In Part 1 of this 2-part Webinar, Anthony Crain, Delivery Manager at Blue Agility, presents simple agile practices, such as agile estimation and innovation accounting, to achieve what more complex portfolio practices often do not. Crain demonstrates how to harness the power of JIRA to standardize and centralize these practices resulting in more dynamic portfolio management and reporting.
Presentation TranscriptMyriam: Good morning, good afternoon, good evening to all. We’re very glad that you’re joining us today for our webinar: Optimized Portfolio Performance with Typical agile techniques into. My name is Myriam Sanie and I’m with the marketing team at cPrime. Joining me today is Anthony Crain, Senior Enterprise Transformation Coach at Blue Agility and Simone Chen, Consultant and Technology Manager at cPrime will be sharing with us some actionable items surrounding portfolio management to help us optimize the process. And we still have several people that are still joining the session. So, before we get started I’ll share with you a few housekeeping items. Firstly, this session will be recorded so you will have the opportunity to review and you’ll be able to share the presentation with those around you. We will also be sending pdf of the slides to you via email within the next 48 hours, so you can be on the lookout for that. If you’re experiencing any difficulties with the technology, please enter a comment in the question box and we’ll do our best to try to help you and if you have any questions for the presenters, you can enter them in the questions box. And then during our Q&A session, I’ll pass them on to Anthony and Simone. And so they will be able to answer that once we get to that portion.
And now a little bit about our speakers. Anthony Crain is a senior enterprise transformation coach at Blue Agility and has been leading agile transformation and organizational change for almost two decades with IBM and Blue Agility. He has led change in a broad range of verticals including banking, healthcare, government, manufacturing, and more. One of his passions, and I can tell you, it really is one of his passions, is turning theoretical concepts into actionable concepts. Also, today, we have a special guest Simone Chen, Consultant and Technology Manager at cPrime. She’s an expert at Atlassian and manages engagements with organizations ranging from startups to fortune 500 businesses and she’s always keeping an eye out for new, crazy ideas in technological developments. So, I’d like to welcome you both and without further ado, I’m going to hand over the baton to Anthony to start us off. Anthony, the floor is yours.
Anthony: This is Anthony Crain and I’ll be talking to you guys about portfolio management. And I know when I first heard about portfolio management, I felt like everything was kind of very theoretical and very hard to figure out what am I supposed to do with all this stuff. I’ve been doing it with a bunch of clients and thinking about it quite a bit. This is what I’m going to show you now, is kind of what I want to come up with and maybe it will give you guys some ideas of how you want to do it.
So, we’re going to start off with a poll. So, Myriam, I think you’re going to take care of this part for me.
Myriam: So, on your screen you should be seeing the poll. And our first question is really we want to find out why is it that you were interested in attending this webinar. You’re already implementing portfolio management, but you want to learn more about best practices. You’re planning to implement portfolio management. You’re trying to implement portfolio management, but you really could use some help. You’re just curious to learn more about portfolio or you don’t know, you’re not sure what you’re looking for. And we’ll give everyone a couple more seconds. All right, so Anthony, it looks like we have about almost half of our audience is just curious to learn more about portfolio management and about 25 percent of our audience is already implementing portfolio management and looking for best practices. About 18 percent are planning to implement portfolio management and 12 percent trying to implement. And then about three percent are not sure that. Is that a result that surprises you?
Anthony: No. Well, I want you to see that only three percent aren’t sure. Feels like eighty percent are. So, this is exciting to me. It’s exciting that so few people said they didn’t need help. You know, as I’ve helped a lot of different customers, they’ve been so far off the mark on how to make their portfolios successful. If we can give some people with some best practices and the rest of you get started, that’d be great.
All right, so this is our agenda and I don’t know if you guys noticed, but this is a part one and part two seminar. So, for part one, we’re going to cover the first four items. And then in our second part of this series, we will cover five, six, seven, eight. But I did want to show you the full agenda, so you can kind of see where we’re going. Part one, we’re going to talk about why we even do portfolio management. What is the purpose? I find that a lot of people are confused by this, so by clarifying it should make everybody’s lives a lot easier and then we can actually build towards having a healthy portfolio that’s actually actionable.
Now once you understand what it’s for, one of the things that’s very powerful is to have what’s called a portfolio baseline that allows you to compare current and future initiatives to the ones that you’ve already done. It’s something that oddly, a lot of my clients don’t do. As I’ve learned, many of my clients don’t even have any history whatsoever. Others have it. But when you look at the history, the data in there is suspect. They believe that the data is not right either. It’s all initial data, but they never actually put their final data in. It’s amazing how few people actually have actual data in their data center portfolio database because so many projects close and nobody records: what was the actual cost? How much time did it take, right? Without that data, it’s going to be very difficult for you guys to ever improve your ability to estimate in the future. So, we need to talk about how you create a historical baseline if you don’t already have one. Now, once you have a historical baseline, you can actually then use it to estimate all of your future stuff. So, the stuff that’s in proposal, for example, so, you have proposals, you know, and you estimate those proposals well, you can use the historical baseline to come up with those estimates very quickly. You can also validate your current state portfolio. So, you have a bunch of stuff in flight, you’ve already asked them into those and they’ve already been approved, but are those estimates realistic?
