Ask HN: How to prioritize on hundreds of issues?

3 points15DCFA8F7y ago1 comments

Hi!

I'm struggling with prioritization, and would like to know how people are tacking in, as I think it is a very common problem. Article and book recommendations are welcome (already read Time Management for Sysadmins).

I work on IT at a SMB company, and our issue control software has ~300 open issues. Issues are related to problems/adaptations on our ERP, developing/fixing problems on dozens of small systems we developed, help desk, systems administration, etc.

We already classify issues by priority:

Urgent/Important (currently ~5% issues) Normal (currently ~20% issues) Low (currently ~75% issues)

I currently split time as this:

- Morning - projects (bigger transitions like ERP upgrading, internal software development, studying and making internal processes better for other departments, etc). - After lunch - high priority issues. - Two hours after - low priority issues. - One hour before leaving - sysadmin stuff (maintenance).

This method works nicely, but... the problem arises when my brain must decide what is the next issue to tackle, when the time comes for low priority issues. Deciding what will be worked, and select on 200+ items with the same priority is very difficult, and I feel that I waste a lot of time deciding the next step versus acting.

Deciding on higher priority issues is easier.

Anyone can relate? Is there any tip, mental model changing that can help?

Thanks!

1 comments

cimmanom7y ago

Try adding a couple more dimensions, each rated on a scale of 1-3 or 1-5 or so:

A field for number of users (or $ in contracts) affected (for instance, does it affect every single user for every interaction; or only enterprise users with the Foo setting on "bar" when the moon is full?)

A field for severity/impact (Direct impact on cash flow? Data loss? Can't use the feature at all? Open up new markets? Decreased feature functionality? Significant bug with workaround? Minor bug?)

A field for centrality of the affected feature (whole purpose of the product? Important to most people's workflows? Peripheral feature that only adds minimal value?)

A field for complexity / level of effort to fix (rate this one on an inverse scale so that the highest numbers are the easiest to fix).

A field for urgency (is there an external deadline attached to this? A vague manager expectation? Is the expected completion time next week? Next year?)

For each ticket, multiply those numbers together, and you'll get a pretty decent and comparatively granular ROI ranking.

j / k navigate · click thread line to collapse