A good KPI dashboard shows only the measures that drive a decision. Each has one agreed definition and one owner, and refreshes automatically from the systems that hold the data. Each role sees its own level of detail. If a number on it cannot change what someone does this week, it belongs in a report, not on the dashboard.
You will recognise this
If any of these came up in your last management meeting, the dashboard is not the real problem yet. The numbers underneath it are.
- We have a dashboard. Nobody opens it.
- Sales says one margin, finance says another, and the meeting is spent on which is right.
- I only see how the month went once the month is closed.
- There are charts everywhere and I still cannot tell how we are doing.
- Every new question means asking the same person for another export.
- When our finance lead is on leave, the report is late.
- We measure everything and act on almost none of it.
Why dashboards stop being opened
A dashboard full of the wrong numbers is worse than none, because it looks like progress. Dozens of charts, traffic lights everywhere, and nobody can say which figure should change what they do on Monday. It gets opened at launch and ignored soon after.
The second failure is quieter: the right numbers, but two versions of each. Sales counts revenue when a deal is signed, finance when it is invoiced, and both are right by their own definition. Put both on one screen and they simply disagree faster.
Start with the visualisation tool and skip the definitions, and the argument survives the new software.
The third failure is lateness. A dashboard that depends on somebody pasting exports into it is only as current as their last free afternoon. It stops when they do.
What is a KPI?
A KPI, or key performance indicator, is a measure you have agreed to steer the business by. Gross margin, cash conversion, on-time delivery and customer retention are common ones. What makes a number a KPI is the agreement around it: one definition, one owner, and a decision it is meant to change.
Every KPI is a metric, but most metrics are not KPIs. Your systems can count almost anything: calls logged, invoices raised, items scanned, visits to the website. A metric becomes a KPI when someone senior has said: if this moves, we act, and here is who acts.
Before anything goes on a screen, we ask four questions. What is the definition, and who owns it? What line should it stay above or below?
And what would we do differently if it moved? If nobody can answer that last one, the number is a metric worth keeping in a report, not a KPI.
How many KPIs should a business track?
Fewer than you track now, and few enough that every one has a name next to it. There is no correct number.
In our view, the limit is attention. A leadership team can act on a handful of measures in a meeting, and each department on a handful of its own. Beyond that, the extra charts are decoration.
Before anything is built, we agree three things with your team: which metrics matter, what decision each one drives, and what to stop measuring. The third list matters as much as the first, because it keeps the dashboard short enough to use.
Nothing on that third list is lost. It moves off the dashboard and into a report, where the person who needs it can still find it.
KPI examples by function
Here is a generic starting set for an established company in Malta, names only. Your own list comes out of the definitions work, and it is usually shorter. Each name needs a definition your team agrees before it means anything.
Leadership
Revenue against budget, gross margin, cash position, the forecast for the quarter, and the one or two operational measures that move the rest.
Finance
Actuals against budget, cash conversion, debtor days, creditor days, overdue receivables, cost per unit or per cover, and margin by product or by site.
Sales
Pipeline value, win rate, average order value, revenue by customer segment, customer retention, and share of revenue from repeat customers.
Marketing
Cost per lead, cost per sale, conversion rate by channel, and revenue by channel, counted on the same customer definition finance uses.
Operations
On-time delivery, order fulfilment time, stock turn, stock-outs, utilisation of people or equipment, and work waiting past its due date.
People
Headcount against plan, staff turnover, absence, overtime, and revenue per employee.
Dashboards by role, with tiered access
Who sees what
Access is tiered by role. Leadership gets the whole business on one screen, with forecasting. Each department gets its own numbers at the depth it works at.
Finance sees the month as it stands, operations sees where work is stuck, and sales sees the pipeline on the same customer definition finance uses.
Tiering is not only about confidentiality. A warehouse supervisor checking a screen at the start of a shift does not need the group cash position, and the managing director does not need every open order. Getting the depth right for each role helps keep a dashboard worth opening.
Flags are raised by rules, not by someone noticing. A KPI marked behind, a job sitting too long in one stage or a regular customer who has stopped ordering goes to the right person. That person decides what to do about it.
What a leadership screen holds
How current is the data on a dashboard?
Current to yesterday. Every system you run loads overnight into the warehouse. When leadership opens the dashboard in the morning, the numbers show the business as it stood at the end of yesterday, not of last month.
For steering the business, yesterday is usually current enough. The loads run overnight, outside the working day, and nobody exports a file or pastes anything.
Reports come from the same place. The monthly pack, if you still want one, builds itself from the same figures and lands in the right inboxes without anyone sending it.
Power BI or Looker Studio?
Both are capable tools, and both sit on the same warehouse. Power BI tends to suit companies already running Microsoft 365 and Excel. Looker Studio tends to suit companies on Google Workspace, and those whose marketing runs through Google Analytics and Google Ads.
We recommend one before the build, based on what you already run and who will use it. The tool is the smallest decision in the project. Because the definitions live in the warehouse and not in the dashboard file, the numbers are the same whichever tool displays them.
What changes, in one business
A distribution business in Malta supplies shops and restaurants. It runs an accounting system, a CRM and a stock system, plus the spreadsheets that fill the gaps between them. It already has a dashboard, with a chart for almost everything.
- The dashboard has so many charts that the managing director has stopped opening it.
- Sales reports margin before delivery costs and finance reports it after. Both sit on the same screen.
- The monthly pack is assembled by one person in finance from three exports and arrives after the month has closed.
- A question like which customers are ordering less than last season goes to that same person, and waits behind the pack.
- Late deliveries are noticed when a customer complains.
- The leadership screen shows a handful of measures, each with a named owner and a decision attached.
