Performance Marketing

How I build an account

1. I start with the objective and ensure the account structure reflects the commercial goal, whether that’s leads, applications, sales, revenue, qualified demand or something else. The platform is the execution layer that doesn’t determine my strategy.

2. Then I look for meaningful differences across factors such as audience, geography, product, customer value, intent and stage of the funnel. I question whether the difference is large enough to change how we should allocate budget, target people, measure performance or optimise.

3. To determine where budget moves, I segment based on where our strongest opportunities are: audience behaviour, products that justify more investment, and what is actually driving performance. I avoid fragmentation that spreads conversion data too thin as this creates unnecessary complexity and makes optimisation harder. So I segment where the data and business case justify it, and consolidate where they don’t.

4. I don’t believe every platform should look the same, so I match structure to the platform. Search, social and other acquisition channels have different mechanics, levels of intent and optimisation requirements. I build each around what the platform is actually good at while keeping the overall acquisition strategy connected.

5. Because not every customer, product or conversion is worth the same amount, I let economics influence the structure. I make sure that budget and structural complexity reflect customer value, conversion volume, acquisition cost and business priorities. The entire goal is to create the conditions for investment to follow opportunity.

6. I build for learning. I organise and design accounts in a way that teaches something. Campaigns I create make it possible to compare performance, identify patterns and make informed budget decisions without the structure itself making the data difficult to interpret.

7. Finally, I want the account to make sense beyond the person who built it and be scalable. So I use naming conventions, campaign architecture, budget models, reporting, and documentation that make the system understandable and adaptable as the business grows, because a good account structure makes better decisions easier.

How I run a campaign

1. I start with the number, not the platform. What are we buying? Is it applications, leads, sales, attention? And what is one worth? Platform choice follows from that.

2. I mine what is already working. Organic content gives you information before you’ve spent anything. I look at what people already respond to, the topics, messages, formats and creative angles that earn attention naturally. Some of the strongest paid creative comes from ideas that have already proven themselves organically.

3. I build so the campaign can be read. One clear objective per campaign. Naming conventions that still make sense months later. Tracking in place before launch. If I can’t tell which decision produced which result, the spend taught me very little.

4. I test deliberately. I don’t change everything at once. Creative, audience, placement, offer or landing experience may all affect performance, but changing several variables simultaneously makes it difficult to understand what actually caused the result. So I learn from each change I make.

5. I intervene when the data gives me a reason. I check in early enough to catch a bad start, but I don’t confuse early noise with a trend. When the evidence is there, I cut what isn’t working, move budget toward stronger opportunities and stay willing to abandon creative I personally like. The audience gets the final vote.

6. I report in plain language, and my reports answer three things quickly: what happened, why it matters, and what should we do next. I care about direction, magnitude and action, more than filling a dashboard with numbers.

7. Every decision I make traces back to the objective. I don’t optimise for numbers simply because they look good. Spend, efficiency and conversion trends only matter in the context of what the campaign is supposed to achieve. I turn those trends into clear visuals and recommendations so the case for scaling, cutting or investigating is easy to understand.

What I actually measure

1. I measure the outcome, not just the event. A lead, application, purchase or signup is a useful signal, but it isn’t necessarily the end of the journey. I want to understand what happens after the platform records the conversion. The closer I can get to the actual business outcome, the more useful the data becomes for deciding where the next dollar should go.

2. I separate signals from outcomes. I look at performance at multiple levels, from media efficiency to the quality and value of what the media ultimately produces. Platform metrics are useful for optimisation, but they are still proxies for business performance. A campaign can have a strong conversion rate and still produce weak commercial results. Conversely, a campaign can look expensive at the platform level while producing customers who are more valuable in the long run.

3. I want performance to be traceable. The useful question to me isn’t simply, “How did the campaign perform?” It’s: where did the result come from, what did it cost, and what happened afterwards? I want to connect spend and conversion data as closely as possible to the outcomes the business actually cares about, while being able to break performance down by dimensions that inform decisions.

4. I measure what can change the decision. I don’t believe in tracking everything just because a dashboard can. I care about metrics that help answer a question: is this working? Is it becoming more or less efficient? Where is the opportunity? What should we scale? What should we cut? What needs investigation? If a metric doesn’t help answer one of those questions, it probably doesn’t belong at the centre of the report.

5. When the numbers are bad, I diagnose before I optimise. A bad number doesn’t automatically mean the media account is the problem. I separate the variables I control — creative, targeting, structure, bidding and budget — from variables that may sit elsewhere in the customer journey, such as the offer, pricing, landing experience or follow-up process. When performance moves in the wrong direction, I first work out where the problem is occurring before changing the account. Otherwise, it’s very easy to optimise the wrong thing.

6. My final metric is the decision. The purpose of measurement is to make better decisions. I want reporting to make it obvious what changed, why it matters, what I believe is causing it, and what I would do next. Because ultimately, the job of performance marketing isn’t to make the platform report look good. It’s to make the business perform better.