But if those products are mature, those initiatives are mature than you might already know the answer, but for the younger ones, you might still be wondering. And so, we can apply the same technique to validate whether the current state numbers are in any way. and so that’s kind of the first half of these things. The second half to me is where it gets really exciting. So, I guess we’re teasing you to come back. Everybody does all of these innovations, right? So, for example, I go into clients and I am asked to help them become more agile. Well, often in fact, always every client at some point asked me the question, Anthony, how do we know if we’re getting an ROI on the IT transformation itself? And my answer to them is, if you don’t know how to answer that, you have a portfolio management problem, so we would love to help you with that.
That’s another opportunity for us to be partners and bring up your name because your portfolio managers should be able to tell you whether the agile transformation is having an ROI. Asking us if it has an ROI is a conflict of interest. We’re going to be like, absolutely. Otherwise, why are we here? So, your portfolio managers need to learn how to look at innovation such as going agile or adopting DevOps or deciding to increase your offshore-onshore blend, right? Whatever that innovation is, we need our portfolio managers to be able to tell us, hey, yeah, you know, you are actually succeeding with this particular initiative. Every business has business objectives, but it’s interesting all the times a business doesn’t understand the technical portfolio and so they’re not sure if the type of portfolio is aligned with their business objectives.
Meanwhile, the technical portfolio owners are actually with the practitioners doing the development. They know what they’re working on, but they not necessarily know what the vision objectives are, and so they couldn’t tell you either if it was aligned. So, we have a problem here where everybody’s doing a lot of good work, but we have no idea if it was in the direction that we’re going in. The other thing is your portfolio’s purpose know, learn that when we get to number one is to return on the investment of I.T, right? So how do you report that? How do you put together a report that says, you know, year by year, here’s how we’ve been doing, here’s how we are doing, here is where we’re getting better, worse, and therefore, you know, what should we do about it? So, at number seven, I’ll show you one style for doing that.
And finally, number eight, is one of the biggest challenges in portfolio management is allocation of human resources to work. There’s lots of different methodologies for doing it. And so, in number eight, we cover all those methodologies. Well, obviously we cover a good eight or so of those methodologies and how they might be useful. So again, one, two, three, four today, let’s get your portfolio up and running. In five, six, seven, eight, let’s take your portfolio to the next level. Please do ask questions as we go. So, number one, what is the purpose of portfolio management? So, bottom line is to maximize the ROI of your IT Resource. That’s it. So, everything else is just complication and, and trying to figure out more basic, plead this thing. Think about portfolio management, the stock market, right?
We invest in stocks to get the most amount of money back out of the money we spent. It’s no different with initiatives in a portfolio. What I find funny about this is that a lot of my clients, they say they’re doing portfolio management, but in fact they’re just doing portfolio planning. What this means is, they decide what to fund from the beginning maybe sometimes six months before the year starts, they decide, or even a year before the year starts, they’ve decided what projects to fund. Then once their project starts, they never make a change after that. They just assume all those products are going to actually meet the business cases that were outlined for them. But we all know that that’s not the case. So, with the portfolio planning, you decide how to allocate funds and resources with portfolio management. You shift that around just like you would in a working with stocks. You invest more in the winners, you perhaps abandon the losers and invest in others that you’re holding and waiting because we thought were less valuable.
But as our projects start to fail, while something that seemed less valuable suddenly becomes more valuable than the one that’s going down. So, we should learn how to shift our investments to those other projects. So, maximize the Roi on IT resources. Specifically, what should we work on? We will always have more work to do than the IT department can manage, who will do the work, right? Big problem in most of my clients, their work in progress or WIP is too high and so they actually get a whole lot less done. And what’s funny is they assign people to multiple pieces of work because they think somehow that’s going to get more work done and it takes a lot of training for them to realize that in fact, the opposite happens. So, if we can allocate and dedicate resources to completing one thing before starting another, then you will actually get more work done in a shorter period of time.
So that’s basically your definition. This is why we do portfolio management. Now, I do have my little metric system called QPEE, and it’s how I like to measure all the work that I do in software development. Whether I’m helping a team, whether I’m helping an organization or anything in between. QPEE measures are my guiding thoughts mode in order to find good metrics. So, you can see here at it stands for quality, predictability, productivity and engagement, and for quality an odd thing, we include value in there, so a high quality product has high value. Some people don’t love that definition. That’s what we have today. Who knows how to look in the future. Defects, compliance, customers that these are all considered elements of measuring quality predictability is all about accuracy. And also, time to accuracy. How fast is it before, you know, you’re right? This is huge when it comes to portfolio manager, if we say it costs a million bucks and it ended up costing $3,000,000. At what point did we realize it was $3,000,000? The earlier we realize that a million-dollar project is in fact a $3,000,000 project? So, the sooner we can make decisions about whether or not it’s still a top 10 item, if we were working on, for productivity, we look at time, cost, scope, your basic iron triangle. An engagement is all about the people who are doing the work. So, I’ve been in accounts where they thought their quality, predictability and productivity were actually quite strong, but there are people quitting left and right because the way they’re getting to it was everybody was working overtime and they’re being command and control and lots of other negative work environment things. So, in addition to making sure that we have good quality predictability in productivity, we need to make sure we have good engagement, that we’re growing the skills of our people, that they’re satisfied, that they’re staying with us, that they’re not working tons over time unless they feel like it.