- Margin has one definition, agreed by sales and finance in the same room, and every screen inherits it.
- Sales, finance and operations each open their own view, at their own depth, from the same warehouse.
- The numbers are current to yesterday when the managing director opens them in the morning.
- Which customers are ordering less is answered from one place, without waiting for that one person.
- A late delivery raises a flag by rule, so operations can act before the customer calls.
The person in finance who used to build the pack now reads it, and spends the time on the analysis they were hired for.
KPI dashboard, automated report or the monthly pack?
They do different jobs. You will usually want both a dashboard and a report, reading from the same warehouse, and neither rebuilt by hand.
| KPI dashboard | Automated report | Monthly pack by hand | |
|---|---|---|---|
| Built for | Steering, day to day | A set audience, on a schedule | The monthly meeting |
| How current | To yesterday | To the morning it sends | Out of date on arrival |
| Who builds it | Built once, loads overnight | Built once, sends itself | Someone, every month |
| Definitions | Agreed once, in the warehouse | The same agreed ones | Re-agreed by hand |
| Depth | By role, tiered | By audience | One version for everyone |
| When the builder is away | Unaffected | Unaffected | Late, or missing |
What you have at the end of the first month
Implementation is the first month. Three to four weeks from an accepted quote, you have a live dashboard and the foundation under it.
An agreed KPI list
Which metrics matter, what decision each one drives, who owns it, and what you have stopped measuring.
Written definitions
Each KPI defined once in plain English and modelled in the warehouse, so every screen inherits it.
A warehouse in your account
The systems behind the first dashboard loading overnight, built in your environment, under your credentials. It keeps running if we part company.
A live dashboard
The first view in Power BI or Looker Studio, current to yesterday.
Tiered access
Who sees what, set by role, so each team sees its own numbers.
How it works
- A data strategy callThirty minutes with Glen, free. You leave knowing which single view would change the most about how you run, and whether the data for it already exists.
- The data auditAbout three hours with your team, mapping where each figure comes from and how it is made. Then a quote for the work.
- The buildThree to four weeks from an accepted quote to a live dashboard. Implementation is the first month. Your leadership team agrees the KPI list and the definitions with us; we build the rest.
- The fractional teamAn ongoing senior team keeps the dashboards live, keeps the definitions current and adds the next view worth having.
What we do not do
- We do not sell off-the-shelf dashboard templates. Every screen is built on your own data and your own definitions.
- We do not point a dashboard straight at your operational systems. It sits on the warehouse, so every screen reads one set of definitions.
- We do not replace the systems your teams work in. We are not IT.
- Custody of your data stays with you, because the warehouse is in your account.
- We do not automate decisions. The dashboard shows what moved; the people paid to judge decide what to do.
Glen Sultana founded 110 Analytics after a career built where the numbers move daily. Our people have run analytics for tier-one companies, S&P 500 firms and high-growth technology businesses. Glen is the person on the first call, and he runs every data audit himself.
- Built in your environment, under your account
- No lock-in: if we part company, it keeps running and stays yours
- Based in Malta
Straight answers
What is a KPI?
A KPI, or key performance indicator, is a measure a business has agreed to steer by, such as gross margin, cash conversion or on-time delivery. What makes it a KPI is the agreement: one definition, one owner, and a decision it is meant to change.
What are some examples of KPIs?
Common KPIs include revenue against budget, gross margin, cash conversion, debtor days, pipeline value, win rate, customer retention, cost per lead, on-time delivery and stock turn. Which ones you should track depends on the decisions your leadership team actually makes. Each needs an agreed definition before it goes on a screen.
What is the difference between a KPI and a metric?
A metric is anything your systems can count. A KPI is a metric someone senior has agreed to act on, with a definition, an owner and a decision attached. Every KPI is a metric; most metrics are not KPIs.
How many KPIs should be on a dashboard?
Few enough that every one has an owner and a decision next to it. There is no correct number, but a leadership team can act on a handful in a meeting, not dozens. Anything that does not drive a decision moves to a report.
Who builds KPI dashboards in Malta?
110 Analytics is a data consultancy in Malta that builds KPI dashboards for established companies, typically turning over roughly €1M to €10M a year. We agree the KPIs and their definitions with your team first. We then build the dashboards in Power BI or Looker Studio on a warehouse in your own account, and overnight loads keep them current to yesterday.
Can each team see only its own numbers?
Yes. Access is tiered: leadership sees the whole business, and each department sees its own numbers at its own depth.
How long does it take to get a dashboard live?
Three to four weeks from an accepted quote to a live dashboard; implementation is the first month. Before that comes a free thirty-minute data strategy call with Glen and a data audit of about three hours with your team.
What does a KPI dashboard cost?
We quote the work at the end of the data audit. The figure depends on how many systems feed the dashboard and what it needs to show first. Tool costs are separate: Power BI licences come from Microsoft or your Microsoft partner, and the standard version of Looker Studio is free.
Do we need a data warehouse to have a dashboard?
In the way we build, yes. Pointed straight at several operational systems, a dashboard tool reads each system's own version of the numbers. The warehouse gives it one set of definitions and one place to read from, and we build it in your own account.
Dashboard terms, in plain English
- KPI
- Key performance indicator. A measure you have agreed to steer by, with one definition, one owner and a decision attached.
- Metric
- Anything your systems can count. Useful in a report; not every metric belongs on a dashboard.
- Dashboard
- A screen showing the current state of the measures a role acts on, refreshed automatically.
- Tiered access
- Each role sees the numbers it needs at the depth it works at, and nothing it should not.
- Overnight load
- The scheduled job that copies each system's data into the warehouse every night, so nobody exports anything.
- Ontology
- The agreed list of your business's terms, each defined once, that every dashboard and report reads from.