Why Is New Software Always Late
Download as PPT, PDF0 likes406 views
This document discusses why software projects are often late and provides recommendations to avoid delays. It notes that over 70% of projects are late according to surveys. Late projects are essentially failed projects that cost customers. Common reasons for delays include unrealistic time estimates set before requirements are clear, scope creep, and not accounting for risks and challenges. The document recommends following best practices such as clearly defining requirements with customers, using appropriate development models, avoiding early firm schedules, ensuring good communication, and allowing time for risks and unexpected issues.
![Lateness in the news “ Anti-terror system is up to five years late” [Computer Weekly Aug, 2007] “ £7bn defence IT project running 18 months late” Defence Information Infrastructure, DII [Computing, Jan 2009] Vista – planned to ship late 2003, arrives Jan 2007! Criminal Justice Enhancement Program - $30 million over budget and running more than five years late [The Age, 2008]](https://image.slidesharecdn.com/Whyisnewsoftwarealwayslate-123630065882-phpapp02/85/Why-Is-New-Software-Always-Late-2-320.jpg)
![A recent survey says … Over 70% of failed projects are late! [M Miller, ConocoPhillips, 2008 ] This is a chronic issue Every developer / manager likely to have worked on a late project and probably will do so in the future!](https://image.slidesharecdn.com/Whyisnewsoftwarealwayslate-123630065882-phpapp02/85/Why-Is-New-Software-Always-Late-3-320.jpg)
![Failed projects? A late project is a failed project. [Lyyttinen & Hirschheim, 1987] Process failure in time & cost Costs to the customer? May as well be a train wreck! Train wreck at Montparnasse 1895](https://image.slidesharecdn.com/Whyisnewsoftwarealwayslate-123630065882-phpapp02/85/Why-Is-New-Software-Always-Late-5-320.jpg)
![What can be done? Plenty! Managers & executives hardly have to be fortune tellers to estimate correctly. Follow a “Best Practice” Code of Good Practice [BCS, 2004] ?](https://image.slidesharecdn.com/Whyisnewsoftwarealwayslate-123630065882-phpapp02/85/Why-Is-New-Software-Always-Late-9-320.jpg)
![Know the Customers needs Avoid making early firm schedules Only give a range of dates until requirements are clearer . “ Ensure that the scope, deliverables, timescales, costs and responsibilities are agreed in advance.” “ Explain fully the corporate objectives that underpin the requirement, the scope, issues, constraints and risks to be addressed. Offer constructive challenge to your customer if: The requirement is unrealistic Any of your customer's expectations are unreasonable” [BCS COP 3.1] .... And if the requirements aren’t clear? 1](https://image.slidesharecdn.com/Whyisnewsoftwarealwayslate-123630065882-phpapp02/85/Why-Is-New-Software-Always-Late-10-320.jpg)
![Don’t let external deadlines dictate Modify features to allow a go-ahead “ Review and agree with your customer any key external pressures and influences for business improvement, plans for organisational change, parallel programmes (with potential mutual dependencies) and the effect these may have on the programme.” [ BCS C.O.P. 3.1] 3](https://image.slidesharecdn.com/Whyisnewsoftwarealwayslate-123630065882-phpapp02/85/Why-Is-New-Software-Always-Late-12-320.jpg)
![“ Being Competitive” Shake the habit of giving attractive early-completion dates Avoid deliberate underestimating In the long run better, don’t have reputation as firm dealing in late software “ Do not overstate the capabilities, performance and benefits of the proposed products or services.” [BCS COP 3.2] 4](https://image.slidesharecdn.com/Whyisnewsoftwarealwayslate-123630065882-phpapp02/85/Why-Is-New-Software-Always-Late-13-320.jpg)
![Sources Failure definition : ‘Information systems failures - a survey and classification of the empirical literature ’, [Lyytinen & Hirschheim, 1987 In P.I. Zorkoczy (ed.), Oxford Surveys in Information Technology - Oxford University Press, Vol. 4, No. 1987, pp. 257-309.] Failure Statistics: New Insights into IT Project Failure & How to Avoid It [Miller, Dawson, 2008] http://www.mgmiller.co.uk/files/paper.pdf](https://image.slidesharecdn.com/Whyisnewsoftwarealwayslate-123630065882-phpapp02/85/Why-Is-New-Software-Always-Late-18-320.jpg)
![Sources Why is software always late? [ Connell] http://www.chc-3.com/talk/why_software_late.ppt BCS Code of Good Practice http://www.bcs.org/upload/pdf/cop.pdf Ethical, Legal and Professional Issues in Computing Chapter 10 , Duquenoy, Jones and Blundell Thomson Fasttrack, 2008.](https://image.slidesharecdn.com/Whyisnewsoftwarealwayslate-123630065882-phpapp02/85/Why-Is-New-Software-Always-Late-19-320.jpg)
Why Is New Software Always Late
- 1. A presentation by Alex Beston CS4524 Professional Topics
- 2. Lateness in the news “ Anti-terror system is up to five years late” [Computer Weekly Aug, 2007] “ £7bn defence IT project running 18 months late” Defence Information Infrastructure, DII [Computing, Jan 2009] Vista – planned to ship late 2003, arrives Jan 2007! Criminal Justice Enhancement Program - $30 million over budget and running more than five years late [The Age, 2008]
- 3. A recent survey says … Over 70% of failed projects are late! [M Miller, ConocoPhillips, 2008 ] This is a chronic issue Every developer / manager likely to have worked on a late project and probably will do so in the future!
