What actually happens in month one.
We build the truth layer first: your economics, your data, your guardrails — established together. Early changes are reversible setup and budget corrections, made only as the evidence clears. By week three, the first daily modeled read of incrementality is live and guiding allocation.
The truth layer first
We connect your economics, systems, and history — P&Ls, daily revenue by channel, costs, inventory, unit economics, ad and data access — so decisions start from your numbers, not a template.
Guardrails, then latitude
We establish economic guardrails together. Early account changes are reversible setup and budget corrections made as checks clear; material changes always require mutual agreement.
A living model by week three
Week 3 turns two weeks of business context into the first daily modeled view of incrementality by channel and tactic — a decision guide, not a claim of perfect causality.
Five workstreams move together.
Our standard 30-day operating setup moves through four weeks of connected actions and outputs across five workstreams. None of them wait for the others to finish.
CMO measurement & decision framework
Turn business economics and multiple measurement lenses into clear guardrails and decisions.
Paid media
Reconcile audiences and signals, rebuild the account, and allocate against business evidence.
Creative strategy & analytics
Combine the creative taxonomy with performance evidence and brand context to prioritize the full-funnel portfolio.
CRO, web merchandising & landing pages
Read page-level behavior, prioritize opportunities, and launch the first focused tests.
Forecasting & demand planning
Understand your planning process and assumptions, then contribute questions, demand evidence, and operating experience. You own the plan.
The four-week launch.
Each week has specific actions and expected progress per workstream. The plan acts decisively without pretending the first answer is final — specific structures, allocations, and tests remain evidence-based working hypotheses inside binding guardrails.
Connect the truth and set the guardrails.
Everything starts with your actual economics and your actual data — before anything structural changes.
Set the economic guardrails
Enter your variable costs, confirm the LTV payback window, set the working efficiency floor and resulting spend ceiling, and weight the attribution models each channel will be judged by. Available capacity is not an entitlement to spend.
Reconcile audiences and signals
Connect ad accounts to Marathon Data, audit exclusions so existing customers aren't counted as new-customer prospecting, verify pixels, events, and CAPI, and document the baseline. First reversible setup or budget corrections happen only as the checks clear.
Define the brand, start the taxonomy
A structured brand deep dive becomes a creative strategy document about what to make and why — not fonts and colors — while every key asset gets labeled by hook, angle, persona, format, and emotional resonance.
Establish the site baseline
Review prior tests, key pages, and merchandising logic; compare conversion, revenue per session, and contribution value by traffic source; build the first living test and landing-page queue.
Gather the planning history and current process
Map how you plan today — owners, methodology, timing — and gather the forecast, P&L history, daily revenue by channel, costs, inventory, and weeks-of-supply constraints. Immediate planning questions are separated from what needs more context.
Turn data into a view — and start the learning loops.
The connected data becomes an integrated operating view, and the first controlled tests go live where the evidence supports them.
Build the integrated operating view
Review the connected data across marketing, P&L, site, creative, and forecast context; identify the real strategic opportunities; pressure-test the forecast with bottom-up and top-down signals.
Activate the execution model, launch the first test
Confirm decision rights and guardrails, assume day-to-day management (or direct the agreed owners in oversight mode), and launch the highest-value controlled media test the evidence supports — with QA before and immediately after launch.
Complete the framework, share the findings
Finish the taxonomy, identify what has actually worked across the funnel — not just high-spend or high-reported-ROAS assets — and align on the coming month's creative direction.
Begin the first website learning cycle
Order the test queue by expected business value, confidence, and effort; document the first test's hypothesis and measurement plan; align the paid-media landing path so campaign and site learning read together.
Explore the planning process, surface the critical questions
Walk the existing planning process with the people who own it, review actuals and assumptions without forcing premature recommendations, and identify the analyses that will make the next working session useful.
The first modeled read goes live.
Two weeks of business context become the first codified daily view of modeled incrementality — and the operating cadence begins.
Set media weights, begin the cadence
Complete the financial settings, set distinct direct-response and brand performance benchmarks, and apply a revisable triangulated attribution model to every active campaign for a daily modeled estimate of incrementality by channel and tactic.
Advance the account transition from evidence
Use the first two weeks of audience, pixel, and tracking findings to decide whether each structural change should be immediate, phased, or deferred — no forced cutover, no permanently fixed architecture.
Deliver the monthly portfolio strategy
Combine the taxonomy, early findings, and business context into the prioritized body of assets with the greatest expected impact — then hand the strategy to your production owner with the measurement approach attached.
Read Test 1, prepare the next test
Confirm the test is collecting clean data, allow the appropriate read window rather than forcing an early conclusion, and build the next-highest-value test from the living queue.
Address the most urgent planning questions
Follow your lead on pressing inventory or planning questions using the context gathered so far; when none are urgent, continue the advisory analysis comparing actual trend, media capacity, and plan assumptions.
Operate from the model — and keep priorities live.
The system is running: daily modeled reads guide allocation, learning loops continue, and priorities stay live as evidence changes.
Read the model, actualize the business
Review the first daily modeled reads without overreacting to short-term noise, compare the model's direction with the actual business, and set cross-functional priorities as a living operating queue — not a fixed roadmap.
Operate from modeled incrementality
The weekly reporting and action cadence is active, budget allocations are increasingly informed by daily modeled incrementality, and guardrail changes still require mutual agreement.
Complete the first creative analysis
The first analysis of recently running creative through the agreed taxonomy becomes a prioritized production plan that works with your existing creative resources — we provide strategy, analysis, and measurement; you own asset execution.
Continue the website learning loop
Complete the first full read the available test duration supports, decide whether to roll out, refine, or stop, and prioritize the next tests on conversion, revenue per session, and contribution value — not conversion rate alone.
Join the monthly reforecast with full context
Participate in your monthly reforecasting process with three weeks of gathered context, contribute practical process improvements, and translate your approved plan into the media guardrails we operate within. You retain the forecast and every planning decision.
What Day 30 means: core data connected and validated, goals and guardrails aligned, initial incrementality estimates guiding allocation, and the paid-media, creative, and website learning loops underway. You get a living priority system — not a fixed 90-day roadmap that goes obsolete as evidence changes.
Focused collaboration around the work.
Live working sessions stay focused on the people and work involved, while asynchronous updates handle routine context.
Business context and approvals
Business priorities, guardrails, plan context, approval needs, and where evidence suggests capital should move.
Actions and allocation
Pacing, budget changes, launch readiness, customer mix, and scale or pause actions with the hands-on operators.
Signal integrity
CTO-led work on pixels, events, CAPI, account access, data connectors, custom signals, and the measurement path.
Brand and asset context
Brand truth, customer, product, current and historical winners, asset rights, taxonomy, and the production process.
Test and implementation work
Hypothesis, implementation owner, QA, evidence, and whether to roll out, refine, or stop a test.
Evidence and open questions
Forecast assumptions, actuals, demand signals, open questions, and the media guardrails implied by your approved plan.
We agree on the minimum live collaboration needed to keep each workstream moving, and use asynchronous updates where practical — your calendar is not the deliverable.
Ready to see what your first month looks like?
A fit call covers your economics, your current setup, and whether this operating launch is the right shape for your brand.