Meetings open with a row about the total
Definitions get written down. Sales counts by order date, accounts by invoice date, and credit notes are deducted differently by each.
Monday, 9 a.m. at a small firm in Murcia: the office manager pulls sales out of Sage, opens the Amazon settlement, manually subtracts fees and carriage, then sends the directors a workbook. Meanwhile the sales lead has built his own sheet, with a different total. That routine swallows a few days every month and leaves two competing truths. We aim to leave you with one: tidy records, sources gathered in one place and dashboards that update without anybody touching them.
Each can be bought on its own, yet sequence matters. Build a warehouse before cleaning the master records and you simply get the old errors in nicer charts. Document reading runs alongside, dealing with the paper still arriving at the loading bay.
Every customer, supplier and product held once, with a valid NIF. We find duplicates, merge them with your approval and set entry rules the ERP itself enforces. With Verifactu approaching, this underpins everything else.
a3ERP, Holded or Odoo, your PrestaShop store, marketplace settlements and Google Ads spend flow nightly into one EU-hosted database. On top sit Power BI or Metabase dashboards that agree with the accounts.
Delivery notes, foreign invoices, consignment notes and contracts that someone currently types in line by line. The system reads them, maps supplier codes to yours and queues each one for a person to approve.
Hardly anyone asks us to “clean master data”. They ask for a dashboard, and a look at the sources turns up one of these symptoms. Recognise two or more and the reporting tool is not where the trouble lies.
Definitions get written down. Sales counts by order date, accounts by invoice date, and credit notes are deducted differently by each.
We clean up and set rules. “Talleres Ruiz”, “TALLERES RUIZ SL” and a record lacking an NIF count as three separate firms in any report.
We automate the loads. Exporting, pasting and reconciling by hand ties up the person who knows the numbers best, right when they are needed most.
We tune refresh frequency. When a product sells out faster than the report arrives, reordering becomes guesswork.
We set up automatic reading. Lines from Portuguese or Chinese suppliers are still copied into the ERP one at a time.
We work out real channel margin. Amazon, Miravia and your own shop get compared on turnover, not on what is left after fees and returns.
A polished chart on dirty data is worse than an ugly spreadsheet. Whoever assembles the sheet by hand at least looks at the rows and sometimes spots nonsense. A slick dashboard radiates confidence, so nobody challenges it even when duplicates inflate active customers by a fifth. That is why we measure data quality before drawing anything, and we show you the error count as it is, without softening it.
We begin with a single area, usually sales or stock, and a small set of measures. Other areas follow once the first report is in daily use. All of it is remote, through secure access to your systems and meetings on Teams or Google Meet.
With accounts and sales we set down how revenue, margin, returns and an “active customer” are calculated. Skip this and the first report is disputed at its first outing.
ERP, shop, marketplaces, advertising and the unofficial spreadsheets. For each we note an owner, how it is accessed and which personal fields belong in the record of processing activities.
The first automated report runs next to the old spreadsheet for a few weeks. Every gap is explained before management retires the manual version.
New areas and measures are added on the same model and loads, so each extra report costs noticeably less than the first.
If you already run Microsoft 365 and Entra ID, Power BI fits naturally: permissions use the accounts you have and Excel stays one click away. Metabase, hosted on your own server, suits teams who would rather query PostgreSQL without a per-user licence. Both can read from the same warehouse. You see the full cost of each before choosing.
It should not. Reads happen overnight or in small batches, ideally through the program’s API or a database copy, never as heavy queries mid-morning. A separate warehouse exists for exactly this reason: dashboards query their own copy while whoever is invoicing in a3ERP notices nothing.
In an account held in your company’s name at OVHcloud, IONOS, Arsys or the Spanish regions of AWS or Azure, or on your own server if preferred. Our work is covered by a data processing agreement. Access follows roles: directors see everything, each sales rep sees their own territory. Views are logged, which makes answering your DPO straightforward.
Mostly that depends on the state of your master records. Reasonably clean data means an early first report, with most time spent agreeing definitions. Thousands of duplicates turn clean-up into the main project. So we give timescales after seeing your data, not at the first meeting. Work is fixed-price or €75 per hour + VAT within an agreed cap.
The Red.es Kit Digital programme includes categories for business intelligence and analytics, but grants are handled only through registered digitalisation agents, and Apply is not one. If you are considering the scheme, check the current terms on the official website. Our quotes do not rely on any subsidy.
Tell us which reports are still assembled by hand and where the figures disagree. We review the sources and suggest where to start.
Your request is with us
Expect an answer within one working day. A reported fault that has halted your team is handled first.
No match found. Try another spelling, or go with the closest provincial capital: every job is done remotely, so the location makes no difference to what we deliver anywhere in Spain.