Guided analysis
Analyse your project yourself, in eight questions
You leave with an ordered plan: the steps to run, in which order, why each one comes at that point, an order of magnitude in days, and how long before a first report someone actually uses. The plan appears here, and you can have it by email.
- Questions, six minutes
- 8
- Forms before the result
- 0
- Expert consultants
- 11
- Projects delivered since 2019
- 100+
Almost everyone knows data has to be cleaned and a model has to be built. What projects get wrong is the sequence: aligning the reference data after the reports are built means taking everything back. So this analysis returns an order, not a catalogue.
What line of business are you in?
The steps are the same everywhere, the indicators much less so. Your answer makes the plan more precise on that point.
What are you starting from?
Count instances, not vendors: three companies on Business Central make three ERPs.
- ERPBusiness Central, NAV, SAP, Sage…
- CRMSalesforce, Dynamics, HubSpot…
- Payroll & HRSilae, Lucca, ADP…
- Line-of-business softwareSQL databases, bespoke applications
- FilesExcel, CSV, manual exports
What do you use today to follow your figures?
What already exists carries business rules nobody has written down elsewhere. Knowing them changes the order of the steps.
Do your customers, items or accounts live in several of these systems?
This question decides the order of the steps, and its answer costs the most when it arrives late.
Do two departments calculate the same indicator differently?
Revenue, margin, headcount: three indicators that often have two definitions under one roof.
How often must the figures be up to date?
Answer for the most demanding use, not for the average.
Who will read these figures?
Several answers are possible. Each audience adds a report, rather than a page in the same one.
Who will keep the reports running once delivered?
There is no right answer: all three are practised, and they do not cost the same.
A platform, with an alignment step
Several instances of the same system and reference data that diverges: this is the scope we work on most. Everything is decided by the order of the steps, and the alignment comes before the consolidation.
- 01
The inventory of your sources
1 to 3 daysOpening your 5 systems, finding the tables that carry the truth, measuring the volumes and the real state of the data.
Why here: A quote built on a description is a wrong quote. What your tables actually contain is discovered by opening them, and this step decides every step that follows.
- 02
Reading your current reports
2 to 5 daysAn inventory of the files in circulation, a record of the formulas that carry a rule, and a note of those still useful as they are.
Why here: Your spreadsheets carry business rules nobody has written down anywhere else. Reading them before building saves the arbitration that would otherwise follow go-live.
- 03
Aligning the reference data
The step projects skip5 to 15 daysMapping tables validated by your business teams, and composite keys so that every record keeps the entity it came from. Your existing setups stay exactly as they are.
Why here: This is the step projects skip, and the one that makes them start again. Consolidating before aligning produces figures that add up without meaning the same thing.
- 04
The data platform
10 to 25 daysIngesting the 5 sources, raw data kept exactly as received, daily orchestration.
Why here: Beyond three systems, copying the preparation into every report becomes unmanageable: each correction would have to be made as many times as there are reports.
- 05
Settling the business rules
The step projects skip2 to 6 daysA workshop with the teams concerned, then the chosen rules written into the transformation layer, once for every report.
Why here: Two departments that calculate margin differently are each right in their own context. Settling it requires both of them in the room, and one person with the authority to decide.
- 06
The semantic model
5 to 12 daysRelationships, measures and hierarchies, shared by every report and by Excel.
Why here: It is the piece that makes a figure mean the same thing everywhere. Without it, every report redefines its measures and the gaps between departments come back.
- 07
The reports
4 to 10 daysAn executive cockpit and business pages by department.
Why here: One report per use rather than one report for everyone. A board and a shop-floor supervisor are not looking for the same thing, and one page serves both of them badly.
- 08
Training your teams
1 to 3 daysA day of training on your own data, then a call on us over the first few weeks.
Why here: Your teams have to be able to change a report without calling us back. That is learned by doing, with someone beside them.
- You report up to 2 instances of the same kind of system. Charts of accounts, dimensions and reference data almost always diverge from one entity to the next: that is where the project is won.
Three things to do this week, without us
They cost nothing, and they save time for whoever takes the subject over afterwards, whether us or someone else.
- Find the spreadsheet whose disappearance would trigger a phone call within the hour. That is the one carrying the business rules to recover first.
- List your systems and, for each one, the person who knows what it contains. It is the hardest information to rebuild, and the most useful.
- Take a customer who exists in two of your systems and compare their records. The gaps you see are the ones the project will have to handle.
- Decide who will have the authority to settle a business rule. Without that person, the arbitration drags and the project with it.
This plan comes from our past projects. It does not replace a scoping: volumes, the real state of your data and the number of rules to settle weigh as much as the number of systems. Two hours with a consultant are enough to refine it, and it is free of charge.
Refine this plan in two hoursFrequently asked
