The short answer
To standardize board package data across a portfolio, agree on one board pack template and a short KPI dictionary first, then replace emailed spreadsheets with one structured way to collect data: direct ERP connections where possible and a validated upload template everywhere else. Check data as it arrives, assemble packs automatically, and roll the process out in waves, starting with three to five companies.
Key takeaways
- Standardize definitions and the template before choosing any tool.
- Collect data one way, not forty ways. Direct connections first, a validated template second.
- Validate on arrival so errors are fixed by the company, not discovered by the deal team.
- Assemble packs automatically from the collected data instead of building them by hand.
- Pilot with a few companies, then onboard the rest in waves.
Ask a fund or holding company finance team what board season looks like and the answer is usually the same: dozens of companies, each sending numbers in its own format, a shared inbox full of spreadsheet versions, and a small team stitching it all together before the meetings.
The scale keeps growing. PitchBook counted more than 13,143 companies in U.S. private equity portfolios at the end of 2025. With 40 portfolio companies reporting monthly, that is 480 packages a year to collect, check and assemble. Doing it by hand does not scale.
Why is board package data so hard to collect across a portfolio?
Because every company reports differently, and nothing forces them to report the same way. The common problems:
- Different formats. Each company sends its own spreadsheet layout, PDF or slide deck.
- Different definitions. EBITDA adjustments, backlog and same-store growth are calculated differently from company to company.
- Late and partial submissions. The pack waits on the slowest company, and follow-ups happen by email.
- Version confusion. Corrected files arrive after the first version has already been used.
- Manual assembly. Someone copies numbers into the master pack, which is where most errors start.
The underlying cause is the same one behind most multi-entity reporting problems, covered in why your companies still run on five versions of the same spreadsheet.
What should a standard board package include?
A standard package should be short, consistent across companies and built from data that can be collected the same way every month.
| Section | What it covers | Where the data comes from |
|---|---|---|
| Financial summary | Revenue, gross margin, EBITDA, against budget and prior year | ERP general ledger, mapped to a standard chart |
| Standard KPIs | Five to ten metrics defined the same way for every company | ERP and operating systems, per the KPI dictionary |
| Cash and covenants | Cash, working capital, debt and covenant headroom | ERP, bank and lender reporting |
| Operational drivers | Two or three metrics specific to the business | Operating systems or the upload template |
| Management commentary | Results explained, risks, decisions needed | Short structured text from management |
The KPI dictionary is the most important document in the whole process: one page that defines each metric, its formula, its source and its owner. Without it, standard packs just present inconsistent numbers more neatly.
How do you collect the data without emailing spreadsheets?
Use the most automated method each company can support, and the same fallback for everyone else.
| Method | Best for | Tradeoff |
|---|---|---|
| Direct ERP connection | Companies on cloud or well-supported ERPs | Setup work up front, then little effort each month |
| Validated upload template | Companies on older or less common systems | Monthly effort for the company, but consistent and checked |
| Structured commentary form | Management narrative, risks and decisions | Needs clear prompts so answers are comparable |
Whichever method a company uses, the data should land in one central place, be checked automatically against rules (totals tie, values fall within expected ranges, required fields are filled) and be returned to the company with specific questions when something fails. The connections are integration work, and the checks, model and packs are data and BI work. We cover the mapping side in detail in how to report across portfolio companies on different ERPs.
Still building board packs from emailed spreadsheets? We will map your current process and the fastest path to one standard pack. Book a working session ›
How do you roll this out across 40 companies?
Roll it out in waves. Trying to switch every company at once usually stalls on the slowest few.
- Define. Agree on the template and the KPI dictionary with the deal team and a few portfolio CFOs.
- Pilot. Bring three to five companies onto the new process, including at least one with messy data.
- Tune. Adjust validation rules, the template and the calendar based on what the pilot companies struggled with.
- Onboard in waves. Add the rest in groups of eight to ten, starting with those whose ERPs connect most easily.
- Retire the old process. Once a wave is live, stop accepting the old spreadsheets from those companies.
New acquisitions should go straight onto the standard process as part of onboarding. Our process follows the same pattern: prove it on a few, then scale.
Where does automation help most?
Automation helps most in the steps that are repetitive and rule-based: reminders and status tracking, reading uploaded files, running validation checks and assembling the pack. AI and automation can also flag numbers that look inconsistent with prior months and draft a first version of commentary from the data for management to edit.
The judgment stays with people. Management explains results, the deal team decides what matters, and every figure traces back to source data.
What usually goes wrong?
- Starting with the tool. Buying software before agreeing on definitions just moves the inconsistency to a new screen.
- Too many KPIs. A 40-metric pack is hard to collect and rarely read. Keep the standard set short.
- No feedback loop. If companies never see what failed validation, the same errors repeat every month.
- No owner. Someone at the fund or holding company has to own the template, the dictionary and the calendar.
Fixing the close and submission process inside each company is business performance work for private equity, and it often shortens the whole cycle. See how we work with funds and holding companies on our private equity and holding companies page.
Frequently asked questions
What should a portfolio company board package include?
Most board packages cover a financial summary against budget and prior year, a short set of standard KPIs, cash, working capital and covenant status, a few operational drivers specific to the business, and management commentary on results, risks and next steps. The financial and KPI sections should use the same definitions across every company.
How do you get portfolio companies to submit board package data on time?
Make submission easy and predictable. Publish a calendar with fixed due dates, give every company the same template or a direct data connection, send automatic reminders, and show submission status to the deal team. Removing manual rework on the company side does more for timeliness than escalating late submissions.
Should we use portfolio monitoring software or build our own process?
Portfolio monitoring software works well when your KPIs and pack format are fairly standard. A custom process on your own data warehouse and BI tool fits better when you need unusual metrics, direct ERP connections or packs that match your firm’s format exactly. Either way, agree on definitions first.
How long does it take to standardize board packages across a portfolio?
Defining the template and KPI dictionary usually takes a few weeks. Piloting with three to five companies takes another month or two, and rolling out to the rest in waves depends on the number of companies and how clean their data is. Most of the effort is in definitions and mapping, not tooling.
Can AI help with board package preparation?
Yes, in specific places. AI can read submitted files and extract figures, flag numbers that look inconsistent with prior months, and draft first versions of commentary from the data for management to edit. It should not replace management’s own explanation of results, and every figure still needs to trace back to source data.
