Data doesn't create value.
Decisions do.
Dashboards, reports and data are tools of the trade. The real job is understanding what is happening, why it is happening and what the organisation should do about it.
I help organisations connect business goals, performance frameworks, reporting and analysis so the right people get the right evidence at the right time to make better decisions.
You've got the dashboards. Why aren't they creating value?
When reporting isn't useful, I don't start by redesigning the dashboard. I start by understanding why it isn't useful.
Do people not use it? Do they not trust it? Or do they understand the numbers but have no idea what to do next?
Those are very different problems. The answer might be better measurement, a clearer performance framework, different reporting, stronger analysis or simply turning off something nobody needs.
Start with what the organisation is trying to achieve.
Reporting should be the output of a performance framework, not a collection of everything your technology happens to be capable of measuring.
What outcome is the organisation actually trying to achieve?
What needs to change or improve to support that goal?
Which indicators tell us how that wider system is performing?
What supporting evidence helps explain why those indicators move?
Not everyone needs the same dashboard.
You can keep an overall view of organisational performance, but different teams need evidence relevant to the decisions they're responsible for.
Marketing doesn't need page uptime dominating its reporting. IT doesn't need ROAS and acquisition mix filling its operational view. Senior leaders shouldn't need to navigate either team's diagnostic detail to understand whether the organisation is on track.
The right reporting gives each audience enough information to understand performance, investigate what matters and act at the level they're responsible for.
One overall view
A shared picture of business performance and the outcomes the organisation is accountable for.
Specific views for specific decisions
Marketing, Product, Operations, IT, Finance and leadership see the measures that help them understand and manage their part of the system.
An analyst's job is half asking the right question and half providing the right answer.
Data, reports and dashboards are tools of the trade. They're not the trade itself.
Reporting tells you what happened. Insight should help you decide what to do next.
I use PEEP to turn an observation into an explanation the business can investigate, act on or test.
What meaningful thing have we observed?
What data supports the observation and shows that it matters?
Why do we believe this is happening, given the evidence and wider context?
If our explanation is right, what should happen if we change, investigate or test something?
When a number moves, the framework should tell you where to look.
If revenue drops, the first response shouldn't be to slice every dimension until something interesting appears. Start with the KPIs that describe the performance of that goal.
Goal: Increase revenue.
Start with the indicators that describe it, such as AOV, conversion rate and CPA. Identify which part of the system has changed, then investigate the supporting measures that can explain why.
Before building another report, justify why it needs to exist.
Reporting estates accumulate. Weekly spreadsheets, monthly packs and dashboards survive because they've always existed, not necessarily because anybody still makes a decision from them.
Where possible, I'll track usage or progressively turn reporting off to establish what is genuinely being used. The reports that remain can then be audited against their audience, objective, KPIs and the decisions or actions they support.
Analyst time is valuable. Removing unnecessary reporting can create capacity for analysis before you've hired another person or built another dashboard.
Report at the speed people can act.
The question isn't how quickly the technology can refresh the data. It's how quickly somebody could realistically make a different decision because of it.
Sometimes hourly really matters.
I've worked with reporting that showed users progressing through core journey steps every hour. When something moved unexpectedly, teams could spot it, diagnose it and be working on a fix before the next report arrived.
Sometimes quarterly is enough.
A strategic measure reviewed and acted on quarterly doesn't become more useful because the dashboard refreshes every 15 minutes. Freshness without a decision cadence is just more processing.
Self-service should answer the routine questions. Analysts should answer the difficult ones.
Reactive analytics has its place, but routine data retrieval shouldn't consume most of an analytics team's capacity.
Challenge the reporting
Start with the biggest chunk of effort. Are recurring reports still needed and what decisions do they support?
Enable self-service
If the data already exists, fix access, discoverability and literacy rather than repeatedly asking an analyst to retrieve it.
Standardise
Reduce duplication, clarify definitions and ownership, and make the reporting that remains tighter and easier to use.
Justify new demand
Before something new is built, establish who needs it, what question it answers and what decision it will support.
Move analysts up the value chain
Use automation, AI, better request processes and stronger analyst capability to spend less time producing numbers and more time validating, explaining and identifying what the organisation can do about them.
AI can summarise the numbers. Someone still needs to understand what they mean.
AI can make routine interrogation and reporting considerably faster, but it can also be wrong. It doesn't automatically understand the wider company, customer or economic context surrounding the data.
The analyst's role increasingly includes validating the output, bringing in context the data doesn't contain, challenging apparent explanations and developing predictions that can be investigated or tested.
Different teams can analyse performance differently. The organisation still needs one number it's accountable for.
