The First 30 Days of an OCI Engagement Done Right

Most OCI engagements do not fail in month six. They fail in month one and get discovered in month six. The patterns are remarkably consistent: access requests that drift for weeks, a kickoff deck that never turns into a plan, a baseline that nobody measured, and a senior architect who was vivid in the proposal and invisible after the signature. By the time the steering committee notices, the budget is half spent and the difficult conversation has become a contractual one. None of this is bad luck. It is the predictable result of a first month run without structure, and it is entirely avoidable.

This article is part of our complete guide to hiring an OCI partner, and it picks up where the selection process leaves off. Across 500+ OCI engagements we have seen that the quality of the first 30 days correlates with the outcome of the whole engagement more reliably than any credential, reference, or rate card ever does. The good news is that a well run first month follows a recognisable shape. Once you know the shape, you can hold any provider to it from day one, and you can spot trouble while it is still cheap to fix.

Week 1: access, governance, and alignment

The first week is administrative, and that is exactly why it is diagnostic. Nothing in week one requires brilliance. It requires discipline, and a provider who cannot be disciplined about the boring parts will not suddenly become disciplined when the work gets hard.

Tenancy access with least privilege

The provider should arrive with a written access request: which compartments, which policies, which groups, and why. Good firms ask for least privilege by default, scoped to the compartments they will actually touch, with break glass procedures documented for emergencies. A provider who asks for broad administrator rights across the tenancy on day one is not saving time. They are telling you they have no access model, and that the audit trail you will need later does not exist in their process.

Named contacts, RACI, and cadence

By the end of week one you should have a one page governance sheet: every named person on both sides, who is responsible, accountable, consulted, and informed for each workstream, the escalation path, and the meeting cadence. A weekly delivery call, a steering review every two to four weeks, and a shared channel for daily traffic is the normal pattern. The tooling should be agreed in the same conversation: where the backlog lives, where documents live, where decisions get recorded. If you wrote a structured request for proposal during selection, much of this was already promised in writing, which is one of the quieter benefits of the approach we describe in our guide to writing an OCI services RFP. Week one is where you check the promises against reality.

The single most important check in week one is also the simplest: are the people in the room the people from the proposal? The bait and switch usually happens here, quietly, framed as a temporary staffing adjustment. Call it out in week one, while substitution clauses still have teeth and while switching providers is still a realistic threat.

Week 2: discovery and baseline

Week two is where the engagement earns its keep for the first time. The provider should be heads down in your estate, and the output should be written, specific, and slightly uncomfortable.

Estate inventory and dependency mapping

A real discovery produces an inventory you did not have before: every workload in scope, its shape, its storage, its network position, what talks to what, and which dependencies will bite during a migration or replatforming. Dependency mapping is the part most often skipped and most often regretted. The database that three forgotten integrations rely on is always discovered eventually. The only question is whether it is discovered in week two on a whiteboard or in week fourteen at two in the morning.

The cost baseline you will audit later

Week two must also produce a cost baseline: what the estate costs today, normalised and documented, with the measurement method written down. This matters for every engagement and it is existential for optimization work. Our own optimization model charges a percentage of verified savings, with no fee if there are no savings, and that model only functions because the baseline is nailed down before anything changes. Across those engagements the average outcome is a 40 percent reduction in OCI spend, and every one of those numbers is defensible only because someone did unglamorous measurement work in week two. If your provider has not measured a baseline by the end of the second week, you have lost the ability to ever prove the engagement paid for itself, under any commercial model.

The risk register

Discovery should end with a living risk register: the things that could derail the plan, who owns each one, and what the mitigation is. A provider who reports no significant risks after two weeks inside your estate has either not looked or is not telling you what they found. Neither is acceptable.

Week 3: the plan made real

Week three converts discovery into a plan you can hold people to. For a migration that means a wave plan: which workloads move in which order, what the entry and exit criteria are for each wave, and where the rollback points sit. For a build or platform engagement it means a prioritised backlog with estimates. Either way, the plan should be specific enough that a delay is visible within a week, not a quarter.

Week three is also when the landing zone decisions get made and recorded: compartment structure, identity and federation, network topology, logging and security posture. These are the decisions that are expensive to reverse, and they deserve written decision records with the reasoning attached, because the person who made the call will not always be the person maintaining the result. Environments should be stood up in the same week, with the path from development to production defined before anything needs to travel it. This is the heart of disciplined OCI implementation work, and it is where specialist depth shows most clearly.

