Reporting people trust
Markets and central teams work from the same definitions instead of debating whose numbers are right.
Global Analytics Manager · London
I make messy marketing data usable, governed and trusted.
I turn taxonomy, reporting and data-quality problems into operating systems teams can actually use. Strategy, governance, automation and leadership in one place.
Programme ownership
Multi-market governance
Team leadership
Hands-on build
governance_flow
RAW INPUT
UK_Meta_Awareness_summer
uk_fb_awareness_Summer26
GB_meta_AWR_summer
validate
CONTROL LAYER
Structurevalid
Approved valuescontrolled
Market relationshipcheck
standardise
TRUSTED OUTPUT
MarketUnited Kingdom
PlatformMeta
ObjectiveAwareness
Illustrative example only. No client data or production taxonomy is shown.
The bit that matters
What this is worth on your account.
You do not hire me for cleaner naming. You hire me because cleaner, governed data changes what your teams can trust, how quickly they can act and how much manual work they carry.
When governance works
Data stops being a recurring fire drill.
The standard, controls, ownership and reporting all reinforce each other.
Less manual QA
Deterministic checks and fixes move repetitive validation away from analysts and into the process.
Faster change
New platforms, markets and reporting needs can plug into a controlled model instead of creating another exception.
Clear ownership
Everyone can see what failed, why it failed, who owns the fix and what needs to happen next.
FirstFind what is actually true
Audit the live data and separate structural, semantic and process failures.
ThenMake it measurable
Turn data quality into something leadership and markets can see and act on.
FinallyMove control upstream
Prevent repeat issues with controlled values, validation and automation.
Selected work, anonymised
Three problems I know how to solve.
Each example is reconstructed with synthetic data. The logic and operating model are real; client information stays private.
01 · TAXONOMY GOVERNANCE
A naming convention should behave like a control system, not a PDF.
This example mirrors the logic of a real multi-market taxonomy. The controlled values and identifiers are anonymised so the method is visible without reproducing client data.
Campaign taxonomy string - live validation
Select a field to see the checks that run against it
Pass
Based on a live governance pattern - all values and identifiers shown here are anonymised
ProblemValid fields can still form an invalid string.
Structure, approved values and relationships all need to agree before the data should be trusted.
What I buildControls that explain the failure.
Every field returns a clear pass, fail or review reason so teams know exactly what needs changing.
Business resultCleaner inputs before they hit reporting.
Less manual interpretation and a standard that can scale across teams, markets and platforms.
FROM CHASING TO AN OPERATING MODEL
Live dataQA engineDetects the issue
PrioritiseIssue queueRoutes to the owner
ResolveMarket actionFix or explain
CloseRe-checkConfirms the fix
Validated
In progress
Blocked
02 · QA & REMEDIATION
The fix is not the spreadsheet. The fix is knowing what happens after a failure is found.
Problem
Manual checks find issues, but progress lives across email chains and individual memory.
What I build
Automated checks, owner-level action files, status logic, escalation and repeatable reporting.
Business result
Less chasing, clearer accountability and a remediation process that does not collapse when volume increases.
FROM RAW API FIELDS TO TRUSTED DIMENSIONS
RAW SOURCES
Campaign string
Advertiser ID
Platform metadata
semantic
layer
cross-checklayer
before report
REPORTING
Market
Business segment
Media type
!A valid-looking code can still be wrong for the account. Independent metadata catches the mismatch.
03 · REPORTING SEMANTIC LAYER
I do not blindly trust the string just because it parses.
Problem
Position-based extraction can turn a malformed or misplaced value into a confident reporting error.
What I build
Calculated dimensions that combine structure checks with independent account and platform metadata.
Business result
Reporting gets a controlled interpretation layer, making discrepancies visible instead of silently accepting them as fact.
Proof of build
Twelve tools. One operating system.
A twelve-tool suite across Python and the Microsoft Power Platform. Each component removes a manual failure point and connects governance from detection through to resolution.
GOVERNANCE SUITE
Detect. Fix. Operate. Control.
PythonPower AutomatePower AppsSharePointBI
01
ValidateKnow what is wrong
compliance_engineField-level validationindependent_qaRe-check reporting logicrefresh_guardBlock broken refreshes02
RepairRemove repetitive fixes
taxonomy_converterRecover safe legacy valuesremediation_exporterCreate owner-ready actionscode_generatorIssue controlled values03
OperateKeep work moving
compliance_trackerTrack movement over timemarket_mailerBuild targeted outreachresponse_readerTurn replies into status04
GovernStop the issue returning
dictionary_auditControl approved valuesrequest_ticketingRoute governed changessemantic_layerProtect reporting meaning
BUILD PRINCIPLE
Automation never invents a taxonomy value. If a fix cannot be resolved with certainty, it is returned for human review rather than pushed into reporting as a plausible guess.
Manager / Associate Director value
Senior enough to own the outcome. Technical enough to know what will work.
I sit between leadership, markets, analysts and technology. That means I can turn a vague data problem into a roadmap, build enough of the solution to challenge it properly, and create an operating model the team can run without me.
Salesforce MCI / DatoramaPower BIPythonSQLPower PlatformMarTech
Set direction
Translate data-quality issues into business priorities, roadmaps and clear decisions.
Lead people
Set expectations, coach delivery, review performance and make ownership obvious.
Build control
Design standards, QA, reporting and remediation that stop relying on individual memory.
Hold the room
Explain technical problems in plain English and keep stakeholders aligned on what happens next.
Track record
Progressively broader ownership.
Global Analytics Manager
Publicis Media / Spark FoundryLeading data governance, analytics, remediation, QA and marketing-technology operations across a global programme, with people-management responsibility across London and offshore delivery.
Senior Data Compliance Executive
AnnalectLed governance reviews, automation controls and analyst delivery, acting as an escalation point for taxonomy and reporting decisions.
Data Compliance Executive
PHD GlobalRan multi-market compliance, rebuilt reporting and trained local teams on taxonomy and data-quality standards.
Affiliates Executive
OMD UKBuilt commercial analysis and automated competitor monitoring, connecting channel data to allocation decisions.