Successes. One of the companies I worked at, a 30 of us started together at that company. By the end of the third year, there was only about five people left of their original 30. When asked, the 25 all said, none of them had actually been on a project that was successful and they were sick of it. So, we have organizations who can turn out failure after failure and this will allocate people to the next project and they don’t realize how that also can demoralize a team. And then innovation effectiveness. We talked about that a little bit, it’s point five on the agenda. That’s the one where we figure out which innovations are the most effective. Again, effective innovations should increase engagement as well.
So those are the QPEE measures. If we take those four categories and think about specifically portfolio management, then the question is how can we apply this QPEE thinking to a portfolio? So same metrics, but now we’re twisting it to align to the portfolio, and here’s a first cut. First, for quality, a high-quality portfolio is aligned with the business, we need to be able to measure, and prove that it is in fact. So especially since the business can’t tell and the tech can’t tell, we need some way to make that more transparent. The other way, however, is production defects. So, as we put stuff into production we need to understand how many defects were putting into production as well. So, these are two great portfolio metrics and as we try different innovations, we want to see alignment going up or production defects going down. For predictability, what they want is estimation accuracy. When you say it’s a year, is it a year? If you say it’s a million, is it a million? And we also want to know how much it costs to estimate our projects. As part of predictability as well. Cost estimation. You could argue that that goes under productivity because of the cost metric, but because it’s tied estimation, people like to put it in predictability. Then initiative success profiling, this is one that I’m stunned that my clients can’t do. I ask them, could you create like a little pie chart of the number of projects that have succeeded, partly succeeded partly failed or failed, and you create a pie chart that shows that by dollars, so how much money do we sink into failed projects currently failing privately, probably succeeded and our successful projects. And the first thing I tend to get asked back is what’s the definition of success?
And I’m like, well, I can definitely give you my definition of success, but the fact that you’re asking that question means you’re not tracking. And if you want to understand portfolio predictability, how about predicting how many of our projects are actually going to fail every year? How many products are actually going to succeed every year? Both of those numbers are important because if you want to get good at reassigning teams to better work, then you got to be able to recognize failures. If you can’t recognize failures, you’re just going to keep people on failing projects. And that’s bad for everybody, for engagement as well as for productivity. For productivity, the things I like to look at here is how long does it take us to complete an initiative? How much does it cost, how big are my initiatives? How about my cost per scope size and that sort of thing, time per scope size, or looking at throughput: how many things we get through the organization in a given year? Also, I like looking at the cost of development. And that one is usually pretty easy to figure out. Finally, engagement is all about innovation effectiveness, being able to prove which innovations are actually not making our portfolio stronger
Myriam: Kyle is asking if you can expand on the estimation accuracy. He’s asking why does it matter? Is it our goal to become great estimators? He says that he has tried not do much on the estimation because what they’re looking to deliver value and not specifically become estimator experts.
Anthony:That’s a phenomenal question and a challenging one too. So, estimation accuracy, the reason that we want to estimate is so that we can predict what the end of the year is going to look like, right? And so, if our estimates are very inaccurate, it becomes very hard for us to predict that. Now there are certain portfolio resource allocation techniques, for example, in SAFe, they like to allow the trains to pull work to themselves. So, in such cases the train simply pull work of the stack. However, those trains still need to be able to say, we think the end result is going to look like this. So why is it important to predict the end result? Well we want to be able to announce to the market what they can expect from us. We want to be able to announce to someone who has a new idea about how long they’ll have to wait before we can get to that idea. If we don’t have good estimation is very difficult for us to paint that picture of where we’ll be in a year and paint that picture of how long it’s going to take before we can get to these new great ideas that are coming into the flight. So those are two of my favorite reasons that I like to improve my estimation. I like being able to tell people what to expect in terms of delay before they get the value that they’re hoping for.
So now you need to create your historical baseline. So again, if you have this in place, that’s phenomenal. The first question I asked someone who says we already have it, is, do you trust it? Do you trust the data in there is good? If it’s not good, then it’s almost as bad as not having it altogether. However, it will give us a head start on having what you have, so that’s good, we just need to repair it. Other teams, other companies, other organizations don’t have anything at all in this space. How do I test whether you have an historical baseline? I ask a simple question. If you have a good and healthy historical baseline, can you show me a report on productivity over the last five years? That’s a simple question. If you have a good historical baseline you feel the answer to that. You say yes, according to her historical baseline, our productivity has either gone up, gone down, stayed flat, gone up and down, whatever it is. You should be able to show me that report on productivity for five years. If you cannot do that, then even though you say you have a historical baseline, it sounds like it either doesn’t exist or the data is suspect. People don’t know where it is or how to use it. Something’s wrong if you can’t produce a simple report. So, assuming that you cannot do that, what do we do about it? So, you’ll notice on these slides, I have a three point slide every time.
