Agile Estimation Reference Story Technique with Examples

hi there this is Ajith here I'm going to talk about the role of reference story in agile estimation what is the reference story a reference story is one of the simple story in the product backlog which is estimated very first time by the scrum team and assigned a story point so that this story can be used to estimate rest of the stories in the backlog this session is particularly useful for those who are coming from traditional mindset of absolute estimation which is where they had been estimating the software in terms of hours days months and now been asked to do relative estimation or story point estimation as part of agile implementation before we run into the reference story and the estimation process around it let's understand what is a story point a story point is a number which is assigned to a story to estimate it and these numbers are taken from the Fibonacci series because stories have to be estimated relatively as you can see in the Fibonacci series which are written down here a number is sum of the previous two number so that way these numbers are forming some relationship between them and this relationship is very important when it comes to estimating their stories now the question comes what are the criterias based on which a story is assigned a higher story point or a lower story point there are three criteria is complicity of the logic and architecture any gaps risk issues open questions more open questions more uncertainty story is going to attract higher story point amount of work or sides number of text boxes drop-down checkboxes calls to a pH is good is going to decide its side put together decides the story point of the story but remember this is always a trade-off between complexity and the amount of work they may be a story which is having just three text boxes but then the logic it is processing is very complex in that case that story is going to attract a higher story point alternatively they may be a story which is I mean say ten text boxes five dropdowns couple of calls to the epa's but then I know if it is using it's very simple in that case this story is going to protect a lower story point comparing to that story let's get to the reference story now in the ideal world a simple stories picked up as a reference story and is assigned a small number from the Fibonacci series that number normally is three and then we ask scrum team to start comparing rest of the stories in the backlog with the reference story based on these three criterias and see if those stories attract a higher story point or an old story point and then continue destination journey it looks very straightforward very smooth but in the real world things are not that easy people literally freak out especially those coming from the traditional mindset of estimation they do struggle even after a couple of Sprint's they are not able to settle down so they are always looking at the scrum master and agile coach for the help for situations like this such a technique which can help them easily transition from their hourly mindset of estimation to the story point estimation and I'm going to talk about such technique by given example here this is a simple story which can be taken up as a reference story it reads given a new bank customer when opens an account then details are entered in the system now we can visualize this story as a empty form like this which is having just my first name customer last name address date of birth is mobile and few other fields now this story will be worked out by a cross-functional team which consists of a UI developer Java server-side developer a DBA and a tester at a scrum master I'm gonna watch this team what do the thing is the end to an estimate of this story so that this story can be delivered in the spin meaning everyone in the team has to tell the total efforts that the story will take which includes the efforts of the development testing DBA work or anything else now UI developer may say that the end-to-end estimate of this story is 30 hours javis a beside me say 35 DBA was not very close enough to the end to and functionality may give some number less or more say his top giving 25 words testers giving 30 of us now the close enough estimate which can be agreed by the team is 30 hours next I'm going to do is draw this table which has our snap to the story points 20 hours to two story 532 350 to 582 8 these numbers comes from based on our prior experience of the projects and something that the agile cou might have been funded so it all depends from project to project set up a set up organization to organization I pick up these 30 of us from here go to this table and see what is the story point it is attracting a story point of three so three becomes the story point of my rafin story now they may be a case where UI developer says that instead of 30 Awards the end to an estimate of the story is going to be 50 overs some one side saying 55 DB is saying 40 and tester saying say 52 the close enough estimate which can be agreed by the team is 50 words I'll take these 50 overs from here to this table and see what is the story point it is detecting story point of 5 so 5 becomes the story point of the reference story in that case now whatever the estimator given here they are close enough estimates which think and easily agree there may be a case where team is giving some random numbers or some outliers which need to be agreed upon for example you are team saying that the end-to-end estimate for this story is 60 words some aside saying 50 DB is saying 30 hours and testers saying 45 now there are clearly two out layers one is sixty another is thirty the scrum master is going to talk to these gates why do they think it is going to take sixty or thirty so they will explain their positions there will be another round of brainstorming between the team and they will agree some consensus being made between them for analysts to make that estimate can be taken up to this story too beside the story find for the reference to me once we decided a story point for the reference story we're going to cancel out this table we're not going to use it any further to keep it simple and for the purpose of our example let's assume the team has agreed a story point of three for this story so let's make it three and see some examples let's say the next story in the backlog is this Login story which is having two text boxes and a simple logic of authenticating the user this story based on 3criteria x' when compared to the reference story is suddenly going to at read a less story point a story point of two now this is a story which is having a search functionality this is story is having some search criterias like customer first name last name date of birth and age based on which a grid is populated this grid is having to scroll bars one is horizontal and another is vertical any click on the scroll bar is going to process some buffering logic and pull the data in this story when compared to reference story on these three periods is certainly going to attract a higher story point a story point of five now they may be a story which is falling between these two story story which is more complex than the reference story and less complex than this story in that case we will assign story point of five to that story and the search functionality story is going to attract story point of eight whatever the stories are discussed here they're more or less simple stories they may be story which is complex story a story which is making a call to the third-party system getting the data from there coming back to our system and processing the logic in that case this is - is going to assign story point of thirteen and they may be stories which are more complex and much bigger than all these stories we talked about in that case these stories will be assigned a story point of 2134 and even more but team need to understand that these stories are epic stories which need to be broken down into smaller stories even in some case a story point of 13 may be on the higher side and team may think that this story cannot be delivered in the spread in that case they need to talk to the scrum master and agile coach to find some splitting technique so that this story can be broken down into smaller stories and can be delivered in the spring what is the highest story point beyond which the story should be splitted it all depends it depends on the agile delivery model and the duration of the Sprint so that's all about the reference story and the process of estimation around it I just mentioned one technique which can help people easily land on the story point estimation method thanks for watching this video let me know your questions or feedback stadium

Channel: Ajeet S. D.

Views: 73,374

Rating: 4.8421955 out of 5

Keywords: agile estimation, agile story point estimation, agile estimation reference story, user story estimation, pmi acp estimation, agile development, agile scrum, agile sprint, fibonacci series, agile reference story

Id: -TLQNoPoyEE

Length: 10min 12sec (612 seconds)

Published: Sat Nov 11 2017