How to Cut Software Spend in a Portfolio Company Without a Risky Migration

Macropyre Team

We think the biggest single software saving in a portfolio company is one expensive system that fits the business badly, and you can retire it a piece at a time, next to the old one, without a risky migration.

We still start with a seat audit, because it costs little and the sheet it produces shows you which system fits worst. The real effort then goes into replacing that one system one workflow at a time, so that any problem stays small and the old system is there to fall back on.

The bill grows because nobody owns the whole list

A sales manager puts a prospecting tool on a company card, the operations team adds a scheduling app the same quarter, and each one renews on its own date long after the person who bought it has moved on to other problems. Finance sees the charges and IT sees part of the list, but nobody holds all of it, so the question of whether the company still needs a tool never comes up. In one SaaS management report, business teams controlled 81% of SaaS spend, and IT managed only 15% of it directly. The same report put unused licenses at 36% on average, measured against recommended use levels.

The audit is cheap, and it points at the real problem

Build one sheet per company before any renewal talk starts. The IT list will be short of the truth, so pull every software vendor from accounts payable and from company card spend, then match seats to real logins from single sign-on or each vendor's admin page. For each tool, record:

  • what it does, which team uses it and who owns it inside the company
  • seats paid, and seats active in the last 90 days
  • yearly cost, with add-ons, overage and support tiers
  • contract end date, notice period and auto-renew terms
  • any price increase already written into the contract

The seat gap is the first saving. Any seat with no login in 90 days comes off at the next renewal or true-up, and people who log in once a month can move to a cheaper or shared seat. Two file-sharing apps doing one job become one, and a team on a premium tier for a single feature should drop down if nobody uses the rest.

Sort the sheet by notice date instead of end date, because a contract with 90 days notice has its real deadline three months before renewal. The reminder email that shows up after that window has closed commits you to another term, so open each talk early and bring the login numbers with you.

The system with the most workarounds is the one to replace

Now sort the sheet by cost and hold the top few lines up against how people actually work. Look for the system with a Monday morning export: someone pulls last week's orders into a spreadsheet, fixes what the system got wrong, and types the result into billing, and has done it every Monday for years. We would expect cutting seats on that system to save little, because most of what it costs sits in its add-ons and in the hours spent around it.

We would not move a portco onto a new platform in one cutover. A big-bang migration puts every workflow at risk on the same weekend, and a slip of a few months can run straight into the sale process.

Replace it one workflow at a time while the old system keeps running

Software people call this the strangler fig pattern: you retire the old system by replacing one piece of its work at a time, and small replacements over time carry less risk than one large change. For a portco, each piece is small enough to go live with little risk, and you pick the order, so the pieces with the best return go first.

  1. Pick one workflow inside the target system, such as one ordering or billing step.
  2. Build the replacement for that one workflow, matched to what the team does today.
  3. Run both side by side with the old system still live, and compare the outputs every day until they match.
  4. Move that one workflow across, then pick the next.
  5. Turn off the license when the last piece has moved, timed to the contract end date with notice given on time.

Your CFO will check the saving, so set the baseline before you start. It is the old system's yearly cost with add-ons, support and the staff hours its workarounds take, set against hosting, support and upkeep for what replaced it. Book the difference only once the old invoice actually stops or shrinks, keep one-time savings apart from recurring ones, and leave out any avoided price increase the contract did not have in writing.

A hold period has room for this, though not every system deserves it

In our plan, the first month goes to the sheet and the renewal calendar in each company. The second takes the seat savings and maps the workflows inside the system you mean to replace, and by month three or four the first piece is running next to the old one. The rest moves across after that, with the final cutover set for the contract end date. Across a portfolio, run the audit everywhere at once and start with the company whose worst-fit system costs the most.

Some systems should stay where they are. We would not replace payroll, tax filing or anything an auditor has to sign off on, because a proven vendor carries that risk better than new software does. We would leave email, office suites, video calls and e-signature alone and cut seats there instead. If the system is small and swapping the whole of it is simple, the pattern may not be worth it, and if the sale is a few months off, take the seat savings and leave the build to the next owner.

We build the replacement inside your business, next to the old system, so if you can already name the one with the Monday export, book a call and we will start there.

Sources