What your committee will ask

How do we know the numbers are right?

The software checks every figure, we check the totals before you see them, and you can check any figure yourself in Excel.

What the software checks:

• Every bridge, whether it explains sales, margin or cost of goods, adds up to the change it explains. Anything it can't split is shown on its own.
• Totals are added up from the rows beneath them, so the headline always matches the detail.
• Where a product has no cost, its margin is left blank and the product is named, so nobody reads a margin that isn't there.
• The ledger and the sales figures are compared account by account, and any account that needs a look is flagged.
• Each month's cost data is checked before it's used: every record must match a known product and plant, and stock must add up. A month that fails is held back with the reasons listed.
• After each month, a check shows that everything each cost centre spent reached a product, and the cost of idle capacity is shown on its own line.

What we check: before you see anything, we compare the headline totals with the figures you already report, and explain any difference.

What you can check: key tables export to Excel with their formulas, so your team can see how each important figure is calculated and work with it.

We already have Power BI. Or Tableau. Or a data team.

Good, keep them. Our software sits between your ERP and your BI tools, and fills the gap between the two.

Your ERP records transactions. Your BI tool draws charts from whatever data it is given. In between, most companies have a pile of spreadsheets: product cost worked out level by level through the recipe, rebates spread over the products they belong to, a demand plan with its sign-offs, a budget built from that plan. That middle layer is what we build.

It feeds your BI tools and your data team. They get figures that have already been checked, with exports and data feeds they can load into any reporting tool they like.

Who owns the software, and who owns the data?

Your data, your definitions and everything built for you are yours, exportable at any time. The engine underneath is ours and licensed to you for as long as you subscribe; if you stop, you keep your database and every export.

Your company gets its own copy of the system, with its own database. Your figures are never used to train models, never pooled into a benchmark and never shown to another client.

The exact terms are in the contract, and you'll see them before you sign.

Where does it run, and who can see our numbers?

It runs wherever we agree it should, including on your own servers. The hosting terms are part of the subscription.

• Your people sign in with their own named accounts, and nobody gets in unless an administrator adds them. Single sign-on through your company directory can be part of the build.
• We see your numbers where we host the system or load data for you. Where it runs on your own servers, your IT team decides what access we have.
• Nobody else sees them unless you ask for it.

Your IT or security team can have a full technical description on request.

Does any of our data reach an AI model?

Not by itself. Nothing leaves your system unless you ask for it.

If you'd like an AI model plugged in, to draft commentary or to answer questions about your plan, that's an option we agree in the scope, and you choose it. The same goes for reports sent by email on a schedule, or alerts. Your data is never used to train models.

What does it cost, to build and to keep running?

A paid pilot, then a fixed fee for the build, then an annual subscription that keeps it running. Every figure, and what each one covers, is on the prices page.

How do we start?

With a paid pilot: you choose one question, and we answer it on your own data with working screens. How an engagement runs has each step.

Our data is a mess. Is that a problem?

No. Most companies' data is messy, and the software is built for it.

• The software matches your column names to its own and a person confirms the match, so your files don't need to change.
• It reads files the way finance systems write them: decimal commas, negatives in brackets, day-first dates, accented characters.
• It starts with what you have, and adds more as more data arrives. A sales file with months and amounts is enough to begin.
• Where something is missing, it says so and leaves the gap visible.

One decision stays with you. When the same product or customer has different codes in different systems, someone senior has to say which one wins. We size that work in the scope, so it doesn't surprise anyone halfway through.

Can you work with our ERP?

Yes. We can connect to your database or data warehouse, or start from the exports your team already produces. Either way, every load is checked before anyone relies on it.

Product cost needs a little more than sales analytics: stock movements, production orders, what each order used, the recipes and each cost centre's costs. These are standard reports in any ERP, and we give your IT team a list of exactly which tables and fields we need.

What if we want to stop?

Then you stop, and you keep your database and every export.

The pilot has its own end, and going on to the build is your decision. Each build has an agreed scope, and the subscription's terms, notice included, are in the contract. At the end there is an agreed run-off period so you can migrate, and the same contract says how your data is returned or deleted.

What do you need from us?

A few hours from three people, one of them senior enough to settle what a definition means.

• Someone who can get the data out of your ERP or data warehouse.
• Someone who owns the definitions, usually the controller.
• Someone who will look at a finished screen and say “that's not what our gross margin means”.

You don't need a project manager, a steering committee or a requirements document to begin. A description of what you need is enough.