2

How many user stories does it roughly take before a product owner has a basic fluency in writing user stories? Meaning that the idea of writing user stories has moved beyond theory, rough execution, and the product owner naturally knows what to do when writing user stories.

Just to be clear, I'm not just asking how long it takes, but how many real stories it takes. Meaning to me, saying how long is not as meaningful as how many stories, though that might be flawed, and if so, then I guess the best answer would explain why, and what would be the best measure.

Also, my assumption is that the product owner would be engaging a team that already knows what they're doing.

| improve this question | | | | |
  • The question formulated this way is unanswerable. What does "gets it" or "move beyond theory" mean? How can you compare wanna-be PO to a veteran who went through a number of different projects? If you want a meaningful answer please ask a question which has one. – Pawel Brodzinski Mar 8 '12 at 13:15
  • I think @blunders wants to know how much does take to learn how to write usable user stories. – Zsolt Mar 8 '12 at 13:51
  • +1 @Zsolt: That's correct, though not just how long it tasks, but how many real stories it takes. Meaning to me, saying how long is not as meaning full as how many stories, though that might be flawed, and if so, then I guess the best answer would explain why, and what would be the best measure. – blunders Mar 8 '12 at 14:22
  • @Pawel Brodzinski: Thanks for the feedback, please see my comment to Zsolt above. If the question's intent is still not clear, let me know. If the intent is clear, but the question might be improved let me know. Again, thanks! – blunders Mar 8 '12 at 14:23
  • 1
    @blunders Pretty much that's what I mean. I also think your starter question was intriguing and now you have plenty clarifications to update it with. :) – jcmeloni Mar 8 '12 at 14:39
2

Different people are good at different things. There is no way to tell how long (or many whatevers) it takes for a specific person to learn something. It is no different with learning to write user stories.

If the PO has a hidden talent as an author or a blogger, he will catch up pretty fast. If he has a tendency to be very structured, he might never get it, because he might never grow above the "as a ... I can ... so that ..."

On top of that, learning something new not only takes practise, but also time. So if I answered: "It takes 1000 stories", it is of no use to lock him up in a cellar, to write the first 1000 stories. New things need time to sink in (a bit like: it takes one mother 9 months for 1 baby, how long do 9 mothers need?)

The 3. and I think the most important factor is real-life feedback. If the PO is constantly hanging out with the team, he will get direct and indirect feedback for his stories every standup. If this is handled equally well in retro your PO might write good stories for this particular project and team in 2 to 3 sprints.

So there's your answer: 3 sprints.

Question: how good are the feedback skills of your team and scrum master?

| improve this answer | | | | |
  • +1 @Huibert Gill: Thanks, though it's unclear where the three sprints answer magically came from; though I must say that I'd agree with that guess though have no real reason to do so. – blunders Mar 9 '12 at 21:00
  • @ Huibert Gill: Just noticed you ask "how good are the feedback skills of your team and scrum master?" - answer is there is no specific instance, though it's assumed everything is average; which includes that on average it's my experience that the team are on-board before the product owner is, meaning in my opinion SCRUM normally originates from from the tech team or scrum master. – blunders Mar 9 '12 at 21:06
  • 1
    I'm mostly conservative with estimates. So if it was 2-3 Sprints, I would opt for 3 sprints. But, what would live be without a little magic every now and then, hey? ;-) – Huibert Gill Mar 9 '12 at 21:06
  • @ Huibert Gill: Agree, just checking... :-) – blunders Mar 9 '12 at 21:07
3

I think the answer will be "it depends". There are a lot of variables to consider:

  • Has the product owner been trained to write user stories, or just given an example and thrown into the deep end?
  • Does your organization have clear standards/guidelines for user story content, or are you corporately just winging it?
  • Does the new product owner have someone available to help coach them?
  • How often will the product owner have to write user stories? Does (s)he have time to abosrb lessons learned and implement improvements?
  • How complex are the products that need user stories?
  • Does the product owner have the knowledge/aptitude to ask the team the right questions to get a useful user story?
| improve this answer | | | | |
  • +1 @Doug B: Yes, I know, but it's not even possible to give a best, worst and average case estimate? – blunders Mar 8 '12 at 17:02
  • @blunders: You can, but as David alludes to those estimates will depend on the quality of the underlying data and metrics used to support the estimates. Garbage in = Garbage out. – Doug B Mar 8 '12 at 18:11
  • Yes, it was never a question of if it's possible, and clearly any answer provided must be taken within the context that allowed it to be reached. I also believe it's possible there is an answer for the question without having to do the survey myself. – blunders Mar 8 '12 at 18:27
0

This is absolutely answerable and measurable. But it will take some serious effort on your part to gather the data. Build a survey instrument and querey those that do this activity. Control bias in your survey tool to the degree possible and ensure you have an adequate sample size. Ask questions that drive to when they perceived a degree of competence in performing this activity and when they perceived they hit "expert" level. I would also use focus-group discussions, as well, to supplement the incoming data.

Since it is a survey, where there is a ton of error, you cannot finish here. It'll likely give you a good starting point on the range of user stories and a rough idea of the frequency distribution. From here, you need to validate this by your own observations taking newbies and begin training and practicing. However, the survey results, if done well, are likely strong enough from which you can estimate and plan.

| improve this answer | | | | |
  • Yes, it was never a question of if it's possible, and clearly any answer provided must be taken within the context that allowed it to be reached. I also believe it's possible there is an answer for the question without having to do the survey myself. – blunders Mar 8 '12 at 18:27
  • You mean like if someone else did the research already? – David Espina Mar 8 '12 at 18:35
  • Correct, though it I'm not saying it would have to be the exact question, just research that would allow the question to be answered. For example, if there was research on the number of sprints it takes a product owner to write effective stories on their own in the real world for a given team size, I could give you a estimated story count based on other research; again, just estimates, but better than nothing. Which might also be to say he their was research on just the length of time alone, it might be possible. – blunders Mar 8 '12 at 19:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.