Performance Consulting
Paid acquisition measured on what it returns, not on the numbers the platforms grade themselves with.
Overview
How I approach this work
Performance problems arrive as a number that moved — return on ad spend is down, acquisition cost is up, a channel that worked has stopped working. Underneath, the same three things are usually true: the account is optimising toward a conversion event that does not represent real value, several channels are claiming credit for the same sale, and nobody can say what a customer is worth or how long it takes to earn that back. Until those are settled, more budget only buys a faster version of the result you already have. So the work starts with measurement and unit economics and moves to the account afterwards, because bidding, structure, and creative can optimise toward nothing except what you have told the platform to value.
Engagements usually begin with an account somebody else has been running, so the first job is understanding what the existing structure is actually optimising for before changing anything — the learning history is an asset, and it is easy to reset it by accident. After that the work follows the same order every time. Measurement first: what is being counted as a conversion, whether it fires accurately and once, whether it passes a real value rather than a flat number, and whether the events the account bids toward correspond to revenue or merely to activity. Where the outcome that matters happens later or offline, that usually means importing qualified pipeline or completed orders back into the platform so bidding learns from results rather than from whatever fired first. Then the economics: acquisition cost by channel and blended, contribution margin per order rather than headline return on ad spend, payback period, and lifetime value where the data honestly supports estimating it. Then attribution, which is where most reporting quietly overstates itself — platform-reported conversions overlap, last click flatters brand search and retargeting, and the only dependable answer to whether a channel is incremental comes from holdouts and geo tests rather than a dashboard. Only then does the account itself repay attention: campaign structure and consolidation, what automated bidding needs in order to work at all, budget pacing, search term and negative hygiene, audience signals, and feed quality where products are involved. Creative gets treated as a media variable rather than a design task, because under broad targeting and automated bidding it is the main lever left — which makes angle variety, format coverage, and testing cadence a system instead of a series of one-off ideas. The landing page and the offer sit at the end of that chain and routinely cap everything upstream of them. What I hand over is the set of decision rules: what gets reviewed weekly, what is left alone long enough to learn, and what genuinely justifies more budget.
Deliverables
What a typical engagement produces
Concrete artifacts from this kind of work.
- Conversion tracking audit
- Every conversion event traced end to end — what fires, whether it fires once, whether it carries a real value, and whether the events the account optimises toward correspond to revenue. Covers consent handling and the server-side path where browser collection has degraded.
- Unit economics model
- Acquisition cost by channel and blended, contribution margin per order, payback period, and lifetime value where the data supports it — the numbers that decide whether spend is profitable, which platform-reported return on ad spend cannot tell you on its own.
- Attribution and incrementality plan
- An honest account of which channels create demand and which take credit for it, with a test design — holdouts, geo splits, staged budget changes — that answers the question instead of a dashboard that restates the platforms.
- Account structure and bidding review
- Campaign structure and consolidation, bidding strategy and the signal it needs to work, budget pacing, search term and negative hygiene, audience signals, and product feed quality where relevant.
- Creative testing system
- Creative handled as the media variable it now is: an angle and format matrix, a testing cadence with enough volume to read a result, and a documented way of retiring fatigued assets and iterating on the ones carrying the account.
- Landing page and offer review
- Message match from ad to page, the friction between arrival and conversion, and whether the offer itself is the binding constraint — because the best-run account still converts against whatever the page will bear.
- Channel and budget allocation
- Which channels suit the offer and the buying cycle, what share of budget belongs to capturing existing demand against creating it, and how much is genuinely held back for testing rather than absorbed into scale.
- Reporting the business can read
- A reporting view built on the numbers that drive decisions rather than the ones each platform surfaces by default, with the gap between platform-reported and actual results stated plainly instead of quietly reconciled.
- Governance and decision rules
- What gets reviewed weekly, what is left alone long enough to exit learning, what triggers a budget increase, and what counts as a failed test — so the account runs to a process rather than to whichever number moved most recently.
- Where I start
- Tracking and unit economics
- Judged on
- Contribution, not platform-reported ROAS
- Scale decisions
- Incrementality, not last click
Related areas
Other parts of my practice that overlap with this one.
Email Consulting
Lifecycle, deliverability, and the customer data underneath — email as a revenue channel, not a send calendar.
Content Consulting
What your content says, whether it persuades, and whether it earns the index.