Insights · Research

Why your monitoring dashboard isn't a decision tool

iLEVEL, Allvue, Chronograph and Cobalt are good at showing the KPI this quarter. None of them answer whether the IC decision that justified holding the position is still defensible. Here is the structural gap, and a real case where it mattered.

6 min readUpdated October 2026Capital Refinery research

The short answer

A monitoring dashboard renders the current KPI. A decision tool would show whether the investment committee decision that justified holding the position still holds. Dashboards lack a structured thesis to compare against, a binding from each KPI to the entry assumption it tests, continuous re-testing as data moves, and a trace from each figure back to its source document. That is a different unit of analysis, not a feature the monitoring stack can add.

700+firms using iLEVEL, the most widely deployed portfolio monitoring platform in private markets
$9.9Bsize of KKR's 2018 leveraged buyout of Envision Healthcare
$7Bdebt backing the Envision Healthcare buyout

The dominant monitoring stack

iLEVEL, now part of SS&C and Solovis, is the most widely deployed portfolio monitoring platform in private markets, used by more than 700 firms. Allvue, Chronograph and Cobalt cover much of the rest of the market, each leaning toward a slightly different slice: fund administration integration, GP-side monitoring, LP-side reporting. Together they form the post-close visibility layer for most institutional private capital, and inside their scope, the category is mature: operator KPI ingestion, dashboard rendering, fund-level rollups, LP-ready reporting, the quarterly board pack.

The problem is not quality. It is scope. The scope was drawn around one question: what are the KPIs this quarter? The more important question is the one nobody on the monitoring team is being asked to answer.

A monitoring dashboard tells you what the KPI is this quarter. A decision tool tells you whether the IC decision that justified holding the position is still defensible. Those are two different things.

The harder question a dashboard cannot answer

The harder question is: given the IC decision made a year and a half ago, is the case still working? Answering it needs four things a monitoring dashboard does not have.

01

A structured thesis to compare against

The dashboard renders KPIs. It does not hold the IC memo; the assumptions sit in someone's head and the model has not been opened in a year.

02

A binding from the KPI to the entry assumption

Without it, 'gross margin 36%' is a number on a screen. With it, it is the gap against an entry case that assumed 38%.

03

Continuous re-testing, not quarterly reconstruction

A figure that moves whenever the operator data moves, not a number the analyst rebuilds by hand for the board pack.

04

A trace from each figure to its source

The operator workbook that produced it, the model line that consumed it, the assumption it tests, the decision that still rests on it.

Question a dashboard answersQuestion only a decision tool answers
What was EBITDA last quarter?Is the entry thesis still defensible?
What is the leverage ratio today?Which assumption broke, and when?
Is the company hitting budget?How long until the breach is irreversible?

The dominant pattern in private equity monitoring is quarterly reconstruction: pulling last quarter's operator data into a refreshed model by hand, and comparing it against a static IC narrative.

What this looked like in a real case

In 2018, KKR acquired Envision Healthcare in a $9.9 billion leveraged buyout, backed by roughly $7 billion of debt. The investment case rested in part on surprise billing as a revenue driver. Through 2021 and 2022, operator KPIs, patient volumes, payer mix, revenue per encounter, flowed into standard monitoring dashboards quarter by quarter.

2018

KKR closes a $9.9 billion leveraged buyout of Envision Healthcare, with roughly $7 billion of debt.

December 2020

The No Surprises Act passes, effective January 2022, undercutting the revenue assumption behind the case.

2021 to 2022

Operator KPIs keep flowing into standard monitoring dashboards quarter by quarter.

March 2022

Independent researchers publicly flag the position as untenable.

May 2023

Envision Healthcare files for Chapter 11.

What those dashboards could not render was the thesis breaking structurally. The KPIs were visible the whole time. The question of whether the thesis underwritten in 2018 was still the thesis being operated was not a question the monitoring layer was built to answer.

Why the existing stack cannot just add this

The natural response is that monitoring vendors will simply ship a decision integrity feature. That is not how this works, and the reason is structural: the monitoring stack's primary object is the KPI, not the decision. Its data model is 'the operator submitted this number for this period.' It has no typed representation of the IC thesis to bind the KPI to.

Adding that connection would mean rewriting the data model from the inside out, and the existing customer base largely does not need that rewrite. It needs cleaner LP reporting and faster ingestion. That is why this layer has to be built separately: not a feature on top of monitoring, but a different unit of analysis, with a different data model and a different feedback loop.

What changes in the workflow once the gap is closed

When the layer exists, the dashboard does not go away. iLEVEL still ingests. Allvue still rolls up. The board pack still gets produced. What changes is that next to each KPI there is a second figure: whether the decision behind it is still defensible. Above the KPI, there is a third: the estimated time before the gap on the assumption that KPI tests becomes irreversible. The team does not rebuild the context from scratch each quarter; the structured record holds it.

Dashboard alone

  • Shows the KPI this quarter
  • No binding to the entry assumption
  • Context rebuilt from memory each cycle
  • The IC asks the deal team to reconstruct the case

Dashboard plus the decision layer

  • Shows the KPI next to the assumption it tests
  • Figures trace to their source document where a locator resolves
  • The record holds the context between cycles
  • The IC sees which positions are still defensible

The IC sees, on each position, which decisions are still defensible and which are operating in the gap. The deal team gets the time it needs to act. The dashboard tells you the quarter. The decision layer tells you whether you still own the thesis.

How Capital Refinery reads the reporting pack

Capital Refinery is built as that second layer, not a replacement for the monitoring dashboard. Each compliance certificate and registered deliverable is tested against its covenant, and the figures that matter for headroom and drift are surfaced as the pack arrives rather than rebuilt by hand.

  • Financial statements are spread with figures traced to source where a locator resolves
  • Ten advisory reconciliation checks flag inconsistencies before a figure reaches analysis; they do not gate it
  • Risk signals are graded against the firm's own configured policy thresholds, not a generic benchmark
  • Open conditions and covenant obligations sit on the same deal record as the ongoing read
Covenants screen: screen coming soon
Live product screen · illustrative data

Questions

What is the difference between a monitoring dashboard and a decision tool?

A monitoring dashboard renders the current KPI. A decision tool would show whether the investment committee decision that justified holding the position is still defensible, which needs a structured thesis, a binding from the KPI to the entry assumption, and a trace back to source documents.

Why can't iLEVEL, Allvue or Chronograph just add this?

Their data model is built around the operator-submitted KPI, not the decision. Adding a typed representation of the IC thesis would mean rewriting the product from the inside out, which their current customer base is not asking for.

What happened with KKR's Envision Healthcare investment?

KKR acquired Envision Healthcare in 2018 in a $9.9 billion buyout with roughly $7 billion of debt. The No Surprises Act, passed in December 2020 and effective January 2022, undercut a revenue assumption behind the case. Envision filed for Chapter 11 in May 2023, despite KPIs remaining visible on monitoring dashboards the whole time.

What four things does a dashboard lack for decision evaluation?

A structured thesis to compare against, a binding from each KPI to the entry assumption it tests, continuous re-testing as data moves, and a trace from each figure to its source document and the decision that still rests on it.

Does Capital Refinery replace tools like iLEVEL or Allvue?

No. It is built as a second layer next to them. It spreads financials traced to source, runs advisory reconciliation checks, grades risk against the firm's own thresholds, and keeps open conditions and covenant obligations on the deal record after close.

Accelerate your deal execution process.

Private markets, connected. Bring automation to your deal execution.

Founder-led access. Every working session runs on your actual documents, not a generic demo.