Spend visibility across a portfolio.

If you lead value creation or procurement for a private-equity portfolio, you do not run the purchasing of the companies you hold. You receive their files. Each month every portfolio company prepares a spend file and sends it over, usually by email.

Then the questions start. Who has filed and who has not. Whether this month has the same columns as last month. Whether "Marketing" means the same thing in two companies. Whether the savings an initiative claimed ever showed up in the numbers.

Until those are settled, the consolidated picture is assembled by an analyst for each review. A figure on it cannot be followed back to the file it came from, and last quarter's version is whatever was saved at the time.

Spend management products are mostly built for one company controlling its own purchasing: requests, approvals, cards, payments. A portfolio owner is in a different position, and a generic data model built for a single buyer does not describe it.

A reporting dashboard gets closer, but it draws whatever it is given. It does not refuse a bad file, remember which file a number came from, or hold savings, contracts and financial inputs against the same companies and categories.

What makes the picture yours is the detail a vendor will not carry: your category list and how deep it goes, the columns your companies can really produce, your formula for savings and for impact on value. Those change as the portfolio changes. When you own the system, a new company or a revised formula is your decision, made when you need it.

In the system that runs at the private equity firm, each company is registered once, under the exact name its files use. An uploaded file is checked twice: first its format, then its data. A file that fails is recorded as rejected, with its errors, and none of its rows are counted. A file that passes is stored in full.

The spend analysis reads from validated rows only: spend over time, by category, company and supplier. Contracts, savings initiatives and a quarterly model of impact on value are held beside it, against the same registered companies. Because each of those is an event on one record:

• Trace. Spend totals can be filtered by company, date range, category and supplier, and every accepted file is listed with its row count, total spend and supplier count.
• Correct on the record. A corrected file is a new upload, and the rejected one stays on the record with its reasons. Uploading again for a month that already has data asks for confirmation and states how many records will be replaced.
• Reports that draw on history. Savings planned against realised, and each quarter's inputs as they stood, are read from what was recorded at the time.

Once validated spend is on record, the next step is an addition: another portfolio company, a further category level, a new view for the board. Each reads the history already held. It extends without a migration.

After the first build the framework becomes available to you. Your own engineers and product people can build the next system on it, or we carry on building for you. Both stay open, and the terms are settled in the first conversation, scoped to the engagement.

How the two stages workThe record that carries forward