S.02 / DASHBOARDS
A company KPI dashboard brings together the few indicators management decides with. I design it custom, starting from the question rather than the chart: what you need to know every morning, which data answers it and what to do with the answer.
02 / PROBLEM
A dashboard you have to interpret for ten minutes isn’t a dashboard. It’s an exam.
Most dashboards are built backwards: the available data gets connected, every possible chart gets drawn and everyone hopes someone finds something. The result is a screen that gets opened in month one and forgotten by month three.
A useful dashboard starts from the questions asked in Monday’s meeting and ends in an action. Anything that doesn’t help answer them is surplus.
And underneath it needs a layer that guarantees the figures are true. That layer is Business Intelligence; the dashboard is the part you see.
03 / SIGNALS
If the dashboard doesn’t change what you do on Monday, half of it is surplus or the question is missing.
04 / FIT
When the questions repeat and the answer is still put together by hand.
05 / KPIS
Four areas, few indicators in each. What gets measured is decided before anything is drawn.
Revenue, margin by line or customer, outstanding payments and variance against budget, with the cut-off date visible.
Team capacity, hours per project, deadlines at risk and stalled work, as in AtalayaIQ’s management view.
Pipeline by stage, unanswered leads, where customers come from and conversion, with pending follow-ups up front.
An HR dashboard: headcount, absences, logged hours and turnover, with personal data protected and role-based access.
06 / QUESTIONS
Every view comes from a question. These are already answered in real systems.
A short, prioritised list (open projects with no activity, records to fix, pending expenses) with the evidence for each one.
EvidenceAtalayaIQ
Hours cost and expenses per project, with explicit coverage: which share of the hours has a rate and which doesn’t.
EvidenceAtalayaIQ
Hours logged per day, people with activity and each project’s weight over time, to see the load before it shows.
EvidenceAtalayaIQ
Pending follow-ups, leads by status and those with recent activity, before they go cold.
EvidenceZentia
A comparison with the previous period at the same logging cut-off, with the change that explains it. No false drops from half-logged data.
EvidenceAtalayaIQ
Leads by source, the most viewed items in the catalogue and an activity map to see what’s working.
EvidenceZentia
07 / ARCHITECTURE
The opposite of most: first what has to be decided; the charts come last.
What the decision-maker needs to know, and how often.
Which number answers that question, and which doesn’t.
What counts, what doesn’t and over what period.
Where the data comes from and how far it can be trusted.
The smallest chart that makes the answer obvious.
What to do with what you see, and where the evidence is.
08 / FROM EXCEL
The step that saves the most hours isn’t the chart: it’s no longer building the report by hand.
Today: someone exports from three systems every Monday, pastes into an Excel template and sends a PDF. If a figure is missing, nobody knows.
The sources are connected where they live, read-only, without migrating systems: the ERP, the CRM, the operations database.
Every KPI is defined in writing and reconciled with its source before it’s shown. If it doesn’t match, it isn’t published.
The dashboard updates itself, says when the data is from and highlights what needs attention.
The spreadsheet can still exist for thinking. What goes away is the work of building it by hand.
09 / EVIDENCE
A management view over seven years of operational data and the sales panel of a real-estate platform.
Business intelligence
A management dashboard that answers what needs attention today and shows why.
STACK Nuxt / Vue / MySQL
Case study: AtalayaIQ
Multi-tenant platform
A sales panel inside each agency’s platform, with its own data and only its own.
STACK Nuxt 4 / Supabase / PostgreSQL
Case study: Zentia10 / DON'T BUILD
A chart with no question behind it is decoration with a maintenance cost.
11 / QUESTIONS
A view that brings together the key indicators management or a team uses to follow how the business is doing and decide what to do. The useful ones have few indicators, well defined, connected to real data and ordered by what needs attention.
What is a management dashboard?The dashboard is the surface: the screen that answers specific questions. Business Intelligence is the layer underneath: source integration, definitions, reconciliation and rules. A dashboard without that layer shows figures confidently even when they’re wrong.
They’re good tools and often the right choice: a Power BI dashboard works well if the data is already in order and the questions are standard. A custom dashboard makes sense when the logic is your own (comparisons at the same logging cut-off, alerts with evidence, prioritised actions) or when it has to live inside the tool your team already uses.
A Power BI dashboard management actually usesThe ones that answer the questions decisions are made on. In AtalayaIQ, the front page sums up the situation “in three lines” and shows at most 15 actions, with no filler.
Whatever the decision needs. For management a daily cut-off is usually enough; for operations it may need to be close to real time. What matters is that the dashboard says when the data is from and which part is still incomplete.
Yes. The same rules that highlight something on screen can trigger an alert or a periodic report. In AtalayaIQ the daily and weekly report arrives ready, with the prioritised action plan.
Read the AtalayaIQ case study12 / SAME SYSTEM
A business problem rarely lives in a single piece. These usually appear in the same project.
Tell me what you decide, how often and where the data that answers it lives today. The dashboard is designed from there.