First, my test question. The thing that you can ask your portfolio to see if they really do have the capability we’re looking at, in this case, historical baselines. The second part will be make it happen. So, you don’t have it. Here’s how you can get to it quickly. Finally, results, if you do this, how will that benefit you? Besides being able to answer the question, what else can we do? So, let’s look at this one, make it happen. If you have no historical baseline, then at this point you’re going to need to get together to some of your veteran thought leaders who have been around a while. You’re just going to sit down and you’re going to brainstorm out as many of these historical niches as you can. If you get them in a room and you start writing them out as they write some down, it’ll remind people of others, and little by little, it will snowball until you have as good a list as you could possibly get. That list might only be six months long, it might be a year long, it might be five years long, right? Just depending on who’s in the room and how great their memories are. Another possibility is that you do have a portfolio but you’re still not able to answer the question because the data is suspect. Well, the good news is at least you have a list of historical initiatives. However, you’re going to want to look through them to make sure that they’re actually good. Eliminate ones that don’t make sense, add ones that are missing, and you’re back to having a list. So, the first thing is get the right people in the room to create that list. Now comes the fun part. If any of you guys have done agile estimation, this next part should seem a little familiar. This uses agile estimation, but it actually uses a specific type of agile animation where you first store things from smallest to largest. A lot of people don’t do this in their agile estimation and the advantages are if you sort the things we’re about to estimate from smallest to largest, when you compare them to each other, you’re comparing apples to slightly larger apples instead of apples to elephant. So, forwarding them into order cannot begin there. First, you create your list of initiatives, maybe put them on post-it notes, make it more fun and interactive. Next, you all work together to sort them from smallest to largest. There are lots of great techniques to do that, the white elephant technique or calm the line technique is a favorite, but as long as everybody’s voice is heard, get them sorted. Now that you’ve got your initiative sorted, the third step is to assign Fibonacci numbers just like you do in normal agile estimation, but you’re doing these not as story points, but as initiative points, we’ll call them IPs, right? So, take your smallest initiative and it’s a one, now look at its neighbor. Is it also a one or is it a two. So we’ll compare them, if the next initiative is still at one, we’ll put one on that. And so on until every initiative has a Fibonacci on it.
If you did, if you simply that down, created your list, assigned the Fibonacci numbers, sorted it by year or by quarter. If you don’t have more than one year of data added up the Fibonacci points for that period of time, you could actually instantaneously graph how your productivity has been overtime. So that’s how we do it. Simone is here and she’s going to show us how she might do that in Jira.
Simone: Great. Right now, what we’re seeing is a particular view in Jira. One thing to make a note of is that he native Jira hierarchy is pretty flat, so it makes it a little hard for portfolio management. There is a suite of various add-ons that we utilize to offset that. We’ve opted to use the Structure plugin. What we’ve done is pulled a series of different issues in our initiatives and assigned the estimates that Anthony had mentioned. So, as you can see we have the Fibonacci sequence in a drop-down format that we can allocate to each of these individual initiatives. The good thing about these add-ons and view is that we can generate aggregate numbers, meaning if we wanted to see initiatives by year or by some other groupings, we can do this, and also do line edits as well as needed.
And so, what you can see on the screen here, if we go ahead and collapse it, is that from historical view, 2015 and 2017 are comparable in terms of aggregate estimates. Whereas something is off about 2016 and so we’re able to then drill down and see, ok, maybe we’re missing initiatives, maybe the sizing was wrong or something else is happening on here about essentially what we’re doing in this view is, or simply going ahead and comparing sizes of each of the various initiatives and seeing the differences across a yearly basis. And that’s it for this particular slide. Anthony, I’ll turn it back to you and flip over to the next slide.
Anthony:If there’s any questions on this, this is kind of out there for some people it’s kind of obvious to others. So, if you have questions is a great time to ask them. I do want to point out that sometimes you won’t have multi-year data, yet. Again, if you can come up with it, if you can mine it, if you can do that mining for those sorts of initiatives, even if the portfolio was not complete, it’s still better than nothing. If you can’t get more than a year or two of data, you can do the same kind of view by quarters, instead of by years.
So, the way we do it is if you have two years or less, we do it by quarter. If you have three years or more, you do it by year. You can use quarters to get enough data points to kind of see trends. So, either one of those will work. Well, let’s go back to the same slide because we want to do the results piece of that slide. So, results, kind of obvious on this one. You can tell if your productivity is getting better or worse over time. Again, it’s all about as good as the amount of initiatives you can mine out. But, you can still get a reading on it and then either blame the data or really discover that you have patterns that we need to start to think about. If you cannot do any history whatsoever, well start now. List out your current initiatives and do exactly the same technique and by the end of the year, you’ll have your first year of history. One year from now, you’ll have your first year of history. I’ve been with companies who have been working on their portfolios for two to five years and they’re still not able to use them to make any decisions. So hopefully with this they’re actually able to make real decisions based on their actual histories. So, either start now mining backwards, or start now going forward.
Myriam: Andy asks, what are we estimating at the initiative level? Is it effort, value, complexities, strategic importance? And he’s also asking for how you define productivity.
