C-01Paid media management
Performance marketing agency that plans on margin, not ROAS
The short answerquotable
Performance marketing at AiS LABZ is paid media managed on contribution margin rather than ROAS. We install server-side tracking before any budget moves, model the margin per order for each product and market, and set a margin floor every campaign has to clear. AI-assisted bid and budget pacing runs daily inside those guardrails, creative is tested one variable at a time, and periodic incrementality tests show which spend produces growth. A senior pod reviews the account weekly across search, social and retail media.
Why measure paid media on contribution margin rather than ROAS?
Margin is the number. ROAS is a vanity metric.
Model contribution margin per order for every product line and market before you set a target. Cost of goods, shipping, payment fees, returns and the closing discount all come out of revenue first, so two campaigns at identical ROAS can sit either side of break-even.
- Set a margin floor per campaign before launch
- Move budget to campaigns that clear the floor with room
- Cut products with thin margin and high returns from paid
- Review allocation weekly against the floor, not the dashboard
Why does tracking get fixed before any budget moves?
Fix tracking before you spend. Pixels miss too much.
Ad blockers, consent banners and iOS and Safari privacy changes strip browser-side conversions before they reach the platform, so bidding algorithms learn from a partial picture. Install server-side tracking first, pass order value and margin with every event, and reconcile against your order system before scaling anything.
- Server-side conversion APIs for Google, Meta and retail platforms
- Order value and margin attached to every event
- Reconciliation must tie out before spend increases
- Consent choices still respected; only blocked-script loss is recovered
The LABZ protocol· for performance marketing
Diagnose. Model. Build. Compound.
The same four phases run every engagement, in Waltham, London or Riyadh. Here is what each one means for this discipline.
Diagnose
Audit tracking, account structure and margin per product line. Find where reported ROAS and real contribution diverge.
Model
Install server-side tracking, reconcile it against the order system, and build the margin model that sets every campaign's floor.
Build
Restructure accounts, launch one-variable creative tests and switch on AI-assisted pacing with margin guardrails, one market at a time.
Compound
Review weekly, run periodic incrementality tests and move budget towards measured lift, so each month is planned on better evidence.
How does AI-assisted bid and budget pacing work?
Let automation pace. Let people set the constraints.
Auctions move hourly with competitor bids, inventory, weather and payday cycles, and a human checking twice a day reacts late. Run automated pacing against the live signal, tuned daily by the pod, inside guardrails: a margin floor, a daily budget cap and a minimum data threshold on every campaign.
- Bids and daily budgets adjusted against live auction signal
- Low-risk, high-confidence moves execute automatically
- Everything else queues for a person to approve
- Markets, margin floors and creative angles stay human decisions
How is creative tested without guessing?
Creative is the biggest lever left. Test it properly.
Targeting is largely automated now, so creative decides the outcome, and most accounts test it on a handful of impressions and record nothing. Isolate one variable per test, whether hook, offer, format or proof point, run to a stated confidence level, and document why the winner won.
- One variable per test, stated confidence threshold
- Angle library maintained across campaigns and markets
- Losing creative retired with the reason recorded
- Winning proof points shared with landing page and brand teams
What does an incrementality test change about the budget?
Attribution flatters branded search and retargeting. Measure lift instead.
Platforms take credit for customers who were already going to buy, which inflates the least useful spend and starves prospecting. Hold back spend in matched regions or audiences, measure the difference, and use the lift figure per channel instead of platform attribution when you plan budget.
- Cut or cap spend that produces little lift
- Move released budget to prospecting where lift is measurable
- Repeat tests; seasonality, competitors and creative fatigue move the figure
- Cadence set in the diagnostic by spend and market count
What you receive
Deliverables, not decks
Every item below is a contractual deliverable with a named owner and a date. Documentation is part of the work, not an extra.
Server-side tracking setup
Conversion APIs from your infrastructure to each platform, with margin attached and reconciled.
Contribution margin model
Margin per order by product line and market, floor set per campaign.
