Google Performance Max for Retail: The Local Growth Engine
Google Performance Max for retail is the difference between spending budget and investing it. For many retailers, it’s the replacement for what “Local Campaigns” used to do. It’s now one of the most direct ways to convert intent to footfall.
It’s normal to hear people refer to Performance Max (PMax) as a “black box”. But the control you lose is manual control. The control you gain is system control. The control of shaping inputs (goals, assets, exclusions, etc) to influence the outcomes you’re after.
One thing you can’t control is the algorithm itself, but you do control the data it receives.
Fuel first
If you’re starting from scratch, there are preconditions to making this work for you. So before you set it up, make sure your Google Merchant Centre and product feeds are in good shape.
Read our comprehensive guide: Google Merchant Centre & Product Feeds.
What’s “Store Goals” in Performance Max?
Performance Max is a goal-based campaign type that gives you access to Google’s inventory from one campaign.
- YouTube: video placements before or during content
- Display: image placements across sites and apps
- Search: text placements for high-intent queries
- Discover: visual placements in the Google feed
- Gmail: placements in the Promotions tab
- Maps: promoted pins and local results
You may not need to bid directly on each of those, but they are available.
Retailers have one important choice to make: the store objective, “Local Store Visits and Promotions”. Setting this goal tells the system that your priority is a physical visit, not an online click.
Driving footfall with an algorithm
PMax uses machine learning to try to find people who are likely to visit. It learns from Store Visit Tracking signals and the other intent, behaviour, and location signals that you layer on top of it.
Practically speaking, it starts to recognise patterns. For example:
- people who are within a realistic travel distance
- people searching “near me” terms at a given time of day
- people consuming content that aligns with your category
If these conditions are met over time, the system starts to build up a profile of your “ideal visitor” and bids more aggressively for those similar users in future.
To feed the algorithm the right signals, you need tracking to be switched on and working.
Learn more in our guide to Footfall Tracking & Measurement.
Campaign structure for retail: Local vs. National
The biggest mistake retailers make with PMax is trying to use a single campaign for both online sales and store-visit goals. The system optimises for one outcome at a time, so if you try to make one campaign do two jobs, you end up with something that optimises to the fastest, easiest signal: e-commerce.
Split your approach into two:
The National Campaign (E-commerce) Targets all of the UK (or shipping footprint). Goal is “Sales”. Uses your primary product feed to compete for delivery orders.
The Local Campaign (O2O) Targets the local areas around your physical locations. Goal is “Local Store Visits”. Relies on your Local Product Inventory Feed (LPIS).
This may seem like duplication. In some cases it is, but what it prevents is the system starving store outcomes while it chases online conversions. It also makes reporting clearer: you can quickly tell which budget is driving which result.
Asset groups: Building the creative engine
An Asset Group is not a folder. It’s the building blocks that PMax uses to dynamically assemble ads at scale. It needs choice. If you drop everything into one big group, you make relevance harder. You limit learning. Prevent the system from seeing what works best for a given search.
Consider structuring asset groups around themes that people are actually searching for:
- Asset Group A: “Running Shoes” (shoe images, simple category video, relevant headlines)
- Asset Group B: “Gym Wear” (lifestyle images, short brand video, category copy)
Relevance is important because PMax is constantly matching your searcher’s intent against available assets. When someone searches “running shoes”, you want the system to reach for the right images and right text, not just what happens to be available in the “catch-all” group.
Local Inventory Ads and the PMax connection
Local Inventory Ads (LIAs) were not built for PMax. PMax was built for LIAs.
PMax is the vehicle that delivers your Local Inventory Ads. If you link Merchant Centre, it pulls your local feed and automatically builds local product ads.
These often appear with:
- product image
- price
- distance to store
- “In Stock” label
You do not need to build separate LIA campaigns alongside PMax in the way you used to. PMax is handling the distribution for you across Search, Maps, and Display.
The catch is that LIAs only work as well as your data. If local feed is messy or out-of-date, you’re wasting clicks and frustrating shoppers. Ensure your LIA data is accurate before you start.
Learn more in our Local Inventory Ads article.
Advanced optimisation: scaling your results
PMax is not “set and forget”. It is “set, guide, and iterate”. The highest-performing accounts are the ones where you give the system a clear starting point and then tighten what it can learn over time.
Audience Signals Audience Signals help to nudge the system into the right starting place. You can upload your customer list (where consent allows) or add remarketing signals, and then tell Google “Start by looking for people like these”. It often reduces the learning phase and limits early waste.
Location Exclusions Even with store goals, PMax will still occasionally waste spend on areas that never convert to visits. The location logic you use for targeting is now a defensive mechanism. Use it to exclude those areas that never result in store visits. Your work on Postcode Targeting now doubles up as a defence layer.
Find out more with our Postcode Targeting vs. Radius Targeting guide.
Seasonality Adjustments Retail moves in peaks. PMax is reactive to trends, but if something shifts suddenly (promotions, major trading weekends, clearance), it can be slow to react. Seasonality Adjustments allow you to tell the system what to expect, so it can respond more quickly.
Ready to build the engine?
Book an Online to Offline Growth Session and we can build and optimise your Performance Max campaigns for maximum footfall.
Reporting: from black box to growth system
The black box is getting lighter. Reporting is still not as it should be but there is more you can use now to inform decisions.
Asset Reports Asset reports show which headlines, images, and videos are marked “Low”. Replace with clearer, more category-relevant options, then schedule a weekly review and refinement session.
Location Reports Location reports show you where visits are actually coming from. Use this to refine your exclusions and store catchments.
Product Reports Product reports show you which items trigger the most visibility, and which products are holding back performance. Remove any weak items and keep focus tight.
Weekly review and refinement Make this a weekly habit. Your job is not to fight the system. Your job is to improve the inputs to the system so that it has better options to work with.
Case study: Local Performance Max in action
A UK homeware retailer we work with previously ran standard Shopping campaigns. We implemented Google Performance Max for retail with a “Store Visit” goal, segmented by store location, with better, stronger assets showing the showroom experience.
The result was a 215% increase in store visits at a 40% lower cost per visit. The algorithm found local customers that manual bidding missed.
Explore the full breakdown of this success with our O2O Case Study.
Conclusion
Performance Max is not something to fear. It’s a tool to master.
Automating the complexity of bidding across channels and placements frees you to focus on what actually moves performance: your structure, your assets, your inventory data and your measurement.
Ready to build the engine?
We can build and optimise your Performance Max campaigns for maximum footfall.
Book an Online to Offline Growth Session
FAQs
Is Performance Max replacing Local Campaigns for retailers?
In most cases, yes. PMax now covers the multi-surface distribution that Local Campaigns provided, with better goal control and more flexibility.
Do I need Local Inventory Ads to drive footfall with PMax?
Not always. Store-goal PMax can still drive visits, but LIAs can reduce friction by showing local availability when your data is reliable.
How long does PMax take to learn for store goals?
Learning depends on data volume and measurement. Expect a ramp-up period, then judge performance once visits have had time to happen and be reported.
What’s the biggest reason store-goal PMax underperforms?
Mixed goals and messy inputs. When you combine e-commerce sales and store visits, or when your assets and inventory data are weak, the system struggles to optimise to the outcome you care about.