Anthony:Great question. So, strategic importance is rolled up under value. The more strategically important, the more valuable it is, so we kind of put those together. We’re going to talk about value in part two. For simplicity, if you want to be able to understand the Roi of your portfolio, you’re going to need to master your value. For this part, we’re just looking at productivity and you ask how we define productivity. We’re defining it basically on effort, initiative point size. So just now we took all of our initiatives and we estimated them in Fibonacci based on size, right? In terms of how big of an initiative is it. So, this helps us to understand when It is working on 10 initiatives, we could just count that as 10, but one of those initiatives was gigantic and some of the others were quite small. So, it will be a little bit inaccurate to count that as 10. So, all this does is allow us to give a little bit finer cut on the level of complexity of the initiative. Now, why does that matter? It matters so that we can predict the next year. how much is asking too much of our development teams. And this goes back to the estimation question. Why is estimation important? Right? Why is estimation accuracy important? We want to be able to ask our development organization to produce as many results as they possibly can, but we need to understand they need to set the expectations. But here’s the funny part. We should not set the expectation for the developers. The developers need to set the expectation for us. Here’s how much we can get done based on our history and by sizing these things, we can give you a better picture. I’m betting most of you guys have heard of the acronym MVP for minimum viable product. MVP is out everywhere. Talking about MVP. I want to write a paper on MVP because I think there’s some things people are missing. MVP is asking what is the most, what does a minimum viable product that would actually be worth the investment that we’re putting into building the MVP?
Who defines that? That’s the product owners, that’s the business, that’s the customer. They decide what MVP is. The development team, however, has to be good at defining what the MPP is and this is something that kind of made up as a minimum possible product. So, the minimal possible product is what the team knows they can build. The MVP is what the customer wants. If the MPP is greater than the MVP, we’ve got a winning business case. But if at some point we realize the MPP is going to be less than the MVP, now we have some decisions to make. There’s lots of cool patterns you can use when MPP is less than MVP. So we as developers, it is no longer acceptable to say, I cannot get you what you want in six months or in three months or whatever the timeframe is. I cannot get you what you want.
That’s no longer the agile world. The agile world says, I can always get you something. It might however be less than what you were hoping it would be. Well, if that’s the game, we need a language to define what the MVP really is so that we can start having good conversations about what we want to do with our funds for the year. So, productivity is measured in these initiative points and these initiatives points are used to describe what the minimum possible product is for an organization for over a given period of time.
So, we now have started our historical baseline. The next question we ask about historical baseline, two more questions now. And again, these are questions you can ask of your portfolio owners over your portfolio. Do you use historical actuals to come up with your future estimate? A lot of people do that with their gut, but they don’t do that in any more formal way than their gut. Others use very advanced mathematics. Again, those have proven to take so long that people just don’t do it. So, the question is, do we really use historical actuals to come up with future estimates? And don’t forget, I told you a lot of my clients would never record actuals.
By the way, I dug into why is it that it’s so rare? You would think that this was so obvious and here’s what I’ve discovered. I asked who owns putting in the actuals? Who is the person responsible for writing down the actual cost and the actual time? Historically it has been the project manager over the initiative. Well, by the time the initiative ends, one, there are often embarrassed what the real numbers are and they’re the last person who wants to write it down. Two, they are already focused on the next initiative that they’ve been asked to run and don’t have time to even think about the old initiative and documenting it closing. So, what’s happening is we have a ceremony, an event, where you close an initiative. We’re asking product managers to own that event. They don’t care to do it, they have no benefit in doing it, it’s all pain to do it, so they’re the wrong person.
But if you have a portfolio, then all the initiative owner has to do is announce this is closed. And then the portfolio manager and the portfolio team, they’re the ones who do the actual closing activities because they’re very motivated to find the truth so that they can help to paint MVPs going forward. So, you want to make sure that you have the right people closing out this data or you will never get it. So, we’re looking now at predictability, remember when we’re looking at those QPEE measures. Predictability was the first p, and the second metric, estimation accuracy and cost of estimation. So, one, you use historical actuals to come up with future estimates and two, how much does it cost you to come up with new initiatives. So that’s the other thing is a lot of people, it takes so long to estimate something that they just don’t have the time. So, make it happen. If you don’t have this capability today to use historical actuals or if it’s taking you too long, coming up with new estimates takes you many weeks with 10 people in the room, all debating. Well the database part of this will be really hard and the infrastructure would happened to be bought and acquired and then configured right? If they’re going to that level of detail, you have a lot of people that you need and it takes a long time and therefore the cost of estimation is quite high. So how do we get this all to be cheaper, faster, easier? So, we go back to that view that we just saw where we were able to put the size of initiatives on and now we’re going to add just to add two new columns to our database.
We want to put down the cost of every project and the calendar time it took to build that project, not man hours. People always ask about man hours vs counter time. Very clear man hours is a piece of cost. Man hours help us understand cost. Counter time help us understanding time. So, if you tack back man hours or person hours, if you want to be politically correct. Person hours is just you’re basically tracking cost twice. So, it’s always costs encountered. So, we’ll add those two columns to our list of initiatives. And if you do this simple thing of adding the columns and then actually get a portfolio team whose job it is to record actuals. You will now be able to determine cost and time ranges for new initiatives in just a minute. How does that work? Quite simple. If someone comes up with a new initiative, you can simply sort for projects that have similar size and look at their cost and time actuals. Getting ahead of myself, let’s take a look at how you can capture the time and costs using Jira.