- 4. But how late, is late ? One survey says 6 times longer than expected!
- 5. Failed projects? A late project is a failed project. [Lyyttinen & Hirschheim, 1987] Process failure in time & cost Costs to the customer? May as well be a train wreck! Train wreck at Montparnasse 1895
- 6. What is a failed project?
- 7. Why is this happening? We have managers, time management software tools, mobile phones, pagers etc surely we’re organised? Is it that people are absent from work? Off sick? Employing the wrong people? Brooks’s Law? Maybe the project is getting bigger – scope creep?
- 8. Unrealistic time estimates. Planning stage is faulty. 1) Just at the point least is known about the product, the crucial delivery date is made. 2) But the Customer is told the wrong time to delivery 3) Project starts, but soon delays .... 4) Cant pull out! They’re committed! Investment is already made, so obliged to wait.
- 9. What can be done? Plenty! Managers & executives hardly have to be fortune tellers to estimate correctly. Follow a “Best Practice” Code of Good Practice [BCS, 2004] ?
- 10. Know the Customers needs Avoid making early firm schedules Only give a range of dates until requirements are clearer . “ Ensure that the scope, deliverables, timescales, costs and responsibilities are agreed in advance.” “ Explain fully the corporate objectives that underpin the requirement, the scope, issues, constraints and risks to be addressed. Offer constructive challenge to your customer if: The requirement is unrealistic Any of your customer's expectations are unreasonable” [BCS COP 3.1] .... And if the requirements aren’t clear? 1
- 11. Use the right model for the customer How can you estimate the delivery date .... .... If the customer doesn’t know everything they want ? .... Or the requirements take so long to deliver ... .... The business has changed before product is delivered! In this case evolutionary or adaptive models. (or bail out? Get a better analysis team?) 2
- 12. Don’t let external deadlines dictate Modify features to allow a go-ahead “ Review and agree with your customer any key external pressures and influences for business improvement, plans for organisational change, parallel programmes (with potential mutual dependencies) and the effect these may have on the programme.” [ BCS C.O.P. 3.1] 3
- 13. “ Being Competitive” Shake the habit of giving attractive early-completion dates Avoid deliberate underestimating In the long run better, don’t have reputation as firm dealing in late software “ Do not overstate the capabilities, performance and benefits of the proposed products or services.” [BCS COP 3.2] 4
- 14. Better Team Communication Bad scenario: The planner has little idea what is involved and fails to consult the implementers, stands aloft, doesn’t consult The planner has an unrealistic expectation of the implementers i.e. 100% availability, 8 hr days? The planner should be finding out how long tasks take and each team members personal capacities. 5
- 15. Assuming all will go well... If you have 20 tasks, each with 90% of success ... What is the probability it will all go well? 12%! 6 ( .9 20 )
- 16. Cant make assumptions? 70% chance on 200 tasks? 0. 0000000000000000000000000000001046% chance of success. In other words can assume it wont go to plan! Allow time for things going wrong! Can only assume there will be some parts going to plan and some that don’t. Survey of the team for unfamiliar territory
- 17. In Summary ... A Project has a strong tendency to be delayed Always follow best practice Only schedule when all the information is available Good communication with the customer and the team
- 18. Sources Failure definition : ‘Information systems failures - a survey and classification of the empirical literature ’, [Lyytinen & Hirschheim, 1987 In P.I. Zorkoczy (ed.), Oxford Surveys in Information Technology - Oxford University Press, Vol. 4, No. 1987, pp. 257-309.] Failure Statistics: New Insights into IT Project Failure & How to Avoid It [Miller, Dawson, 2008] http://www.mgmiller.co.uk/files/paper.pdf
- 19. Sources Why is software always late? [ Connell] http://www.chc-3.com/talk/why_software_late.ppt BCS Code of Good Practice http://www.bcs.org/upload/pdf/cop.pdf Ethical, Legal and Professional Issues in Computing Chapter 10 , Duquenoy, Jones and Blundell Thomson Fasttrack, 2008.