How to deal with missed deadlines?

New / EngUfpO77
May 7 15 Comments

I’m building the prototype for a startup. We have missed two deadlines I gave and we keep learning more aspects of the problem that needs to be solved.

How do you deal with project estimates and missing deadlines? How do you communicate the realities of software development to non technical stakeholders?

comments

Want to comment? LOG IN or SIGN UP
TOP 15 Comments
  • New bmmC14
    estimate * 2
    May 71
    • New bzRF81
      iirc it's 2x+300 hours
      May 7
  • Microsoft rand(NULL)
    You need to deliver a workable prototype super fast. Cut absolutely everything which can be cut. Forget about code quality, quality engineering, etc. Fuck it. Just make the prototype work first. If you raise the money or start earning the money then you will deal with quality and non-vital features. Just put yourself in a box: what can you deliver over the weekend, in 1 week, 2 weeks? Like a hackaton. Like your life depends on it. Make the timeframe drives you, not otherwise.
    May 70
  • New / MgmtMcAfee
    You need to track your hours.
    Get a baseline.
    Then track hard projects and easy ones.
    Map those sized projects to a Fibonacci number.
    Then create a roadmap.
    That roadmap should be broken into sprints.
    Estimate by points 15-20 max per person, per week, and create a bi-weekly sprint.

    Have frequent checkpoints during the sprint to see if you are tracking your targeted velocity. If not, adjust. Then update the roadmap.

    After some time, you will figure out how long things take.

    NOTE: Break the entire project down into tiny little granular deliverables. Add EVERYTHING!!!! Even time for meetings and reviews. Then double that time.

    Over time, after a few months. You can nail your estimates better.

    Max your Fibonacci numbers out at 13, for a large and 1 for a tiny.

    PM me if you need more help.
    May 71
    • New / MgmtMcAfee
      Get this road map to your smaller team first. Work out timelines, and adjust. Then, ONLY, escalate if needed. You will find that everyone will work with you after this and you can gain a lot of respect and support with timing. They will often cut things out of scope for you!
      May 7
  • Facebook ImranKhan
    For a big and complex enough project there is no way of providing good estimates. So called engineering leaders I have seen have cracked this problem by giving aggressive deadlines and handing over the actual work to minions who are there to support him no matter what in the hope of a good performance review by him. When the deadline is missed, there are two approaches taken. 1) the suave way is to point out things beyond control and call it with some fancy names to confuse the usually illiterate management 2) the brute force but sometimes naive way is to throw some of the selected trouble making minions under the bus for not pulling their weight. This can work magically if those trouble makers were questioning the leaders and there is a justifiable motive that can be attributed to them not making the deadlines.
    Anyway, the best approach is to provide an estimate with ifs and buts to make it vague so it can be made whatever one wants at the end.
    May 70
  • IBM / Engqpwoeirutd
    Go rogue, abandon the concept of deadlines. Have Target dates, but not deadlines. Build in the expectation that you will almost always go past your target date. Focus on the engineering, make sure everyone is pulling their weight and no on is slacking off. <--- if you can do this, you can deliver better updates faster in the future.

    Deadlines are artificial pieces of trash. Get your clients to understand that you focus on quality over completion time. Ever wonder why there seem to be less bugs in Apple products than anything else? This is because of their development philosophy. They don't release anything that isn't fully and throughly tested internally. As a result we all have to wait longer for Apple updates than others.
    May 74
    • Facebook / Engkjdthcj
      Lol have you ever delivered anything with that attitude? Or maybe you deliver one item a year and just say you focus on quality
      May 7
    • IBM / Engqpwoeirutd
      No, because I'm not in the position to make those decisions. But I regret every single time I have to ship shitty code because someone higher up decided that we all like giving up our nights and weekends to meet some arbitrary deadline.

      I've come to have the attitude that deadlines are all just target dates, and I don't fret too much if we have to delay a release by a couple of days or even a week because the reality of coding time didn't align with someone's napkin math.
      May 7
    • Facebook / Engkjdthcj
      Makes sense
      May 7
    • IBM / Engqpwoeirutd
      PS every time I was able to focus on quality instead of deadlines: the code that was written during that time never seems to need to be revisited / refactored.

      When you have the "just ship and fix later" attitude all you're doing is creating technical debt. Nobody ever wants to clean up their technical debt.
      May 7
  • New / ITbluerail
    Always prepare the worst case.
    May 70
  • Google gliz
    Tell them there are risks upfront.
    May 70
  • Apple etubrute
    Steve McConnells cone of uncertainty.
    May 70
  • Oscar 🐨koala
    Deliver small pieces. Like someone else said call the leadership and have an honest talk. Don’t end up like accenture vs avis.
    May 70

Join verified employees in our anonymous social network!Download the app!

close