Ecommerce paid media, engineered not gambled.

Most brands think they have an ads problem. Usually they have a math problem, a creative problem, or a tracking problem wearing an ads costume.

Paid media is the most misunderstood channel in ecommerce. It looks simple: put money in, get sales out. So people boost posts, chase whatever's trending, and blame the platform when it doesn't work. Then they either quit, convinced ads "don't work for us," or they keep spending on faith, hoping this month is the month it clicks.

Neither is a strategy. Profitable paid media is a system with four moving parts: the economics that tell you what you can afford to pay, the tracking that tells you the truth, the creative that finds your buyer, and the discipline to scale only what's actually working. Get those right and ads become the most controllable growth lever you have. Get any one wrong and you're just lighting money on fire with a nicer dashboard.

The landscape shifted after iOS privacy changes. The brands winning now aren't the ones with the biggest budgets. They're the ones testing the most creative, fastest, against the cleanest numbers.

1. The foundation: tracking and economics

Before a single ad goes live, two things have to be right, and almost nobody does both. Skip them and everything downstream is guesswork.

Tracking, because bad data makes bad decisions

You cannot optimize what you can't measure, and after the privacy changes, measurement got harder, not easier. Your pixel and server-side tracking (CAPI on Meta, proper conversion tracking on Google) need to be firing and verified before you spend, with consistent UTMs so you can actually tell what drove a sale. This is the unglamorous plumbing that decides whether every decision after it is based on reality or on a number that's quietly lying to you. When an account is underperforming, broken or half-broken tracking is one of the first things I check, because it makes good ads look bad and bad ads look fine.

Economics, because you should know the answer before you spend

Here's the question most brands can't answer: what can you afford to pay for a customer and still make money? Until you know that, you're not running ads, you're donating. The math isn't complicated. Work out your breakeven ROAS from your margins, then your target ROAS above it, then your maximum cost per acquisition. Build a simple best case, average case, and worst case, where worst case is breakeven, not disaster. Now you have a number to steer by. An ad isn't "good" because the creative is pretty. It's good because it beats your max CPA. Everything else is opinion.

2. Creative is the new targeting

This is the single biggest shift in paid media, and most brands haven't adjusted to it.

It used to be that the platform found your buyer. You picked interests and lookalikes, and Meta's targeting did the work. The iOS privacy updates gutted that. The platforms can no longer see enough to reliably hand your ad to the exact right person. So the job moved. Now the creative has to do the targeting.

What that means in practice: the setting, the person on screen, and the hook are what qualify your audience. An ad shot in a busy-mom kitchen with a frazzled parent talking about mornings calls out busy parents before a word of copy lands. The creative says "this is for you" to the right viewer and "keep scrolling" to the wrong one. You're not relying on the algorithm to find your buyer anymore. You're building the buyer into the ad.

This is why UGC-style creative wins right now. Not because "authentic" is a buzzword, but because a real person in a real setting qualifies the audience and holds attention in a way a polished brand film can't. Authenticity is the performance driver: it lifts click-through, hook rate, and watch time, which is exactly what the algorithm rewards with cheaper distribution. The brands that treat creative as a checkbox lose to the brands that treat it as a testing system. More on that below.

3. How to run Meta ads

Meta (Facebook and Instagram) is where you create demand. People aren't searching for you, they're scrolling, and your ad has to interrupt them and make them want something they weren't looking for. That puts almost all the weight on creative.

The account structure should be lean. Fighting the algorithm with dozens of tiny, hyper-segmented ad sets is a 2019 move. Let consolidated campaigns (Advantage+ or a simple CBO setup) do the optimizing, and keep prospecting separate from retargeting so you can read them honestly. Then put your energy where it belongs: the creative.

Test angles, not headlines. A new headline on the same idea is noise. A new angle is a real test. The angles worth rotating: the pain, the desire, the proof, the mechanism (how it works), and the objection (the thing stopping them from buying). Every video needs to earn the first three seconds or it's dead, and every static needs to land at a glance. Run three to five creatives per ad set so the algorithm has something to choose between, and name everything with a convention like angle_format_hook so your reporting is actually legible a month later.

4. How to run Google ads

Google is the opposite of Meta. Here, people are already searching for what you sell. You're not creating demand, you're capturing it, at the exact moment someone types their intent into a box. That makes it some of the highest-intent traffic you can buy, and it makes the fundamentals different.

Structure campaigns by intent and goal, with tight ad groups by theme. Pick the campaign type by objective: Search for capturing direct intent, Shopping and Performance Max for product-led ecommerce, and so on. Conversion tracking here is non-negotiable, more than anywhere else, because Google's automated bidding is only as smart as the conversion data you feed it. Feed it garbage and it optimizes toward garbage.

Then the ongoing work is mostly discipline: mine your search terms report to add negative keywords (so you stop paying for searches that will never buy) and to find new themes, write several responsive search ads matched tightly to the landing page, and set a bidding strategy that matches your data volume. Manual or target-CPA when data is thin; smarter automated bidding once you have enough conversions for the machine to learn from.

