What does transition-in really cost?
Hypothetical, to show the shape: a three-year service desk and infrastructure contract opens with a ninety-day transition: knowledge transfer from the incumbent, tooling migration, security onboarding, staff hiring and clearance. Revenue during transition is partial or nil. Effort is full. Firms that price transition as two weeks of shadowing discover a quarter of unrecovered cost in month one.
Incumbent cooperation is assumed in too many bids and guaranteed in almost none. Read the transition schedule for what the outgoing supplier must provide, what the buyer will provide, and what happens when neither arrives. An incumbent-written tender (how to spot one) makes transition hostility the base case, not the risk case.
What do service credits do to the margin?
They convert missed SLAs into deductions against steady-state revenue: monthly, compounding, and sometimes uncapped individually while capped in aggregate, or the reverse. Identify every metric, its measurement period, its trigger, its calculation, its cap, and how overlapping failures interact. Then model two simultaneous failures, not one. The worked method is on KPIs and abatement; this page does not redo the arithmetic, it tells you to run it before the gate.
Termination rights sit beside the credits. Repeated SLA failure can give the buyer exit rights long before the term ends, which strands the transition investment with no steady state to repay it. Read termination alongside credits, always.
What does exit demand?
Handover assistance, knowledge transfer, parallel running, data extraction and return, staff transition obligations, scoped in the contract schedules you sign at award, performed years later when the account is at its thinnest. Price exit effort now: the team that hands over, the documentation the buyer can demand, the months of dual running. Security terms (certifications, controls, personnel requirements) run through all three phases; hold them at close or no-bid.
This page excludes industrial-control and SCADA services. Different risk universe, different evidence. IP ownership and licence terms belong to their own reading; do not collapse them into the credits analysis.
What does the gate look like?
Transition costed as a project with its own hours. Steady state modelled with two-failure credits applied. Exit priced as handover effort. Security evidence held today. If the three phases together do not clear margin, or if transition depends on an incumbent’s goodwill, the call is no.