Applying Agile development techniques to managing my own work

Product Management is a multi-facted role. Building roadmaps, thinking of new products, interviewing customers, working with developers, sales, marketing, support, training - the list is endless (I wrote previously about it here).

I recently started using an Agile-like approach to managing my day to day activities. This is what I mean:

1. I have a list of projects I am working on. That could include everything - tactical activities such as 'Sales needs this document for this sales cycle - by this evening' to 'Make progress on customer development on a product that will not be launched for six quarters'

2. Every three weeks I start a new sprint. I reassess and prioritize everything on my plate and have a strict prioritization. Similar to agile sprints, things do pop up. Just like developers might get pulled into emergency issues, requirements etc.

3. I then draw a line - projects that I plan to address in this sprint, and those that I plan to postpone for another day.

4. If I am pressed for time, I postpone projects and tasks from the bottom of the list

5. I track progress each week against each project. If one is getting left behind, and I do need to focus on it, I figure out what else needs to move down to make way for it

6. I refer to this list in my nightly planning, moving around tasks on my kanbanflow board

So far, so good! Things are working out well, and this has really helped me do the difficult task of prioritizing according to goals vs. prioritizing according to what seems most urgent. I will report back on progress in a few weeks.

Words of caution: Lean startup methodology

I am a big fan of the Lean Startup methodology, and related concepts such as customer development and minimum viable product. The Lean Statup Machine weekend was super helpful for me in learning and practicing the methodology more, and learning from an awesome set of mentors. As I learn this methodology more and more, there are some words of caution for newbies:

1. There is a big learning curve as far as practicing the methodology is concerned. Several pieces of it are science, and can be learned quickly, while others are more art than science. Take for example customer interviews. One has to learn how to structure customer interviews for maximum learning, from being flexible to focusing more on customer problems and not your solution. As you do more and more interviews, you do get better at them.

2. It is imperative that you select your interviewees carefully. What exactly is your target market? Often, one can build a solution for the general population, but identifying a particular target which might have the largest need for the solution is imperative. And one can stumble across this target by interviewing several potential targets. How do you determine best prospects? In words of Paul Graham, that is a function of "How much of a problem they have, and how quickly they come to a decision"

3. Asking if a prospect has a problem is also a tricky process. Let's face it, we hate problems. We do whatever we can to deal with them - come up with a solution, or rationalize them/fail to recognize that they exist. Let's take the example of buying an airline ticket. If you asked someone if they have a problem doing this twenty years ago, the answer would be no! I call up my travel agent or visit him/her, and in no time, I have my ticket. But then again, how easy is it to buy tickets online? If we cannot see a solution, we will likely rationalize problems. They key is not to have the whole population admit that there is a problem. They key is to find a few people who think that it is a big problem, and are willing to be your first customers, and to find a lot of other people who are willing to admit to it being somewhat of a problem, who will jump to the solution if it is really good.

4. Experts are often wrong. As the mentors at LSM NYC said, their opinion is also just one data point. They can probably give you the best advice on how to practice the methodology, but their advice is less useful on a particular concept.

5. It is great if you can have someone review your iterations - hopefully someone who knows about this methodology, but even someone who is just willing to be logical is good enough.


More to be added as I think of them....