Performance Max with store goals and digital coupons, explained.

Someone has told you to just run PMAX. A rep, maybe, or the person who built your website. Nobody stopped to explain what it does to the rest of your account. So here it is in plain language, with the uncomfortable parts left in.

Performance Max is one campaign type that reaches across several Google surfaces at once. Search results, shopping listings, YouTube, maps, display inventory. You don't buy those surfaces one at a time. You hand over the assets and the product feed. Then you name a goal, and the system decides where the money goes.

That last sentence is the whole story. What you buy now is the objective, and the placements are the system's business. The objective you name is the only steering wheel you get.

Pick the goal and you've already picked the plan

Set it to online purchases and the system hunts people likely to check out on your site. Set it to store goals and it starts hunting people likely to walk in. Same campaign type. A different buy underneath.

A store goal pulls the whole machine toward geography. A radius around your locations, plus the local surfaces where somebody checks whether the sectional is actually in stock. Your Business Profile is suddenly inventory. Get your long-weekend hours wrong and you're paying to send people to a locked door.

The creative changes too. A store goal wants images that look like your floor, not stock photography of a loft in Copenhagen. Show the fitting room. Show the delivery van, and the front window in grey November light. People decide whether a place is worth the drive by looking at it.

Store visits, though, arrive as a modelled figure. The signal comes from people who turned location sharing on. No single trip shows up in your report. The platform waits until enough of them pass through before it publishes anything. That's the whole construction, said plainly. If your reporting hands you the number as a turnstile count, push back before you plan on it.

A coupon is how a walk-in leaves a trace

The digital coupon solves one problem. A modelled visit is good enough to plan against and too soft to prove anything. So the ad carries an offer that somebody claims on their phone. They walk in, show the screen at the counter, and your staff redeem it. Now the visit has a date and a basket attached.

That is the honest version. The coupon buys you a receipt, and the discount is the price of that receipt. Decide whether the trade is worth it before you set the value. Price it from margin, the way you price everything else.

A furniture store handing out a deep discount on a sectional is giving away real money. An offer on an accessory line costs far less. It still produces the trace. This desk would rather discount something small and countable than something large and beloved.

And the coupon leaks. Some share of redemptions will come from people who were already in the car. They found the offer, took it, and cost you margin on a sale you had won. Nothing in the report separates them from the customers you actually bought. Plan as though part of the redemption cost is subsidy.

The parts nobody volunteers

Start with the reporting. Performance Max reports thinly. You see results at the campaign and asset-group level, and much less about which surface did the work. The money moved. Where exactly it went is a question the report will not fully answer. Read it beside your blended numbers. No channel grades its own homework, and this one is especially keen to try.

Then there's your own name. People searching your store by name were already yours. A campaign told to find cheap conversions will find them first, because they convert easily. The report comes back beautiful, built on demand your sign out front created.

So exclude your brand terms where the account allows it. Keep a separate branded search line you can read on its own. Then watch that line the week Performance Max switches on. If branded clicks drop while PMAX conversions climb by about the same amount, you have your answer.

It also runs on your feed. A feed with stale prices or empty availability gives the system nothing to sell. A supplier rewrites one field overnight, and your best-selling chair is disapproved by the end of the week. Nothing announces it. Nobody notices until the weekly number sags.

Volume matters too. Starved of budget and conversions, the campaign never gets past guessing. Give one campaign enough to learn on before you open a second.

Five questions will tell you who built it

What goal is this optimizing for, and who chose it? A shrug means the goal is whatever the default was.

Are my brand terms excluded, and what did branded search do the week this launched? Without that chart, nobody was watching the two lines together.

How many items in my feed are disapproved right now? That number should arrive the same day you ask for it.

What is my cost per store visit, week by week? A monthly average hides the two weeks that decided the season.

What happens if this campaign goes dark for two weeks? Save that one for last, because it stings. If nobody can predict what falls and what holds, then nobody knows what it's doing. That test costs two weeks and settles an argument that could otherwise run all year.

Ask all five this week. Whoever manages the account should answer them on the spot. How they answer will tell you more than the dashboard does.