Why software deadlines are hard to predict

Jul 19, 2017
Kevin Monk

Tags: deadlinestime-managementsoftware-development

We have a rule that we don’t give deadlines for software changes. This may sound evasive and non-commital but we don’t believe it serves us or our clients to attempt the impossible. We’re not alone in thinking this way, many software companies have a similar no-deadlines policy.

Why’s it impossible? Because if it’s intellectually novel then you’ve never done it before. It is estimating the unknown. There are some things that are easy-ish to predict. We can predict the time to:

  • do the washing up;
  • build a wall;
  • walk to the station; or
  • make a cup of tea.

We find it harder to estimate the time to:

  • solve a cryptic crossword;
  • find all the items in a treasure hunt;
  • make a complex birthday cake; or
  • learn a new skill.

The key difference between the two groups is the familiarity and repetition of the subtasks. Anything novel will have pitfalls, barriers, and leaps in understanding that we couldn’t possibly have foreseen. Software development mostly falls into the latter group.

The good news is that whilst we can’t give deadlines we can do other things. We can give:

  • very approximate estimates;
  • deadlines for familiar tasks; and
  • the position of that development in our pipeline.

This is how we manage our workflow:

  • We use agile software development practices.
  • We meet every 2 weeks for a “Tech review”. In the Tech Review…
  • Naomi gives us a summary from the Customer Support Team of any recurring themes or problems from our doctors.
  • Rob feeds back from Account Management on our client’s priorities.
  • Between Naomi, Rob and the Tech team, we establish the priorities for the next 2 weeks.
  • We review any new feature requests in the last 2 weeks and ‘bumped’ requests that didn’t make it into the last review but have been mentioned again or an increased importance.
  • We apply a form of Pareto Analysis and the Eisenhower Method to decide priorities.

This no-deadline approach has a problem called Parkinson’s Law:

“work expands so as to fill the time available for its completion”

We mitigate Parkinson’s Law through internal deadlines, hiring hard-workers, and the intrinsic motivation to satisfy our clients needs and see SARD succeed commercially.

We hope that this clarifies why we’re unable to give timescales for the majority of our work. As always, I’m happy to discuss this directly: 07736 066 408.

Find this interesting?

You should sign up to our newsletter because it will keep you up to date with Medical HR, technology and what SARD JV are up to. We promise not to send you a load of sales bumf. Newsletters have a typical email open rate of around 15-20%. We aim to keep our open rate above 60%. That means we'll only add you to the list if you really want to be added and work hard to send you interesting stuff.

Subscribe to our mailing list

Comments