 Delivery Architecture — Strategy to Daily Work |
Delivery Architecture — Strategy to Daily Workhttps://www.compassperformanceinc.com/library/delivery-architecture Program & Project Delivery Delivery ArchitectureStrategy does not become results in one step. It descends through six altitudes, and every altitude answers one question, holds one decision right, runs one cadence, and produces one artifact that the altitude below inherits. Work travels back up the same structure: a blocker at the daily altitude escalates until it reaches the altitude that owns the decision. The architecture Six altitudes, each with one question it alone can answerMost delivery failure is an altitude error: a steering group debating a task, a team asked to decide what should be funded, a benefit owned by nobody because it sat between two projects. Naming the altitude fixes most of it. Altitude 01 · DirectionEnterprise strategy3–5 years, reviewed annually “ What must be true about this enterprise three years from now? ” The gate Pass criteria — Objectives are measurable, few, and owned by a named executive. Evidence — Objective set with baseline and target measures; concentration read. Skip it and — Fifty priorities, none funded — every altitude below inherits the confusion. Do the work here Altitude 02 · SelectionPortfolio12–18 months, re-sequenced quarterly “ Of everything we could do, what will we do — and in what order? ” Pass criteria — Every item scored on the same scale, sequenced against real capacity, and declines recorded with a reason. Evidence — Scored intake register, capacity read, council decision log. Skip it and — Everything started, nothing finished, and the best people spread across nine efforts. Altitude 03 · IntegrationProgram6–24 months “ Which projects must land together for the benefit to be real? ” Pass criteria — Each benefit has a measure, a baseline, an owner and a realisation date; every cross-project dependency has an owner and a date. Evidence — Benefit register, dependency board, escalation log. Skip it and — Projects each declared green while the benefit the program promised never arrives. Altitude 04 · DeliveryProject6 weeks – 12 months “ What exactly are we delivering, by when, with whom, at what cost? ” Pass criteria — Chartered scope with success measures, a method chosen, gates carried, and a sponsor decision at each one. Evidence — Project workspace record: charter, approvals, issues, risks, status reports. Skip it and — A project with energy and no baseline — success can never be proven. Altitude 05 · IncrementSprint / iteration1–4 weeks “ What will be finished, and demonstrable, by the end of this iteration? ” Pass criteria — One sprint goal, a commitment inside capacity, work in progress held under the limit, and the increment demonstrated. Evidence — Sprint board with committed and completed points, blocker record, review note. Skip it and — A sprint that ends with everything eighty percent done and nothing usable. Altitude 06 · ExecutionDaily workToday “ What did I finish, what am I doing next, and what is in my way? ” Pass criteria — Every task has one owner and one due date; every blocker older than one day has been escalated by name. Evidence — Stand-up log, task list, escalation trail into the altitude above. Skip it and — Busy days, no movement — effort spent where it changes no measure. Down and back up How one objective becomes today’s task — and how a blocker gets a decisionCascade down Escalation up Not all work deserves the same route Transformative, incremental, feature, or task — each enters at a different altitudeGovernance that fits an ERP program will suffocate a two-week enhancement. Governance that fits an enhancement will lose an ERP program. The route is chosen deliberately, and the finish line is stated before the work starts. The five tests Is it a project, or is it a task with ambition?Test 01Is there a measurable outcome? If no measure moves, it is a task or a wish — not a project. Test 02Does it change how work is done? Changing the way the work happens is a project. Doing the work is operations. Test 03Does it need more than one function? Cross-functional effort needs a charter. Single-function work usually needs a due date. Test 04Is there a sponsor who can pay for it? No sponsor, no money, no project — however good the idea. Test 05Would you defend it at a portfolio council? If it cannot survive scoring against everything else, it should not consume capacity. Where the work is done The desks that carry each altitudePortfolio Council DeskScore every candidate on one scale, sequence Now / Next / Later, test the sequence against real capacity, and record every decline with a reason. Open the desk → Program Integration BoardBenefit register with measures and owners, cross-project dependencies with dates, integrated milestones, and escalations to steering. Sprint & Daily BoardOne sprint goal, a commitment inside capacity, a hard WIP limit, and a stand-up log where blockers escalate by name. Project WorkspaceThe chartered record: stages, ground rules, members, issues, risks, approvals and status reports. Project Delivery MethodsChoose DMAIC, a predictive plan, or Scrum — and inherit that method's gates. Process ArchitectureMap current state, run the Process X-Ray, and rank the improvement opportunities inside scope. Where execution actually breaks.White space between functions, accountability that follows the outcome, and cadence that closes commitments. |
YOUR NEXT STEP
Start with a confidential conversation about what you want to achieve
Opens the approved calendar in a new tab. An appointment is confirmed only by the calendar provider.