Performance Max adds inventory on Discover, Maps, and additional placements that Smart Shopping couldn’t access. It also gives advertisers more insight into how automation is performing, with reporting on audience signals, asset performance, and search term categories that Smart Shopping largely kept hidden.
The transition was not optional. Google automatically upgraded all remaining Smart Shopping campaigns by September 2022, carrying over budgets, product feeds, and historical performance data. Advertisers who had been running Smart Shopping campaigns didn’t need to rebuild from scratch, though many found that Performance Max required adjustments to creative assets and audience signals to perform at the same level.
How Performance Max Carries the Torch
If you’re setting up a Shopping-focused campaign today, Performance Max is your path. You still need a Merchant Center feed, linked accounts, and conversion tracking. The bidding strategies work the same way: maximize conversion value, or set a target ROAS and let the system optimize toward it. The algorithm still analyzes real-time signals like device, location, audience behavior, and competition intensity to set bids auction by auction.
The biggest practical difference is that Performance Max asks for more creative inputs. Beyond your product feed, you can supply headlines, descriptions, images, and videos that the system mixes and matches across placements. You can also provide “audience signals,” which are suggestions (not hard targets) telling Google which types of customers to prioritize. The algorithm uses these as starting points, then expands to other audiences as it finds conversions.
Budget management works similarly to Smart Shopping. You set a daily budget, and the system allocates spend across channels automatically. You have no ability to control how much goes to Shopping versus YouTube versus Display. This was a common frustration with Smart Shopping, and it remains true with Performance Max. If you want granular control over individual channels, standard Shopping campaigns still exist alongside Performance Max and let you set manual bids and choose placements yourself.
When Automated Shopping Campaigns Make Sense
The automated approach behind Smart Shopping (now Performance Max) works best when you have enough conversion volume for the algorithm to learn effectively. Google generally recommends at least 15 to 30 conversions per month for automated bidding to stabilize. Accounts with fewer conversions often see erratic performance because the system doesn’t have enough data to predict outcomes accurately.
Larger product catalogs also benefit more from automation. If you sell thousands of products, manually setting bids for each one is impractical. The algorithm can evaluate every product in every auction and shift budget toward whatever is converting best at any given moment. For smaller catalogs with just a handful of products, a standard Shopping campaign with manual bidding can give you more control without much additional effort.