Mosaic Lab · Private equity
Data-first due diligence for private equity, from buy-side through the hold to exit.
Why now
Illustrative split. Sources: Bain & Company, 2026; StepStone, cited in McKinsey & Company, Global Private Markets Report 2026.
Read the bars as points of IRR: each adds up to the 20% target, split by where the return comes from. The number on the right is a different thing: the annual EBITDA growth the deal needs to reach 20%.
For deals made between 2010 and 2022, StepStone attributes 59% of returns to leverage and multiple expansion and 41% to revenue growth and margin improvement (StepStone, cited in McKinsey & Company, Global Private Markets Report 2026). The last-decade bar uses that mix.
Today, debt costs 8–9% and funds 30–40% of the price, against half at 6–7% in 2015 (Bain & Company, 2026). The same 2.5x over five years now needs 10–12% annual EBITDA growth, against about 5% in 2015 (Bain & Company, 2026). The value creation plan carries real execution risk: 71% of deals in Bain's study fell short of their projected margins (Bain & Company, 2026).
12 is the new 5, so diligence has to test what really drives earnings.
Our approach
Our approach mirrors how investors think, not how consultants report.
Understand the investment thesis and the value drivers.
Build the customer-product data cube and define the right metrics.
Test each driver using granular analytics.
Findings that confirm, challenge or reframe the thesis.
How we engage
Value drivers written as testable hypotheses
Data needs mapped to each hypothesis, data room indexed
Built from source, tied to the management accounts
Revenue, margin, costs, cash and assets
The gap between story and evidence, with model inputs
Evidence for each driver, traced to source
We support your deal team at whichever steps you need, from a single data cube to the full process.
What we offer
Buy
Test the equity story before IC
Hold
Track the thesis while you own it
Exit
Evidence the story before buyers test it
Build capability
The same method, taught to your team
Delivered asExcel databooksinteractive dashboardswritten reportsmanagement presentations
The analytical toolkit
Revenue
Contracts
Margin
Costs
Earnings
Working capital
Assets
Cash and debt
All eight areas run off the same reconciled data, built from transaction-level records in the data room: invoices, the general ledger, payroll, the orderbook, the fixed asset register and bank statements. Messy data is a finding in its own right.
Revenue is where most of the return now sits. The spotlight that follows shows how we test it.
Our approach applied to one driver
How we test whether the growth in the plan is real. Revenue is one driver among several; the same pattern applies to margin, asset economics or conversion.
Spotlight: revenue · 1 of 3
Illustrative: one year's revenue bridge for two businesses, each starting at 100.
Most revenue diligence stops at a CAGR, the top ten customers and a monthly trend. That shows how fast a business grew, but not where the growth came from or whether it will repeat.
Business A keeps its customers and sells them more: it loses 4% and wins 14%, so its growth is cheap to sustain and likely to repeat. Business B loses 17% of its revenue every year and has to win 27% to grow at all; any slowdown in new wins shows up at once as decline.
The market prices the difference: public SaaS companies with net revenue retention above 120% traded at a median 9.3x revenue, against 3.1x below 100% (Software Equity Group, 2024; direction only).
Upsell is more of the same products; cross-sell is products a customer has not bought before; downsell is lower spend from customers who stay; churn is customers who leave.
Same 10%, very different businesses. Telling them apart takes revenue by customer, by product, by month.
Spotlight: revenue · 2 of 3
Revenue by customer, by product, by month, so you see the drivers the way management runs them.
New, upsell, cross-sell
Gross and net retention
Usually in the long tail
Cohorts by year won
Products per customer
Price versus volume
It closes the gap with management, and gives you the record to test the forecast against.
The cube is built from the transaction or billing data in the data room and reconciled to the management accounts and audited revenue before anyone draws a conclusion from it.
Read: You need a revenue cubeSpotlight: revenue · 3 of 3
Illustrative · Three questions every IC memo should answer
Which sources of growth does the forecast rely on, and in what mix?
Has the business ever delivered that mix?
Are the newest customers behaving like the oldest?
Every growth line in management's forecast maps to a part of the bridge, whether or not the model shows it: new customers, upsell and cross-sell, and retention. Set each against what the business has actually delivered, and the gaps show where the IC discussion belongs. In this illustration the plan needs twice the best year of new customer wins. That may be possible, but it is an assumption to underwrite, not a number to accept.
If the IC memo cannot answer the three questions, the revenue diligence is not finished.
How we work
Our workbench lets one senior practitioner hold this standard across a full workstream.
Any figure in a deliverable can be walked back, step by step, to the cell in the data room it came from.
Who you work with
Private equity investing: financial and commercial due diligence, LBO models and deal structuring.
Buy-side and sell-side data workstreams for PE: customer cubes, ARR and cohort analysis, buyer dashboards.
Reviews of vendor financial due diligence, commercial due diligence and information memoranda to set targeted buy-side workplans.
The person in the meeting does the work.
Contact
Tell us where the deal stands and which steps you need, from a single data cube to the full process.