Executive Summary
Each company in your portfolio defines and reports its core metrics differently. This makes tracking performance across the group a manual and rather error-prone exercise.
You are making multi-million-pound capital allocation decisions using inconsistent data. This slows you down, can undermine your credibility with LPs, and makes quarterly Investor Reporting a difficult process.
The answer isn't another business intelligence tool. It's a decision to standardise: a 'Group Level' reporting structure built on a common Semantic Layer that ensures a single definition of truth across all your companies.
Why reporting breaks down across a portfolio
You’ve just acquired your fifth B2B SaaS company. Each is successful in its own right, a leader in its field. The investment logic was sound. But now, in the first quarterly review, you ask a simple question: "How does PortCo A's Net Revenue Retention compare to PortCo C's?"
There’s a pause. The CFOs of both companies present healthy-looking numbers, but the footnotes tell the story. One includes upsells from non-core services, the other doesn't. One calculates it monthly, the other quarterly. They aren't comparable. You haven't acquired a portfolio; you've acquired a collection of data silos, each speaking its own language.
In my experience, this is a common problem when companies scale, just amplified across a portfolio. The localised, often scrappy, reporting that got each company through its early funding rounds becomes a serious handicap at the group level.
The problem isn't the tools, it's the definitions
This isn't just an inconvenience. It points to a more fundamental problem with the operating model. You have likely invested in companies with modern data stacks: Snowflake, dbt, Looker. But you’ve simply inherited their organised chaos. Automating a broken process just generates inconsistent data at speed, and adding ten sources of that together doesn't create clarity. It creates a misleading picture.
The issue isn't the dashboard tool. It's that each company's business logic is hidden in a web of undocumented SQL scripts. The problem isn't that the analysts aren't working hard. It's that the Metric Definition of 'Churn' was never agreed upon and mandated from the top.
I saw this with a private equity-backed group that had acquired three e-commerce brands. They spent two months trying to consolidate performance for their board meeting. The result was a 100-page slide deck full of asterisks and footnotes, which understandably eroded the board's trust. They had data-rich, insight-poor companies, and the manual effort to bridge the gap was not sustainable.
How to create a unified reporting structure
To fix this, the group has to stop thinking like a collection of individual companies and start acting like a unified portfolio. This means shifting focus from cleaning up data to getting the data architecture right. The solution, in my view, is a top-down decision to create a single, unified reporting structure.
1. Standardise the data model First, we help define a set of 'Group Level' KPIs that are not up for debate. This can't be a democracy. The Operating Partner, usually with our guidance, defines the 20 or so core metrics that matter for creating value: ARR, NRR, CAC, LTV, Gross Margin. This becomes the standard every portfolio company must adopt.
2. Implement a governed semantic layer We don't suggest ripping out each company's BI tool, as that's too disruptive. Instead, we build a thin, standardised semantic layer that sits on top of their data warehouses. This layer contains the agreed-upon logic for the mandated KPIs. It ensures that when someone in PortCo A looks at 'ARR', it is mathematically identical to the 'ARR' in PortCo B. This is the foundation of a Single Source of Truth.
3. Automate the roll-up With a consistent semantic layer in place, aggregation becomes straightforward. We can now build a single 'Group Level' dashboard suite that provides a real-time, like-for-like comparison of performance. Preparation for board meetings can shrink from weeks of manual work to a few hours of proper analysis. This level of rigour is also very helpful for any future Due Diligence process.
The political challenge of standardisation
Let's be direct. Implementing this is more of a political challenge than a technical one. You are telling successful founders and CFOs that their way of reporting, which has worked for them so far, is no longer sufficient. Some will resist. They will often argue their business is 'unique'.
You have to be prepared to enforce the standard. This requires a robust Data Governance framework, not as a box-ticking exercise, but as the way to make your operating model work. It often feels like you're moving slower for one quarter to be able to move much faster for the next three years.
The reward for this difficult work is control. It's the ability to confidently allocate capital, identify underperforming assets, and report to your LPs with clarity. It's the shift from managing a collection of companies to operating a true portfolio.