Most portfolio monitoring demos are a tour of dashboards. Dashboards are the easiest part of the product to build and the least useful part to evaluate, because every vendor's tool looks competent in a forty minute call.
Two questions separate the tools, and neither one is usually asked.
Quick Answer
When evaluating a portfolio monitoring platform, ask which systems it reaches and by what method, naming them rather than counting them, and ask what it does when two of those systems disagree. The first tells you whether the tool can see operations or only finance. The second tells you whether the numbers are governed or reconciled by hand. Everything else, including the dashboard, follows from those two answers.
Why This Matters More Than It Used To
Higher capital costs have made multiple expansion and leverage less reliable as return drivers. S&P Global's 2026 survey found 72% of GPs now rank operational improvements as their top value creation lever, and 60% agree higher capital costs are forcing a sharper focus on portfolio company performance.
Operational improvement requires operational visibility. A fund cannot drive sales productivity, pricing, retention or resourcing from a general ledger. It can only audit those things after the fact.
The same survey found 37% of GPs dissatisfied with the quality of the non-public operational metrics they receive. FTI Consulting's 2026 Private Equity AI Radar, based on 200 fund and operating leaders, found only 7% of portfolio companies have reached enterprise-scale AI deployment, with 36% using AI across multiple use cases and 35% naming talent availability as the primary constraint on scaling further.
Deloitte's 2026 survey of US private company leaders in the $100 million to $1 billion plus revenue band found 72% cite data quality or availability as an obstacle to realising value from digital and AI investment, while 63% are actively investing in digital transformation rather than piloting it. The money is going in ahead of the foundation.
Question 1: Which Systems Does It Reach, and By What Method?
Not the integration count. The list, and how each source is connected.
An "80+ integrations" number tells you nothing, because it does not say whether the eighty are eighty accounting packages. A platform that reads finance systems and calls itself portfolio intelligence will give you clean, reconciled, backward-looking reporting. It will not help you influence operations, because it cannot see them.
The list should cover four categories:
- Finance. NetSuite, QuickBooks, Xero, Acumatica, Sage, whatever your portfolio actually runs.
- CRM. Salesforce, HubSpot, Dynamics. This is where forward revenue lives.
- HRIS and resourcing. Headcount, attrition, capacity, cost of delivery.
- Product or delivery systems. Whatever carries your specific value creation thesis, whether that is usage data, support load or project delivery.
Ask for the named connectors in each category, and ask what happens for a portfolio company that runs something obscure. The answer to the second question tells you more than the answer to the first.
Question 2: What Does It Do When Two Systems Disagree?
Nobody asks this and everybody should.
A group with three CRMs has three sales stage models, so the same pipeline number means something different depending on where it came from. Finance and CRM will disagree on a booking. HR and finance will disagree on headcount cost. This is not an edge case. It is Tuesday.
There is one workable answer: the number that stands is the number the portfolio company has validated. Everything maps back to a single definition, the portco signs off on it, and the fund reads a figure the management team has already agreed is theirs. That matters twice over, because a number a portfolio company has disowned is a number nobody acts on.
Anything else is one of two things. Either the platform is a viewer, with reconciliation done by hand in a spreadsheet every month, which is the work you were trying to remove. Or it is quietly picking a winner between two systems and not telling you which, which is worse, because the number looks governed and is not.
Then ask what happens when a portfolio company changes its stage model, because they do, usually without telling anyone.
What Data Governance Actually Means Here
Governance sounds like a compliance word. In a fund it is narrower and more practical: settling the definition of every metric once, centrally, and holding it.
ARR, bookings, pipeline and headcount are defined differently in every portfolio company, and often in two systems inside the same company. Without one agreed definition per metric, benchmarking produces numbers that look precise and are not comparable, and forecasting produces confident projections built on a moving base.
This is the step funds skip, because it is unglamorous and it happens before anything appears on a screen. It is also the step that determines whether anything on that screen is worth acting on.
The Uncomfortable Part
If you hold a controlling stake and you are still running the position off finance data alone, you are choosing not to use the control you paid for.
Nobody says it that way, because finance data feels like the responsible choice. It reconciles, it audits, it satisfies the LPs. It also arrives too late and explains too little to change anything, and the hold period is where the multiple is actually built.
Some caveats worth stating plainly, because none of this is absolute.
If you are locked in, you are locked in. With three years left on a Chronograph or iLevel contract, nobody is ripping it out. Run it for finance and LP reporting, which is what it does well, and put the operational layer alongside it. You will carry two sets of definitions, which is a real cost, but it beats waiting for a renewal date.
If you are minority-only today, buy for the fund you are raising. One fund we work with holds minority positions, so financials are genuinely most of what they can get. They chose the broader platform anyway, because their next fund is majority, and at that point CRM and HRIS coverage stops being optional. A platform decision outlasts a fund.
And if reporting really is the requirement, finance-only tooling is sufficient and this whole argument does not apply to you.
In This Series
Previously: Part 1: Analytics in Private Equity, Where Finance-Only Data Runs Out
Next: Part 3: How Funds Get Operational Data Live. It covers the sequence that works, and what it looked like when one portfolio company's month-end reporting went from weeks to days.
Frequently Asked Questions
What should you ask a portfolio monitoring vendor?
Two things. Which systems it reaches and by what method, named rather than counted, across finance, CRM, HRIS and your delivery systems. And what it does when two of those systems disagree. The first tells you whether it can see operations. The second tells you whether the numbers are governed or reconciled by hand.
What is data governance in a private equity context?
Settling the definition of every metric once, centrally, and holding it. ARR, bookings, pipeline and headcount are defined differently in every portfolio company and often in two systems inside the same company. Without one agreed definition per metric, benchmarking and forecasting produce numbers that look precise and are not comparable.
What if we already have Chronograph or iLevel?
Run it for what it does well, finance and LP reporting, and add an operational layer alongside it. You will carry two sets of definitions, which is a real cost, but it beats waiting three years for a renewal date to get operational visibility.
We are minority-only, so we cannot get much beyond financials. Does this apply to us?
Today, less so. It applies to the next fund. Minority positions limit what a portfolio company will share, but the platform decision outlasts the fund, and moving to majority positions is exactly when CRM and HRIS coverage becomes the job.
Can AI help with private equity due diligence?
Partially, and less than most funds expect. S&P Global's 2026 survey found 31% of GPs have somewhat or fully integrated AI into due diligence, while 49% cite a lack of internal expertise and 43% cite data privacy concerns as the main barriers. Sixty-four percent still rate AI as ineffective for deal sourcing. Diligence analytics improves most when a fund has a clean portfolio baseline to judge a target against, which is a data problem before it is an AI one.
What is the biggest mistake firms make when adopting portfolio analytics?
Buying an AI layer before checking what it can see. Confident answers built on partial data are worse than no answers, because someone acts on them.
