Estimates

  • Comments posted to this topic are about the item Estimates

  • Deadlines can be achieved even on a complex project if:

    1 estimates are created by people who have experience in the field.

    2 estimates are peer-reviewed honestly and thoroughly.

    3 estimators are insulated from customer stated timescales.

    4 if the total estimate is too big, the deliverables or the resources should be tweaked: never the estimates.

    5 each estimator's estimates are tracked against delivery to establish their over/under estimate record.

    6 estimator's under/over estimate track record is used to validate their estimates.

    7 continued, regular, honest estimates are encouraged as the project progresses.

    8 proper project management is applied to tweak features/quality and resources to correct any divergence as soon as it becomes apparent.

    and finally:

    9 reward those whose estimates are proved to be accurate.

    Points (5) and (6) are important: each person is prone to being optimistic or pessimistic, and both traits need to be corrected within reason before an estimate is adopted.

    Points (7) and (8) ensure that estimates are updated and that corrective action is applied. "Head in the sand" isn't a good way to meet deadlines.

  • Hey Steve, Regarding educating people in the community ... Have you ever thought about hosting a weekly/bi-weekly/monthly webinar? It could be 30-60 minutes long on a variety of SQL Server topics with time for Q&A at the end.

    Just a thought!

  • I hate estimating because of the continual changing of scope and the lack of time set aside for proper requirements analysis. It may be different in some DBA worlds but certainly application or systems development suffers from inappropriate level of contingency based on the flexibility of requirements. This is where Agile steps up until it gets abused by people using it as an excuse for doing everything off the cuff, undocumented and lacking best practices.

    Gaz

    -- Stop your grinnin' and drop your linen...they're everywhere!!!

  • I have thought about it. Tough to make the time on a regular basis, but I wouldn't mind it. Mostly we'd like to try and get good interaction with people, maybe with something like Twitter. Maybe we'll get it done at some point.

  • david.wright-948385 (9/9/2014)


    Deadlines can be achieved even on a complex project if:

    ...

    4 if the total estimate is too big, the deliverables or the resources should be tweaked: never the estimates.

    You ignore #4, which is a big one.

    I do agree estimates can be better, but far, far too many people have little experience and their estimates are not tracked and feedback given. I'd argue that most of this is because mistakes, failures, bad choices are hard to deal with and we don't want to be punished.

    This is one of the things that lean enterprises are trying to get around with lots of metrics, transparency, and continuous delivery.

  • Steve, I'd say you hit it just right. If estimates were used by management as a real tool, managed and monitored them to make improvements, we would all be using them - and not as some form of punishment. But I've seen many clients who look at estimates in the same way we look at a car repair estimate. Not as an estimate, but as an exact amount. In the end, much comes back to either good management - or less than effective management.

    The more you are prepared, the less you need it.

  • Steve Jones - SSC Editor (9/9/2014)


    david.wright-948385 (9/9/2014)


    Deadlines can be achieved even on a complex project if:

    ...

    4 if the total estimate is too big, the deliverables or the resources should be tweaked: never the estimates.

    You ignore #4, which is a big one.

    I do agree estimates can be better, but far, far too many people have little experience and their estimates are not tracked and feedback given. I'd argue that most of this is because mistakes, failures, bad choices are hard to deal with and we don't want to be punished.

    This is one of the things that lean enterprises are trying to get around with lots of metrics, transparency, and continuous delivery.

    You only need enough accuracy to know whether it makes business sense (which could be a combination of qualitative and quantitative measures) to implement, and to what to standard/specification to implement to. Anything much past that level of accuracy will cost ungodly amounts of effort and cash, and often enough will sink perfectly reasonable and valuable deliverables.

    I'd be careful with rewarding based on "who's the most accurate". Most of the time you provide benefit by implementing something that provides the business outcome they want and pushing away those outcomes that can't be achieved, NOT by predicting the cost of a 2-year project to the penny.

    ----------------------------------------------------------------------------------
    Your lack of planning does not constitute an emergency on my part...unless you're my manager...or a director and above...or a really loud-spoken end-user..All right - what was my emergency again?

  • crussell-931424 (9/9/2014)


    I hate estimates because they tend to be how long it would take to do something and we never seem able to dedicate that time to the project. There always seems to be "hot" items that need to be done now taking time away from the project. Every one of those pushes back the delivery of the project.

    There are two ways around that: get a decent project manager that will either allow for background/fire fighting activity or provide separate resources to cover it, or allow for it yourself in your estimates 🙂

  • david.wright-948385 (9/10/2014)


    crussell-931424 (9/9/2014)


    I hate estimates because they tend to be how long it would take to do something and we never seem able to dedicate that time to the project. There always seems to be "hot" items that need to be done now taking time away from the project. Every one of those pushes back the delivery of the project.

    There are two ways around that: get a decent project manager that will either allow for background/fire fighting activity or provide separate resources to cover it, or allow for it yourself in your estimates 🙂

    Sometimes it is the cost of the context switching that is difficult to include in estimates and isn't covered when you report on how long you worked on the "distraction".

    Gaz

    -- Stop your grinnin' and drop your linen...they're everywhere!!!

  • david.wright-948385 (9/9/2014)


    Deadlines can be achieved even on a complex project if:

    1 estimates are created by people who have experience in the field.

    2 estimates are peer-reviewed honestly and thoroughly.

    3 estimators are insulated from customer stated timescales.

    4 if the total estimate is too big, the deliverables or the resources should be tweaked: never the estimates.

    5 each estimator's estimates are tracked against delivery to establish their over/under estimate record.

    6 estimator's under/over estimate track record is used to validate their estimates.

    7 continued, regular, honest estimates are encouraged as the project progresses.

    8 proper project management is applied to tweak features/quality and resources to correct any divergence as soon as it becomes apparent.

    and finally:

    9 reward those whose estimates are proved to be accurate.

    Points (5) and (6) are important: each person is prone to being optimistic or pessimistic, and both traits need to be corrected within reason before an estimate is adopted.

    Points (7) and (8) ensure that estimates are updated and that corrective action is applied. "Head in the sand" isn't a good way to meet deadlines.

    I totally agree. I believe that makes approximately two of us on the entire planet based on my experience in the workplace.

    Another one I would add is, "Do not be afraid to drop the project if the honest estimate suggests it's not worth doing. Find something worth doing and do that instead". This is one people fall down on a hell of a lot - Pet Project Syndrome.

    I'm a DBA.
    I'm not paid to solve problems. I'm paid to prevent them.

Viewing 11 posts - 1 through 10 (of 10 total)

You must be logged in to reply to this topic. Login to reply