Too Much Work in Progress? How to Use Cost of Delay to Rank and Choose Johanna Rothman https://mastodon.sdf.org/@johannarothman www.jrothman.com [email protected]

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman What’s Your WIP*? • How many of you are “working” on at the “same” time? • Projects to ship to a customer of some sort (multiple “primary” projects • Efforts (investigate this, please…) • Initiatives or background research (not af fi liated with primary work) • Anything else? • Count them all and put the number in the chat. I can anonymize your answers. • * WIP is short for Work in Progress 2

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Multitasking Costs Time and Money • Impossible to predict anything: either time or money • Little’s Law explains why: WIP is a function of how long things take times the time it takes you to fi nish • Multitasking extends cycle time and lowers throughput • Everything costs more 3

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman What Happens with Multitasking 4

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman More Multitasking Effects • People have feelings!! • Work dissatisfaction • Personal frustration • A small story from my long-ago past … 5

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman No. • It’s a complete sentence • We need to be able to say no in a reasonable way. • Some ways: https:// www.jrothman.com/mpd/portfolio- management/2017/02/visualize-your- work-so-you-can-say-no/ • “No” or “not yet” is an issue of ranking 6

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Manage the Project Portfolio at All Levels • Expose all the work • Make sure you have it all • Rank the work (a big challenge) • Only assign one project to an entire team (sometimes an even bigger challenge!) 7

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Rank: Start with the Zeroth Question • Zeroth question: • Should we do this work at all? • Is this work still valuable? • If the work is no longer valuable, skip it and move to the next piece • Old work might not have value any longer 8

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman WIP & Aging Connected in Portfolio and Roadmaps • The higher the WIP, the longer everything takes • The longer everything takes, the older the un fi nished work gets • Which is more valuable, the older work or the newer work? 9 Average Lead Time = Average Work in Progress Average Throughput

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Is old work still relevant? 10

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Cost of Delay Focuses on Value • Cost of Delay is: • The cost the organization incurs when they delay a product release • Speci fi c costs on next slide 11

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Four Separate Costs of Delay • Cost 1: People Costs • Cost 2: Sales Introduction Costs • Cost 3: Delay to Maximum Sales • Cost 4: Overall Sales Reduction 12

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman How to Calculate the Costs of Delay • Cost 1: People Costs (Still have to pay people) • Multiply: (Weekly salary costs for all the people on the project) * (Number of weeks delay) • Cost 2: Sales Introduction Costs • Multiply: (Weekly anticipated sales) * (Number of weeks delay) • Cost 3: Delay to Achieve Maximum Sales • Multiply Max sales and subtract anticipated maximum sales for the number of weeks of delayed introduction (area under the curve) • Delay 4: Overall Sales Reduction • Area under the curve of anticipated maximum sales over lifetime minus lower sales 13

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman How High WIP Affects Each Cost • Cost 1: If you continue to pay people and continue to multitask, you might be in the Sunk Cost Fallacy • Cost 2: Do you expect a big splash or a longer launch? Big splashes create much higher CoD • Most orgs can easily determine Costs 1 and 2. (I rarely bother with Costs 3 and 4 because no one remembers calculus and areas under the curve) 14

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Cost for That First Time Delay • Cost 1 is about cycle time • If you measure velocity, you have no idea how long things will take. Points do not tell you how long things take. • However, you can ask the team to offer a guess about how long they think they need • Assume you have 6 people, each of whom make the same weekly salary of $2000/ week. The team costs 6 * 2000 * number of weeks. • A 4-week delay costs $48,000 15

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Cost 2: Lost Revenue from Delayed Release 16 Customers pay $1000 Expected Sales Expected Revenue Week 1 50 $50,000 Week 2 100 $100,000 Week 3 200 $200,000 Week 4 (& onward) 300 $300,000 Take the maximum sales of 300 and expected revenue of $300,000 (because we can’t depend on the delay not a ff ecting max sales) Multiply that by 4 weeks = $1.2 Million

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Add Costs 1 and 2 Together • A four-week delay (given our salary and sales numbers) is: • $48,000 in salary costs • $1.2 Million in missed sales • That’s why I rarely need to use Costs 3 and 4. Costs 1 and 2 are so high, managers say, “oooohhhh.” • (Also, Costs 3 and 4 are totally predictive and a late release changes all predictions) 17

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Use CoD to Rank Adoption/Internal Projects • Internal waste often has much higher costs • Sad story of a broken build system in a 3000-person organization • $5.4 Million per week (salary costs plus delay for shipping the external product) • Time literally is money 18

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman How to Rank with CoD • Start with the zeroth question: Is this work still valuable? • Write down all of this: • Weekly salary costs for everyone you need to pay to release the product/feature • Number of features remaining until you can release • Cycle time/feature (or team guess from before) • Then add in the expected weekly revenue (next slide) 19

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Example Comparison of Four Projects, Part 1 • When teams don’t know their cycle time, everything is suspect • Project C looks like a winner, but I don’t buy the revenue or the 3- week cycle time. Can we trust these numbers? • Turns out Project C needed more people 20

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Example Comparison of Four Projects, Part 2 • Project D looks like WSJF (Weighted Shortest Job First). A possibility. • After discussion, the client chose, based on risk due to cycle time: • Project C • Project D • Project A • Project B 21

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Collaboration and Low WIP Matters • Break this feedback loop with: • Reduce WIP explicitly • Increase collaboration to reduce WIP & increase throughput • Monitor aging • All of that will reduce your various Costs of Delay and make it easier to rank work 22

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Links You Might Enjoy • How to see how long things take: https://www.jrothman.com/mpd/2019/09/measure-cycle- time-not-velocity/ • Why story points lead you astray: https://www.jrothman.com/mpd/2024/04/how-to-move- from-story-points-and-magical-thinking-to-cycle-time-for-decisions/ • How to measure WIP: https://www.jrothman.com/mpd/2023/03/how-to-measure-all-your- work-in-progress-to-make-better-decisions/ • Flow Metrics: https://www.jrothman.com/newsletter/2024/01/ fl ow-metrics-and-why-they- matter-to-teams-and-managers/ • Series that prompted this talk: https://www.jrothman.com/mpd/portfolio-management/ 2024/08/how-to-calculate-the-cost-of-delay-to-rank-all-the-work-part-1/ 23

© 2024 Johanna Rothman https://mastodon.sdf.org/@johannarothman Let’s Stay in Touch • Pragmatic Manager: • www.jrothman.com/ pragmaticmanager • Please link with me on LinkedIn • https://www.linkedin.com/in/ johannarothman/ • https://leanpub.com/ divingforhiddentreasures-spanish 24