Finally, week three should ship at least one quick win. A rightsized set of compute shapes, an unattached storage cleanup, an alarm baseline where there was silence before. Quick wins are not the point of the engagement, but they fund trust, and trust is the currency that gets you through the harder weeks later. A provider who cannot find a single early improvement in a real estate is not looking very hard.

Week 4: first delivery and review

By the end of week four something real should exist: a workload running in production on OCI, a landing zone accepted against written criteria, or a signed design that engineering can build from without further debate. The artifact matters less than its nature. It must be finished, reviewable, and owned, not a status update describing progress towards a thing.

Week four also holds the first steering review, and the format of that review tells you a great deal. A good one walks through the plan against actuals, the risk register with movement since week two, the baseline with any early deltas, and a short list of decisions needed from you. A bad one is a slide deck of green traffic lights and no decisions. Course corrections agreed in the first steering review are cheap. The same corrections agreed in month four are change requests.

Early warning signs in the first month

The first 30 days are also your best diagnostic window, because problems that surface early are still cheap to act on. Five patterns deserve an immediate, direct conversation.

The A team vanished after kickoff. The principal architect ran the kickoff workshop brilliantly and has not been seen since. This is the single most common failure in consulting delivery, and the remedy is contractual: named key personnel, stated allocation, and your approval for substitutions. Enforce it in week one or accept that you never will.

No written outputs. Meetings happen, the mood is good, and nothing exists in writing. Verbal progress is not progress. Every week of a healthy engagement leaves artifacts behind: a governance sheet, an inventory, a baseline, a plan, decision records. If two weeks pass without a document you could hand to an auditor or a new team member, raise it formally.

Access stalled by the provider's own process. Some delay getting access is normal and is often on your side. But when the provider's own onboarding, security review, or internal approvals are the bottleneck for the access they requested, you are seeing their operational maturity directly, and you should expect the same friction at every later stage.

Scope creep before there is scope discipline. If new items are drifting into the work in month one without a written change process, the engagement has no boundary and the budget has no ceiling. This is most dangerous on hourly engagements, where every undocumented addition simply becomes more billable time.

No baseline measured. Worth repeating because it is the least visible and the most damaging. Without a week two baseline you cannot verify savings, demonstrate improvement, or hold anyone to an outcome. A provider who avoids measuring the starting point is avoiding accountability for the finish.

What good looks like

Strip away the methodology branding and a good first month reduces to three habits. First, written artifacts every single week, small, specific, and reviewable, so that progress is a fact rather than a feeling. Second, a cost and performance baseline you could hand to an auditor a year later, measured before anything changed, with the method documented. Third, quick wins delivered early, not because they are large, but because they prove the team can ship inside your environment, with your constraints, before the big commitments land.

Notice what is absent from that list: heroics. The first month of a good engagement is almost boring, and that is the highest compliment available. The drama lives in engagements where week one was skipped. If you are still choosing between providers, the ability to describe their own first 30 days in this level of detail, unprompted, is one of the sharpest evaluation questions you can ask, and we cover the rest of that evaluation in our guide to choosing an OCI consultancy.

How the commercial model shapes the first month

The shape of the first 30 days flexes with the commercial model, and a provider who runs the same generic kickoff regardless of model has not thought hard about either.

Under a fixed Project fee, the first month is where plan discipline pays for itself. The provider carries the overrun risk, so a serious firm invests heavily in week two discovery and week three planning, because their own margin depends on the estimates being right. From your side, the week four review is where you confirm that scope, acceptance criteria, and wave plan are all in writing, since the fixed fee is only as fixed as the scope behind it. We go deeper on this model in our piece on fixed price OCI migrations.

Under a Managed Monthly retainer, the first month is about the run book. By day 30 there should be documented procedures for the routine operations the retainer covers, monitoring and alerting wired into a 24/7/365 coverage model, and an escalation matrix tested at least once. A retainer without a run book is just a promise with a monthly invoice attached.

Under an Optimization fee charged as a percentage of verified savings, everything hinges on week two. The baseline is the contract. Both sides need it measured, agreed, and signed before changes begin, because the fee is calculated against it and the no savings, no fee promise is only meaningful if savings can actually be verified. With 20+ years of combined Oracle experience behind the analysis, the technical findings usually come quickly. The measurement discipline is what makes them bankable.

The 10 item kickoff checklist

Use this as the agenda for your own side of the first month. Every item should be done, in writing, by the day shown.

None of this requires a heavyweight methodology or a program office. It requires a provider who treats the first month as the foundation of the engagement rather than the victory lap of the sale, and a buyer who knows what to insist on. Hold the first 30 days to this standard and the remaining months largely take care of themselves. Let them drift and no amount of escalation later will buy back what the first month gave away.