It depends on how much time is left before the deadline and how long optimization or customization will take. We can’t exactly say how much time optimization needs, but we can set a time limit for it. If we can finish optimization in that time, it’s good, and we can continue with the project. If not, we should stop optimizing and stick to the old algorithm to meet the deadline. Both ways are fine. After the project is completed, we can still optimize later. It might be hard, but the service and quality of your product are very important too. If we can request more time for the deadline, then the main focus should be on optimization.

Balancing algorithm optimization with project deadlines is always a challenge, but it’s crucial to strike the right balance between efficiency and timeliness. Here’s how I approach it: 1. Prioritize High-Impact Optimization: I focus on changes that deliver the most significant performance gains, ensuring the effort invested aligns with the project’s goals. 2. Set Realistic Milestones: Breaking the project into phases allows iterative testing and refinement, ensuring steady progress. 3. Transparent Stakeholder Communication: I keep stakeholders informed about the trade-offs between achieving perfection and meeting deadlines. This approach ensures that efficiency is not sacrificed while meeting project timelines.

- Profile first - measure before optimizing to identify actual bottlenecks - Apply the 80/20 rule - focus on optimizations with highest impact/effort ratio - Set performance thresholds - define acceptable minimums for critical metrics - Consider maintenance costs - complex optimizations can increase technical debt - Document optimization opportunities - track potential improvements for future iterations The goal isn't perfect efficiency, but rather meeting project requirements while maintaining reasonable performance. When deadlines loom, prioritize working solutions that meet specifications, then optimize incrementally in subsequent releases if needed.

What I have seen from my experience is when working on complex project with tight timelines it’s all about prioritization and planning. Deciding algorithm or approach is initial step. Based on it we will decide project timelines, which should include testing, deploying, resources etc. If estimate for optimised algorithm is taking long time then deliver the project in phases which is always a good idea. First phase can have deliverable that are acceptable in current scenario, cause without matching current trend you can’t sell your product. Then keep the more complex part or further optimization for upcoming phase. Convey the deliverables to the stakeholders, so that they know and agree with what they will get and when.