5. Meta vs Google: which and when

This is the question I get most, and the honest answer is that they do different jobs, so for most ecommerce brands it isn't either/or.

Google captures demand that already exists. If people are actively searching for your product or category, Google puts you in front of them at the bottom of the funnel, ready to buy. High intent, often higher conversion rate, but capped by how many people are searching.

Meta creates demand that doesn't exist yet. Nobody wakes up searching for a product they've never heard of. Meta's job is to interrupt the scroll and make them want it. Bigger ceiling, lower intent per click, and it lives or dies on creative.

Where to start depends on your business. If there's real search volume for what you sell, Google is often the faster path to profit because the intent is already there. If you're creating a new category or your product is a visual, impulse, or story-driven buy, Meta is where the growth is. Most brands past a certain size end up running both, with Meta filling the top of the funnel and Google catching the intent it creates. The wrong move is picking one on vibes. Pick it on where your buyer actually is.

6. The creative testing system

If creative is the targeting, then your ability to produce and test creative fast is your real competitive advantage. This is where budget matters less than process. The winners aren't the biggest spenders. They're the ones testing the most creative, fastest, with the clearest briefs and the tightest feedback loops.

Three things make that system work.

The brief is the whole game. Whether you shoot in-house, hire creators, or use a marketplace, brief quality is the number-one determinant of what comes back. A strong brief specifies the hook, the body, the CTA, the environment, the type of person on camera, how the product is handled, the tone, and the length, so there's almost no room for interpretation. A vague brief gets you vague content. A specific brief gets you an asset you can actually test.

Multiply your winners. When something works, don't just run it, mine it. Take one winning ad and spin it into five new hooks, three new CTAs, and a few different shot styles. You're not guessing at new ideas, you're pressing on a proven one from new angles. That's how a single winner becomes a month of profitable tests.

Read the ad libraries. Your competitors are running a live experiment and showing you the results for free. In the Meta Ad Library and Google's Ad Transparency Center, the signal is longevity: an ad that's been running for months is converting, because nobody pays to keep a loser live. Study the ads that persist, the shot styles, the hook formats, the kind of person on camera, and let that inform your own tests. Not to copy, but to learn what's working in your market right now.

7. The metrics that matter

Paid media drowns you in numbers, most of which don't matter. Read at the account and campaign level, not the per-impression noise, and watch the ones that actually predict profit:

• ROAS and CPA against your targets. The scoreboard. Everything else is diagnostic.
• Hook rate (did they stop scrolling) and hold rate (did they keep watching). If the hook rate is low, the creative is losing them in the first seconds, before your offer ever gets a chance.
• Click-through rate, the test of whether the ad earned the click.
• Add-to-cart and conversion rate, which tell you whether the problem is the ad or the page it sends people to.
• Frequency and fatigue, your signal to refresh creative before performance decays.

The point of metrics isn't a prettier report. It's to tell you which single thing to fix next. A low hook rate points at creative. A great hook rate but low conversion points at the offer or the landing page. Read the funnel, isolate the weakest link, fix one thing, then let the next batch of data tell you if it worked.

8. Why your ads aren't profitable

When ads lose money, brands almost always blame the targeting. It's almost never the targeting. It's one of four things, and diagnosing which is most of the job.

The economics never worked. If your margins can't support the cost of acquiring a customer through ads, no amount of optimization fixes that. Sometimes the honest answer is that paid media doesn't make sense for you yet, or not until the back end (email, retention, repeat purchase) makes each customer worth more.

The creative isn't stopping anyone. Post-iOS, weak creative is the most common killer. If the hook rate is low, you never had a chance. This is usually the fix with the most upside.

The offer or the page is leaking. Great ads sending traffic to a page that doesn't convert is money poured into a bucket with a hole. Sometimes the ad account is fine and the website is the problem.

The tracking is lying. Broken attribution makes winners look like losers and gets them turned off. More than one "failing" account was actually fine underneath a measurement problem.

The method is always the same: diagnose traffic, then funnel, then sale. Find the single weakest link. Fix that one thing before you touch anything else. Changes get made on data, not on a hunch.

How I'd build this for you

Everything above is the what. Here's the how. I run paid media as an engineered system, not a gamble. I start with the economics and the tracking, so we know what a customer is worth and we can trust the numbers. I build the creative around the avatar and their real objections, because the creative is the targeting now, and I run it as a testing system, not a one-off. Then I read the account at the level that matters, isolate the weakest link, and scale only what's actually beating the number.

It works. For one ecommerce brand, disciplined paid media plus the retention systems behind it added over $100,000 a month in revenue. I keep client names private and the results real, and I'll show you the actual thinking on a call.

Paid media also doesn't work in a vacuum. The ads fill the top of the funnel, but the website has to convert the click and the email and SMS systems have to make each customer worth more, or the math never gets easier. That's the whole point of the free diagnosis: I'll look at where your ad spend is actually leaking, whether it's the economics, the creative, the page, or the tracking, and tell you the fastest fix, whether or not you ever hire me.