Simone: One thing to note is that the columns that you see here really reflect the fields that are on your Jira issues, so you can use a combination of default system fields as well as custom fields, in our case, the estimate, what the custom fields and the costs as well as timing values be. Now the one nice thing about Structure that you can have multiple different views associated with one particular plan, so to speak. So, what I’m going to do is go to the expanded view so what I did was just change from my basic historical view that only focused on estimates to my new historical views and obviously naming is something you can change. Now you can see that we have costs as well as initiative time in weeks associate each of our initiatives and what we have rolled up as well are the costs by year, as you can see here. So, let me just collapse real quick so you can see that the initiative time rolled up across the different years is this amount and same here for costs. Now, once again, because of this particular view, we can actually go in and modify on the fly. Given these historical values is probably not something we necessarily do, but when we move forward onto future and current portfolio assist become increasingly important. With that, back to you Anthony.
Anthony: We’ll take questions if you have any, but I do want to point this out. Doing this aggregate view. So, notice you can see all 2015, right? The total cost. You also always know, even if you haven’t had this data, you know your cost of IT, that’s a number that you can always mine for over the last five years, easily. As you try to guess your initiatives from the historical perspective, you can add this up and see what the gap is. So, notice that in 2015 it was 100,000 dollars. Well, if the cost of IT was $300,000 a year, you know we’re missing two thirds of our initiatives. That can be quite helpful when we’re trying to understand productivity from year to year: two thirds of our initiatives cost as much money. Also, you have an aggregate estimate on size. It cost us 100,000 dollars to build eight initiative points. It took us 20 weeks to build an eight initiative points. These simple numbers across the top can be used to predict years on end. And we can look at our productivity over time. So now you have the Iron Triangle, the iron triangle is always been scope, time and money. But the one that has always been tricky is scope. What is scope? So now developers are fully in charge of scope. They decide how big initiatives are and we trust them because that’s agile. We trust them to give us sizes so that we can use those for prediction of the future. So, in this case, eight initiative points, 100,000 dollars, 20 weeks, that was 2015. You can see in 2016 if this was real data it is one initiative point for $2000, so if you divide that out, that’s actually doing better in 10 weeks. So, questions? Did any come in or should we move onto the next topic?
Myriam: So, a couple questions came and I think that you’ve started to address that, but maybe we can go a little deeper in that. And so, our first question is from Ramin and he says he has a team of 3000 engineers and business has agreed to fund 100 more engineers, but they also said that they want to see visible improvement and the business outcomes that are enabled by their investments. The question is without historical data, how much improvement can you promise, how would you explain it and how could you allocate new resources across existing teams? So that’s the first question. It’s a long one, but I think you get the drift. And then Sarah, on the other hand said, how do you handle historical baseline of productivity to put, if your company changes the people on the team regularly, which I think is kind of a complimentary question and maybe you can clarify in terms of historical, what can we do if we don’t have actual historical data or if that data would be very variable because of changes in resources.
Anthony: This is the fundamental problem, if you don’t have historical data we cannot answer these questions. I go back to the first question that was asked which is “why is estimation accuracy important?” Part of it is that we estimate it all. So, in a situation where you have no historical data and people are asking you to predict the future. You can only go back to the old way of doing things, look at your gut, and say that this is what I think we can do. And that has proven to be not an effective way to do portfolio management. So honestly, if I was getting questions like that, I would go to my management and say, these are the kind of questions we’re getting and they can be answered. Here is the good news. These questions can actually be answered. The bad news is we got to have history and we have not, as a company, been good at history and if we do not do good at history, we will never be able to answer these questions in a legitimate fashion. So, we created an initiative to craft this data. And if we can mine the history, let’s do it, let’s give it a two week time period to mine data and work with what we find. And no matter what, let’s start capturing current data. In the meantime, however you’re only doing it, unfortunately, is the way you’re going to have to keep doing it because you don’t have access to any history whatsoever. So, I would love to sit down with both you guys and talk more deeply about your specific situation and maybe I’d have something more exciting to answer. So, results, you now have cons and time ranges for every new initiative. And you can see again, in terms of tooling, it’s quite simple. You just need a place to put them and you need you to record the actuals and need of mine history, if you can. This slide is talking about the future. So, as of right now, if you’ve followed the first couple of slides I showed you guys, you would have a list of your historical initiatives by time, their costs, their scope and their time. Now we want us to take a look at our new proposal. So, there are new requests coming in all the time. There should be some kind of a tool where you can bring those proposals and look at them and compare them to your history. Again, we can turn to Jira for this. So, the question that we’re asking at this point, again, another way to test your portfolio is, can you use your portfolio to determine if your current proposal estimates the reason? So how do you make it happen? If you can’t? Well, go back to your tool and this time you’re going to create a list of your proposals. So, what we have so far is our history, our historical projects. Well, now we’re going to add in all of our proposals as new record, and once again we’re going to give Fibonacci numbers to all of our proposals. But in this case, we’re going to compare them to the historical initiatives. So, in the beginning, first we compare historical to historical, now we’re comparing proposals to historical to find ones that are kind of similar. Like, this one is a little bit like these others. And so, once you do that, it will pick up the Fibonacci number of its neighbor. So, let’s take a look in Jira how you might put Fibonacci numbers on proposals by comparing them to their neighbors.
