Portfolio OS.
The target operating model that closes the delivery discount: four functions installed in order, two aligned · transition mechanics · the four handover logs · engagement economics. Companion to Closing the Delivery Discount in B2B SaaS.
Published · Updated
Most turnaround plans for vertical B2B software start with new capital. This one starts with the delivery spend the company already carries: legacy engineering payroll, outsourced support overflow, integration overhead, and professional services run at a loss. Portfolio OS sets out how that spend is redirected onto the company’s own books, and what has to be true first.
The fix is an order: mandate before technology, talent alignment before hiring, team before rebuild. The four installations are the target operating model; the order of operations is the plan that gets there. The document covers the decision rights written down — one owner and a deadline per ruling — Week One and the first 100 days with their named outputs, the transition mechanics that prevent a hollow-out, the four handover logs that stay with the company, and the engagement economics that tie the operator’s return to incremental EBITDA. It closes with eight failure modes and the scale at which the model is proven.
Key findings
-
The fix is funded from delivery spend the company already carries — inflated legacy engineering payroll, outsourced support overflow, post-acquisition integration overhead and professional services at negative gross margin — redirected onto the company’s own books.
Method
-
Senior capability is sourced at 40–55% of tier-1 rates, employed by the company rather than rented from a bench.
MethodPublic compensation surveys, Exhibit 1
-
The order is fixed: mandate, then casting, then team, then rebuild and routing. Each decision that crosses the product–services boundary gets one owner and a clock.
Method
-
Phase 0 is a fixed 100-day engagement with named outputs: the listening report, the redirect envelope quantified, the knowledge-holder map, the baseline set and the first senior hires made.
Method
-
The 5–10 people who carry the legacy architecture are retained under written agreement by week 4. Their departure mid-transition is how programmes of this kind actually fail.
Method
-
EBITDA typically crosses zero around month 12 and builds to run-rate through months 24–36. A board that has not seen that curve in advance treats month 6 as failure.
Modelled
The order is the argument. Each step is worthless without the one before it.
Why this order
Mandate first, because decision rights are what separate installation from advice. Casting second, because the audit decides who is miscast before anyone is hired. Team third. The rebuild last — because a rebuild without the first three produces a newer version of the same asset.
- 1
Funding the Fix: Redirect, Not Add
The delivery spend trapped in the business, and why the model starts with money already being spent.
- 2
The Order of Operations
Mandate, talent alignment, team, then rebuild and routing — each step worthless without the one before.
- 3
Four Functions Installed, Two Aligned
Product and org design, people, operations, engineering — with strategy and sales aligned, not run.
- 4
Phase 0: The First 100 Days
Four stages, four named outputs, Week One and the six questions, and the 24-month shape with its gates.
- 5
The Transition: Changing the Engine in Mid-Air
Hire ahead, retain the knowledge holders first, two teams, customer continuity, freeze and rationalise.
- 6
Engagement Economics: Seats, Fees, Carry
The mandate in practice, the operating fee as non-recurring transformation cost, and carry on incremental EBITDA.
- 7
What We Don’t Offer, and What Breaks the Model
Scope stated plainly, what transfers and what is rebuilt, and the eight failure modes.
The six questions we ask every team in week one.
Week one changes nothing: meet the executive team, meet the top five customers’ relationship holders, read the last four board packs — and publish a one-page statement of what will be measured, against what baseline. The read that follows asks the same six questions at every team and every product line, so the answers can be compared. In a group built by acquisition, the differences are the finding.
Walk me through one real enterprise commitment, from the sales call to the release that carried it.
The actual route, not the policy. Ask for a named one.
Who decides what enters the core product — and who has to agree before they can?
Whether the boundary has an owner or a committee.
Why is this work done by this team and not another?
Structure inherited from deals rather than designed.
Where is your team doing manual data entry?
Manual entry marks every place your systems don’t connect.
What is bought in, and who manages that relationship?
Spend outside the delivery pool, and who controls it.
How do you know, inside the month, whether you are on plan?
Whether anything is measured at a useful frequency.
The read ends in three to five hypotheses, each with what would disprove it. Findings make a report. Hypotheses start a design.
Read the full paper.
You already have access.
This browser has already been given access to our papers. No need to enter your email again.