Executive loops for a VP of Engineering use the 30-60-90 day plan in different ways, and none should be assumed. Sometimes it is a written take-home sent before the final round. Sometimes it is a short presentation to the CEO or a panel. Sometimes it is a verbal question in a conversation. Many companies do not ask for one at all. The content below works for all three versions; only the length changes. If you are given the format, ask how long you have and who will be in the room.
Assumptions to state before the plan
A plan without context reads as a template, so state assumptions and invite correction. The worked plan below assumes an organisation of about eighty engineers in nine teams. Roughly the last three quarterly commitments were missed. The CEO's framing is that "engineering cannot hit dates", and a director role is open. All figures are illustrative.
Then name the hypotheses, because "missing deadlines" is a symptom with several possible causes:
- Dates are committed to customers before engineering has estimated the work.
- Too much work is in flight, so everything moves slowly and nothing finishes.
- Unplanned work, such as incidents, escalations and support, absorbs capacity nobody planned for.
- Teams depend on each other for most deliveries, and the handoffs cause the delay.
- Priorities change mid-quarter, so finished work is not the work that was promised.
- Skills, staffing or management gaps.
Saying you do not yet know which of these is true is the strongest opening available. It tells the panel you will diagnose before prescribing.
The worked 30-60-90 day plan
| Phase | Focus | Main actions | Outcome by the end of the phase |
|---|
| Days 1-30 | Listen and diagnose | One-to-ones with every manager and staff engineer; skip-levels; peers in product, sales, finance and customer success; a post-mortem on the last three missed commitments; baseline delivery data | A written diagnosis naming the top causes with evidence, shared with the CEO; one honest list of every current commitment |
| Days 31-60 | Change how commitments are made; first win | No external date without a team estimate and a range; cut work in flight; contain unplanned work; re-baseline current commitments; start hiring for the open director | One re-baselined milestone landed on its new date; fewer projects in flight; every commitment has one owner and a range |
| Days 61-90 | Install the system | Run the next quarter's planning on the new process; agree capacity trade-offs with product and sales; monthly delivery report to the executive team; draft a twelve-month engineering strategy | An agreed quarterly plan with explicit trade-offs; a predictability measure tracked and reported; leadership gaps filled or hires under way |
The phases overlap in practice; the checkpoints tell the CEO what to expect and when.
Days 1-30: listen with a purpose
Listening is not the outcome; the diagnosis is. Ask every manager the same few questions so the answers are comparable. Which commitment slipped last? When did they first know it would? Who did they tell? What did they stop to recover it?
The most informative exercise is a post-mortem on the last three missed commitments. For each one, establish who set the date and on what basis. Compare scope at commitment with scope at delivery, and find where the time went. Pull a baseline from the work tracker: how many projects are in flight per team, how long work takes from start to finish, and what share of each sprint was unplanned. Rough numbers are enough. You are looking for the pattern, not a dashboard.
Ask your peers early what "on time" means to them. Sales may mean the date given to a customer; product may mean the quarter. Different definitions of a deadline are a common, quiet cause of "missed" ones.
By day 30, write the diagnosis down, keep it to two pages, and walk the CEO through it. A plausible illustrative finding: "Most missed dates were committed before engineering saw the work. About a third of capacity went to unplanned support. Fourteen projects were in flight across nine teams." Alongside it, publish one list of current commitments with honest status, even when the news is bad. That list is your first deliverable, and often the first time the leadership team has seen everything in one place.
What a new VP of Engineering does not do in month one matters as much. You do not reorganise, change the methodology, replace managers, or promise the CEO a date. Each of those needs a diagnosis you do not yet have, and a new leader's early mistakes are hard to reverse.
Days 31-60: change the commitment, then keep one
Act on the diagnosis you wrote, not on a generic playbook. Working from the illustrative finding, change three things.
First, change how dates are made. No date goes to a customer without a team estimate, and dates are given as ranges with a confidence level. Agree this with the sales and product leaders as peers. Do not impose it as an engineering rule, because they own the customer conversation.
Second, cut work in flight. Going from fourteen projects to eight, by pausing six with the product leader's agreement, will usually do more for predictability than any process change. Say which six and why.
Third, contain unplanned work, with an interrupt rota per team and a single intake for support escalations, so that the third of capacity becomes visible and planned.
Then re-baseline the current commitments and land one of them on its new date. One kept promise does more for engineering's credibility than any presentation.
Use this phase for the people decisions the diagnosis supports. Start hiring for the open director role. If a manager is clearly struggling, begin the support and conversations now, because you cannot resolve it by day 90 if you start at day 80.
Days 61-90: make it the way the company plans
Run the next quarter's planning on the new process. Show capacity per team, the floor held for operational work, and the commitments that fit. Then show the ones that do not fit, and ask product and the CEO to choose. That conversation is the main outcome of this phase: trade-offs made in planning rather than discovered in delivery.
Start a short monthly delivery report to the executive team. It should show the share of committed milestones delivered within their range, and the trend in how long work takes. Report misses alongside hits. A measure that only ever shows good news will not be believed.
Finally, draft a twelve-month engineering strategy and present it for challenge.
Activities are not outcomes
The commonest weak 30-60-90 day plan is a calendar of activities: meet stakeholders, learn the architecture, build relationships, identify quick wins. None of it is wrong. None of it tells a CEO what will be different on day 90, or lets anyone check. The strong plan names an outcome at each checkpoint: a written diagnosis, a kept date, an agreed plan with trade-offs.
The opposite failure is arriving with the answer, such as "introduce Scrum" or "set OKRs", before diagnosing anything. That tells the panel you will apply your last company's fix to this one. The best candidates hold the hypotheses openly, are willing to find that the cause is upstream of engineering, and still promise a visible result inside sixty days.
A 30-60-90 day plan is judged by what will be true at each checkpoint: a diagnosis with evidence at 30, a kept promise at 60, a planning system at 90. Name the outcomes, not the meetings.
© 2026 Preptima. Originally published at preptima.com (opens in a new tab).