Simone: Great. So, going back to Structure, once again this the historical initiatives and what we’re going to do is look at our future state portfolio. And, as Anthony mentioned, we going to be comparing again historicals, the initiatives that you see here will be present in our next view. So, the primary difference in terms of the field you see is your status because you want to include your historicals against your future state items, those that are in your backlog, in analysis or anything that’s not yet in your implementation phase.
So, in this particular view you can see that I have all these different initiatives that are estimated to be this size and you can easily compare side by side the various costs as well as the various time ranges for the different initiatives. And as you start seeing the new ones that have values, you can start comparing against the historical ones of similar tiers. And so, the nice thing about this particular view right now, it’s off to disorder against status and estimate. However, we can definitely sort to get the parameters, I can sort by cost, I can sort by overall estimate as well, depending on what kind of views you want present. However, all-in-all, the goal is to have your historical side-by-side against your future state portfolio that way you can fill in and compare your cost as well as time values.
Anthony:In the meantime, let’s click on initiative estimate for a second, here’s the real exciting part for me guys. We sort by initiative estimate and now a new proposal comes in, right? Let’s say that’s the proposal that we think is the one. Well now we can look at this database here and we can see, gosh, all of our other closed proposals, cost $2,000 and took 10 initiative weeks. Well let’s say port 40 was a new one and we’re saying this one’s going to cost a $1000 and take 8 initiatives weeks. So that’s half the price of our history, and a little bit faster than our history, eighty percent of the time that it took our historical one. With only one data point that doesn’t tell us much, but as you accumulate a few of these very quickly by dropping the simple Fibonacci number on a new proposal, you’ll be able to see the range that it should conceivably fall in. So, looking at the threes, for example, we have three of those, right 800, 2000 or the eighth down there. We have 10,000, 100,000. As you get real data in by giving, by simply comparing it to its peers, picking out a Fibonacci number, you will be able to see the minimax cost, the minimax times. And thus, in just minutes, get yourself, a level one estimate on a project without having to have a whole bunch of experts in the room debating more points. So, ask beautiful questions, we’re not going to look at them just yet. Let’s go the next piece of the slide. I just want to make sure you guys see all the parts of part one before we take any more questions.
If you do what you just saw, if any proposals are out of the expected bounds for their Fibonacci number, you can question them. If they’re lower than the lowest low than their doubtful that it’s going to cost that little or we really misunderstood it. The proposals are much higher than the high. Why is this proposal so much more expensive or it takes so much more time than others of its kind? So, we can find the anomalies, this whole idea quickly find anomalous proposals and dig deeper just on those and the ones that are inside the range, you can wait until the last responsible moment to dig deeper on. You also can quickly create estimates and proposals that aren’t estimated yet. So, if you have any proposals that have not yet been estimated, you can give them a quick range just like that.
All right, next slide. And finally, I think this is the final point 4. We can validate the current portfolio. So, we also have initiatives that are currently in flight. Well, how are they? Right? So now we’re looking at estimation accuracy and time accuracy. The question is, are your current initiative estimates reasonable? Is what we came up with before having this cool portfolio approach. Is it in any way good? We have another litmus test against our hardworking, deeper analysis type of estimates that we’ve already done on our in progress, in flight thing. So, how do we make it happen? Super easy. You’re going to take all your current initiatives, put them in the tool and set their state to current, and now give them Fibonacci numbers just like everything else. And compare those Fibonacci numbers to the ranges on the same Fibonacci numbers for your proposals and for your history.
We also need to keep track of changes to estimates. So that’s a big deal. This second bullet is very slippery. Also, keep track of changes to estimate. I had a client who told me, Anthony, we are 98 percent accurate in our actuals to our estimates. And I said 98 percent accurate? You guys should write a book and you guys should go on the road and be speakers because that’s incredible. I don’t even know why I’m here then. And someone else said in the room said “I don’t trust that data.” I’m like, “well that happens a lot. Why don’t you trust the data?” And they said, because we’re comparing it to the final estimate. We’re allowed to re-estimate projects. That’s a good thing to be able to update assessments as you learn, but they overrode the estimate field every time. They bought a legitimate portfolio management tool that allowed them to change that number in place and so they had no history of the previous estimates. So, I’m like, “wait, so you’re saying you’re 98 percent when compared to the final estimate, why aren’t you a hundred percent” right? Just change that number to the right number. So that was completely irresponsible. They truly had no idea what their initial estimates were and therefore they had no idea of how to figure out their time to accuracy. At what point did you know it was really going to be a year? Who knows? We knew it was a year because the last time we updated it was two weeks ago. So, two weeks ago we knew it’d be a year. How much before that did we know it’s a year. Who knows you’ve lost it. Let’s take a look at how we can use Jira to track this information.
