Query Cost ~ USD $$$

  • All,

    I have done lot of fine tuning work in sql2005. I just consolidated it in a spreadsheet and sent it to my manager. He suddenly asked one question. 'How much USD Dollar amount you saved ?'

    I don't know the answer for this. can anyvody know the answer for this question?

    is it possible to calculate how much $$$ amount saved after this tuning work? if it is a basic question, i think i am an infant for this question.

    karthik

  • Nobody can give a sensible answer to this question, but you.

    I could say "all the Euros you saved * 0.77", but I'm sure it won't help. 🙂

    You have to understand what business processes are involved in the procedures you optimized and estimate how time saving translates into money saving. For instance, last year we changed our warehouse barcode readers management system, that used to stop working for several minutes every week. The operators had to stop working during downtimes and those stops costed 100K Euros / Year. That was the money we saved.

    -- Gianluca Sartori

  • Whatever figure you decide on, remember to send 80% of it to the people on here that have helped you over the last year 😀

    More seriously, try to think if anything you have done has put off the need for bigger hardware, or the costs that would have been incurred if you had not done whatever it is you did.

    Have you recently asked for a pay rise or something? This is an old managerial-type trick...don't get your hopes up.

    Paul

  • Paul White NZ (5/25/2010)


    Whatever figure you decide on, remember to send 80% of it to the people on here that have helped you over the last year 😀

    Only the last one year?

    Gail Shaw
    Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
    SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability

    We walk in the dark places no others will enter
    We stand on the bridge and no one may pass
  • Paul White NZ (5/25/2010)


    Whatever figure you decide on, remember to send 80% of it to the people on here that have helped you over the last year 😀

    More seriously, try to think if anything you have done has put off the need for bigger hardware, or the costs that would have been incurred if you had not done whatever it is you did.

    Have you recently asked for a pay rise or something? This is an old managerial-type trick...don't get your hopes up.

    Paul

    HAHA...HAHA...:-D

    Good Deal 🙂

    karthik

  • GilaMonster (5/25/2010)


    Paul White NZ (5/25/2010)


    Whatever figure you decide on, remember to send 80% of it to the people on here that have helped you over the last year 😀

    Only the last one year?

    Look at it as a down-payment 🙂

  • I'm confused....

    If you can't explain the benefits of the fine tuning we you did, then why there was a need for tuning in the first place? Tuning for the sake of tuning doesn't pay off...

    Example:

    If you fine tune an app rarely used, not business crucial but on a high end hardware (solely purchased for that app), what's the benefit of e.g. cutting down the time for a nightly job from 10min to 2min ?



    Lutz
    A pessimist is an optimist with experience.

    How to get fast answers to your question[/url]
    How to post performance related questions[/url]
    Links for Tally Table [/url] , Cross Tabs [/url] and Dynamic Cross Tabs [/url], Delimited Split Function[/url]

  • lmu92 (5/25/2010)


    I'm confused....

    If you can't explain the benefits of the fine tuning we you did, then why there was a need for tuning in the first place? Tuning for the sake of tuning doesn't pay off...

    Example:

    If you fine tune an app rarely used, not business crucial but on a high end hardware (solely purchased for that app), what's the benefit of e.g. cutting down the time for a nightly job from 10min to 2min ?

    i think this is one of those things that a lot of people get caught up in to. i've had *so* many arguments with some of the people i work with with regards to this.

    as far as i'm concerned, before deciding whether or not to address a problem, you need to first determine the seriousness of the problem, and then the time it would take to address it. you'd then rank your problems and solve them in order.

    but, other people have other opinions, and some times what will happen is that a large amount of time will be invested into solving a seemingly minor problem. i'm guilty of this too - especially when it comes to "cool" things, like solving interesting SQL problems. i'll spend way more time than necessary, just to optimize a query which would run once a month from taking an hour to taking a minute.

    was it a good solution? sure, making an intensive process take less time is of course good. but is it worth the time invested? well, chances are, no, it isn't. but it's easy to get caught spending the time anyway.

  • A lot of managers live and die by the motto: If you can't measure it, you can't manage it. I can see some value in this, but it can also be taken too far.

    For one thing, despite the seeming concreteness of the term "measurement", there can be plenty of shades of gray in anything you measure. I can look out my window and decide that today is much sunnier than yesterday; I can make (manage) many decisions based on this measurement (put on sunblock, take sunglasses, leave the umbrella at home). However, I don't know the exact number of lumens reaching my eye; I can take a measurement (it's sunny today), but I can't drill it down to a granular quanta (number of lumens). So, unfortunately, too many managers won;t accept "sunny" as a valid measurement (despite its efficacy) and send you back to quantify it further.

    In your case, it seems you can measure the benefit of the work you performed (the app works faster; perhaps you've logged execution time, CPU load, etc.), but not in the units and granularity your manager would like. However, you may be able to translate from one to the other.

    For example, let's say you worked on a data entry app and improved the performance by 10% (by whatever measure you're using). That means that the data entry personnel can either: enter 10% more data in any particular time frame or perform the daily "quota" of entry in 10% less time. So, the company can benefit from either a 10% increase in production or a 10% decrease in data entry staff-hours. Either way, you could convert this 10% into USD using the salaries of the Data Entry team as the basis: if your company is paying the DE folks 100K/year, you saved the company 10K. If the app revolves around entering, say, subscriptions, and each subscription brings in a certain number of dollars, you could also measure the value of the tuning as 10% of the daily/monthly/yearly income from those subscriptions.

    This may not be truly accurate (e.g. your company may never see that extra 10K appear on the books), but this is probably the type of calculation your manager is looking for.

  • I use to have to calculate ROI(s) for a lot of IT projects (and previous to that, Engineering projects).... it is not an easy task.

    As stated before, what is your measurement/basis?

    - Money saved in terms of CPU time?

    - Productivity improvements (Staffing/Employees/Customers) in terms of getting a job done quicker due to a query running faster? (I like these kind of ROIs, but they do not produce as much $ as you might think).

    - Some other task that now performs faster/better?

    How do you quantify the $ for the people that use to complain about a report taking 2 minutes or more but now runs in 20 seconds or less? Guess part of that might go into "Tech Support savings"?

    Here is a good article that might help in trying to figure out ROI/"how much money did you save":

    http://articles.techrepublic.com.com/5100-10878_11-5034665.html

    Plus another article:

    http://blog.smartdraw.com/archive/2009/03/16/how-to-calculate-roi.aspx

    Good luck! 🙂

    Argue for your limitations, and sure enough they're yours (Richard Bach, Illusions)

Viewing 10 posts - 1 through 9 (of 9 total)

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