If Marketing, Product and Finance all report a different version of the same KPI, having three technically defensible definitions doesn't remove the accountability problem.
Agree the number
I typically work closely with Finance to establish the agreed organisational number, definitions and reconciliation needed for performance accountability.
Keep analytical flexibility
Teams can still create alternative cuts or definitions for legitimate analytical purposes, provided they're explicit about what they mean and don't create competing versions of organisational performance.
Good analytics should help you anticipate what might happen next.
Forecasting can help organisations act before the quarter is over, but predictions should stay close enough to the evidence to remain useful and uncertainty needs to travel with them.
I prefer ranges, assumptions and explicit caveats over a precise-looking number that pretends nothing material could change between now and the end of the period.
State the assumptions, show the range and be clear about what the model cannot know yet.
From performance frameworks to the insight function around them.
Engagements can solve a specific reporting problem or help reshape how analytics supports decisions across the organisation.
Performance frameworks
Connect business goals, objectives, KPIs and supporting measures.
Reporting frameworks
Define audiences, information needs, cadence, ownership and decision use.
Dashboard strategy
Create focused self-service views around the questions different teams need to answer.
Reporting rationalisation
Audit, consolidate and remove reporting that no longer earns the effort required to produce it.
Performance analysis
Diagnose KPI movement and identify the evidence behind changing performance.
Insight development
Use PEEP to move from observations to explanations and testable predictions.
Forecasting & scenario analysis
Use current evidence to anticipate outcomes while making uncertainty and assumptions explicit.
Analytics operating model
Improve self-service, definitions, requests, data literacy, AI use, governance and analyst capability.
Measurement, reporting, insight and optimisation shouldn't live in separate silos.
They are different parts of the same decision-making system. When they move to the same beat, at the same cadence and towards the same destination, the evidence becomes considerably more useful.
Capture the evidence the organisation actually needs.
Report and analyse performance in the context of the business goal.
Turn evidence into explanations, predictions and choices.
Act, experiment and feed what you learn back into the system.
Five fingers can do useful things. Make a fist and they become considerably more effective.
Data validity, measurement, reporting, insight and optimisation are the same. Connect them and you remove hand-offs, reduce silos and create one route from business objective to evidence to action.
Analytics grounded in how organisations actually make decisions.
I've spent more than 19 years working across analytics, measurement, reporting, experimentation and optimisation, with experience spanning commercial businesses, large organisations and complex public-sector environments.
One evidence chain.
One point of ownership.
One view of the whole system.
You're not buying someone who only knows how to build the dashboard. I can work across data validity, measurement, performance frameworks, reporting, analysis, insight and optimisation, so the pieces don't disappear into separate silos.
That continuity matters. The person challenging whether the number is valid also understands why it is being reported, what the stakeholder needs to decide, how it should be analysed and what happens when the evidence suggests something should change.
The aim is a simpler system where everyone is working from the same evidence, at the right cadence, towards the same business outcomes.
Analytics, insight & reporting consultancy FAQs
What's the difference between reporting, analysis and insight?
Reporting shows what happened. Analysis investigates the patterns and drivers behind it. Insight connects a meaningful point to evidence, an explanation and a prediction that gives the organisation something it can investigate, change or test.
Can you help us redesign our dashboards?
Yes, but I normally start by understanding the audiences, objectives, KPIs, decisions and reporting cadence first. Redesigning a dashboard that shouldn't exist or that answers the wrong questions only makes the wrong reporting look better.
What is a performance framework?
A performance framework connects business goals and objectives to the KPIs and supporting measures used to understand performance. It gives reporting structure and provides a clearer diagnostic path when an important number changes.
Can you rationalise existing reports and dashboards?
Yes. I can assess usage, audience, objectives, duplication, KPI relevance, decision value and production effort, then help consolidate, redesign or retire reporting that no longer earns its place.
Can you help create a self-service analytics model?
Yes. That can include reporting structure, common definitions, dashboard design, access, stakeholder data literacy, request processes, AI use and analyst capability so routine questions don't require an analyst every time.
How do you use AI in analytics?
AI can help interrogate data, automate routine reporting and accelerate analysis, but its output still needs validation and context. I focus on using it where it reduces low-value effort while keeping human judgement around data quality, explanation, prediction and decision-making.
Do you work with existing BI and analytics tools?
Yes. The approach is deliberately tool-independent. The objective is to make the organisation's existing data and reporting more useful, whether the delivery layer is Power BI, Looker Studio, spreadsheets or another platform.
Got plenty of data but not enough answers?
Start with the decision or performance problem. We'll work backwards to the evidence, reporting and analysis you actually need.