Simone: You’re able to sort through by her costs and time. So, from a view standpoint, it’s very similar to your future state. The primary difference is really the sample data that you’re pulling into this particular view.
Anthony: Are you going to show the success history?
Simone: So, in terms of the history portion, are you referring to any particular initiative success or which particular one?
Anthony: Yeah, so one of the things now is what happens when we change an estimate, right? So, we decide, oh, you know what? It’s not going to be $800 for portfolio 10, portfolio 15, we set at 9,000. So instead we want to change that to double it. It’s going to double the cost you thought, so 18 grand. So, we change it right now. When we finally closed the project, it turns out it was 18 grand. But when did we realize it was 18 grand? How much of the project had gone by? We can now go into the history or portfolio 15 and we’ll be able to find the date we set it to 18 grand and say well that was two weeks ago. And then we scroll back to the previous time it was set and say, oh my gosh, it was when it was first started and it was set to 9 grand. So, this company who said they were 90 percent accurate, no, they were 50 percent accurate. Their time to accuracy was 98 percent. So that’s a terrible time to accuracy. Time to accuracy should be very low. Agile says that by demoing every sprint, our time to accuracy, will be a lot lower. We will figure out that it’s a million-dollar project and not a half million dollar project in just a few weeks instead of, until almost the entire project is going to bill. So, if we go to the history, you guys can see here in the history when we actually change these estimates and thus be able to get a time to value metric and also be able to tell what your true estimation accuracies are. So, five days ago it was 9,000, now it’s 18,000. If we closed it today, right? We’re like, well that’s not 90 percent accuracy and predictability, right? They changed it at the last minute. That’s not the same as estimating it correctly in the first. Now we do want people to be able to. We just need to be able to understand our time to accuracy. Back to that first question, why is that important? Because if we have good time to accuracy, if we know that all of our agile projects, about a third of the way through the project, time to accuracy gets to within eighty percent, then we now can announce the market. When a project is a third of the way done, what they can expect to get versus what we thought back when we first came up with a project idea for the first one. Alright. So please put your questions in on this part. Let’s move on to the next piece. Now if you do this, you’ll be able to understand the red, yellow, green status for inflight initiatives in terms of the ability to meet their business case. So, if they are way outside of the estimation bands by over 20 percent, you can set them red and say, these projects are way outside of their estimation band. We need to re-take a look at them to make sure that we’re going to be successful. That’s the business case. If they’re within 20 percent but higher low then you might step up yellow, and if they’re within the high and low for other close initiatives with the same Fibonacci number, then you can set them green as you think these look like they’re in the range so it can be more relaxed about them. You can also estimate your time to accuracy. When are your estimates within eighty percent of actuals, that’s so critical to know. Now we can set expectations for customers on what we’re going to deliver.
Myriam: So, for our next poll, we’re asking you how long does it take for your organization to create a level one estimate? And how much does it cost? Over two weeks and more than five people working on the estimate? Over two weeks, but just a few people or more than five people. But we can do it quickly. Not exactly quick but we’re doing better than 1 and 2. Super quick, using relative estimation, just historical projects. Or five you you’re not sure? So right now, I’m showing that about almost 30 percent of our audience doesn’t know or is not sure. And then about evenly divided about 25 percent, 25 percent over two weeks for just a few people, or more than five people and quickly. And 25 percent, not exactly quick but doing better than one and two, and about 20 percent say it takes them over two weeks and more than 5 people and only about 8 percent say that it’s super quick using relative estimation for historical projects. So that is some interesting data there.
Anthony: So, to the eight percent who are using relative estimation on historical projects. That’s awesome. I hope that you still found something valuable in here since I showed you something that you’re already doing. And if not, hopefully when you saw the agenda for the second half, there are things there that you haven’t mastered yet that I can help up your game on. For those of you who are in the 2 and 3, that is something to be very proud of. But again, because they don’t call me unless there is trouble, most of my clients have been in one. Those of you who are in one, you’re in good company, but it looks like that audience is shrinking. So, that’s really cool. I think we have one more slide after the poll. I helped one of these clients that was doing it, that would’ve answered 1 to the pole to move to this relative estimation technique. Their estimation accuracy was equal. In other words, they compared their estimation accuracy before they switched to relative estimation. The percent of accuracy was the same. But the cost to estimate, level one estimates decreased by 99.8 percent. It used to take them five days, now it takes them five minutes. And also the number of people in the room was fewer. It also decreased the cost of level 1 estimates by 98 percent. It used to cost them about $308 per project to estimate it, down to $8 a project or $76,000 a year. So, these are from an actual company that went from big room estimation to relative estimation, at the initiative level. And these are the kind of results that they saw. So, this is our final slide. We are out of time.
Myriam: So, I’d like to, once again, thank you Anthony and Simone for providing really some great insights on a complex topic that is portfolio management and how we can optimize this process. We’d like to thank you all very much for taking the time to join us today on this Wednesday. We hope you found this session of value. As a reminder, we will be sending out the recording as well as the slides out to you and we really hope that we’ll see you all again, part two of this webinar. We appreciate your time and enjoy the rest of your day.