Account restructure
Campaign, audience and bidding structure rebuilt around margin, ready for guarded pacing.
Creative testing programme
One-variable tests with a documented angle library shared across channels and markets.
Incrementality test reports
Periodic holdout tests giving a lift figure per channel for budget planning.
Weekly performance review
Spend, margin and new customers by channel, changes, automation actions, tests in flight.
Which channels does a performance marketing agency manage?
Choose channels by where margin sits, not by habit.
Google and Microsoft search capture demand that already exists. Meta, TikTok and LinkedIn create demand and must earn their place through creative testing. Retail media on Amazon and grocery platforms converts well at the point of purchase but is rarely incremental for a brand people already search for.
- Google and Microsoft search, Shopping and Performance Max
- Meta, TikTok and LinkedIn paid social
- Amazon and retail media networks
- One account structure adapted across seven markets, margin floor per market
What does the weekly review cover?
Five minutes a week. Numbers your order system agrees with.
Expect a short weekly review from the senior pod: spend, contribution margin and new-customer volume by channel, what changed and why, what automation did alone, what was held for approval, and the tests running. Because tracking is server-side and reconciled, the revenue figure matches your order system.
- Weekly: spend, margin and new customers by channel
- Monthly: margin view by product, market and channel
- Incrementality data and next quarter's budget position
- Month to month after setup, no long contract
Further reading from the AiS LABZ blog
Is this for you
A good fit and a bad one
Built for
- Brands spending enough that margin, not ROAS, decides the outcome
- Founders who want ad platform numbers reconciled to their order system
- Teams scaling across markets who need one structure and one margin model
Not the right service when
- Businesses that cannot share cost of goods and order data
- Anyone who wants a dashboard ROAS target chased regardless of profit
Common questions
Questions we get asked
ROAS counts revenue, not what is left after you have paid for the goods, the shipping, the payment fees and the discount you gave to win the sale. Two campaigns at the same ROAS can differ completely in what they contribute. We model the margin per order and plan budget against that, so scaling a campaign makes the business more profitable rather than just busier.
Yes, with guardrails. Bid and budget pacing runs on automated systems tuned daily, because machines react to auction changes faster than a human can. What stays with our senior pod is the strategy: which markets to fund, what margin floor each campaign has to clear, and which creative angles to test. Automation handles the pacing; people decide the constraints it runs inside.
An incrementality test measures sales that would not have happened without the ad, usually by holding back spend in matched regions or audiences and comparing outcomes. It matters because attribution platforms take credit for customers who were going to buy anyway — most visibly on branded search and retargeting. Running the test periodically tells you which parts of the budget are producing growth and which are billing you for it.
A monthly management fee set in the diagnostic against scope: channels, markets and creative testing volume. A three-month setup covers tracking, the margin model and the account restructure, then the engagement runs month to month. We never take a percentage of spend, because that rewards spending more rather than earning more. Ad spend is paid directly to the platforms from accounts you own.
Diagnostic takes one to two weeks; tracking and the margin model another one to two. Data is cleaner within the first month, because the platforms are finally learning from complete conversion signal. Margin improvement typically shows in months two and three as budget moves to campaigns that clear the floor. Incrementality results follow the first holdout test, which needs a few weeks of stable spend.
Yes, and it usually works better that way. Your team owns brief, brand and creative production; our pod owns account structure, pacing and measurement, and both sit in the same weekly review. Everything we build, including the margin model and the test log, lives in your accounts and stays with your team.
Three things. Access to your ad accounts and analytics, so nothing is rebuilt in accounts we control. Cost of goods, shipping and payment fee data by product, because the margin model is only as accurate as its inputs. And a named decision-maker who can approve budget changes within a working day. If any is missing, the first phase is scoped around getting it in place.
Free 15-minute diagnostic
Find the constraint first.
Fifteen minutes with a senior on the Performance Marketing pod. We tell you what we would look at, what it would cost and whether it is worth doing — including when the answer is no.