Since this summer we have been starting to work according to Scrum. In our Scrum implementation we need to set a time estimate and a market value on all items in the Product Backlog (an item is normally a new feature).

The time estimation is a fun process where the developers votes with the help of a deck of cards. The cards hold numbers from 0 to 100 in the steps 0, 1/2, 1, 2, 3, 5, 8, 13, 20, 40 and 100. They can also choose the card "?" or "Coffey break" :-).

If the team members votes more than 2 steps / cards from each other they need to take a discussion and do another vote. In the end, the team needs to agree on one number before they can move on. This number will then represent the amount of time the item will take do deliver. It is a strange way of doing estimations but it seems to work...

In the first sprint, time estimates was not so accurate but now, after a few sprints, estimations are getting better and better.

The market value, is a little harder to decide. We are today using a "score card" where a feature is evaluated against a set of conditions that have different value. The sum of the condition values gives the market value. But it is still more of a guess than a precise science.

How could we do this better? - By asking the users! And yesterday, for the first time in our history, we published a form where EPiServer editors could rank 20 suggested features. Within 24 hours we got 25 answers which was over my expectations. It will be fun to follow this during the next couple of weeks.

Are you curious to know the top three in the list right now? Here they are...

1. "Change sorting order using Drag & Drop" 

2. "List of broken links"

3. Improved Forms editor.

If  you want to vote yourself, visit http://www.episerver.com/en/Surveys/Help-us-prioritize-functionality/