You have just been handed a 10% cut in the engineering budget. How would you go about delivering on your committed roadmap?

Budget cut would mean, lesser engineering bandwidth and hence we need to cut down our roadmap to focus on the topmost priroity items. Hence, I would follow the belwo steps -

• Reassess and reprioritise my roadmap based on available bandwidth
• Stakeholder alignment on the revised scopa and what that means in terms of product features to be dropped
• I would also try to minimise the scope for each of the prioritised features to ensure atleast for the MVP we are just picking up the must have user stories
• I would work with the tech team to also readjust the reserved bandwidth for tech debt/ performance tuning initiatives
• Finally, I would make sure to have clear expectation setting with the leadership team based on the revised roadmap and thereby revised targets

Clarifying questions
1. is the 10% across the board ie. FTE and non-FTE? across the board
2. is it a global cut across your portfolio or specific to a particular vertical? global
3. Is the expense challenge applicable to the remainder of the fiscal year or some other timeframe (assume for the next 12 months)

Restate the problem statement
You have a team of engineers delivering on a portfolio of activities with a committed roadmap and are required to cut your budget (FTE and non-FTE) by 10%. What is your approach to do so?

Let's itemize the key drivers of savings:
1. attrition of FTE and delayed backfill or backfill with more junior FTE
2. GL analysis and accounting clean up to ensure no erroneous outflows for things like compute and hardware
3. rationalization of efforts and removal of duplicative work
4. sourcing global and change budgets where resources can be billed out to work on LOB specific outcomes that are already aligned to their roadmap (ie. delivering deployment capability for new pattern could be POC on a particular use case that has funding)
5. reduction of outflows: that being paid to central tech and eng service teams, reducing headcount of support teams and removing superfluous tasks from their mandate
6. capitalization of hardware costs
7. reduce vendor footprint where able such as removing certain sandbox services that would be considered non-core

My next step would be to prioritize by t-shirt sizing the impact in order of least invasive and easiest to absorb as well as quickest to actualize - L to highest impact on operations and process heavy to implement H. It may look something like this:

1. L
2. L
3. M
4. H
5. M
6. M
7. H

I would then start to forecast out expected savings starting at L until I reach the desired cut.

You could also list out the potential mitigants to any service level impact for each of these potential actions. In a large enough unit absorption should be easy enough and in most any unit 10% should be fairly absorbable unless it is not the first cut delivered in the timeframe. (This could have been a clarifying question but there are still mitigants in that case, you'd just add in which drivers are no longer available to action)