Tuesday, September 28, 2010

Logo design contest for Gigantt

Help me find an awesome logo for my company by participating in the design contest. The prize is $223 and anyone can enter.

I did some mock-ups of the UI concept. I took the plot of Ocean's Eleven and created complete version of all the steps it takes to rob a casino in a "fractal" Gantt chart, which is going to be the basis of Gigantt.

(click image for bigger version)

This is a very crude and early sketch of the UI, so bare with me. But as I hope you can see, the idea is to take a very complex plan and by giving users a way to zoom-in infinitely they can drill down further and further into the details. 

For comparison, the same exact plan when plotted in MS Project looks like this:

You have to scroll 4 screens down and 2 aside in order to view the whole plan, unless you collapse and hide parts of it (and miss critical links). Just inputing it into Project took me 2 hours and editing it is nearly impossible. 

So, anyway, that's a sneak peek into what I'm working on. Many more fantastic features are planned on top of this ZUI.

Now go forward this to your designer friend who could win quite a few Shekels designing a logo that would capture all of the above.

Thursday, September 23, 2010

Planning at the speed of thought

How do you plan your projects? Some like to plan top-to-bottom; others prefer to start from the end goal and plan their way back. But no matter how you choose to go at it, one thing is common to all approaches - planning happens at the speed of thought. If your planning gear can't keep up with the speed at which you internally compose your mental plan, then it's bound to frustrate you.

That's why mind-maps are so great. Any GTD fan who has ever done a brain-dump using a tool like Freemind or MindMeister knows what it feels like to have your fingers catch up with your brain. It's liberating. Being able to serialize your thought process without skipping any detail along the way is key to successful planning. Thus, any project planning tool that introduces overhead into the process is unacceptable. And nearly all of them seem to do! There's always a form to fill, some fields to populate. You often have to switch between keyboard and mouse. These are costly interruptions to the thought process that add precious seconds to every item you jot down.

Thankfully, most mind mapping tools are built to eliminate this overhead. There seems to be a clear keyboard-based convention for creating mind map. Hit the [INSERT] button for a new child item; hit [ENTER] for a new sibling; navigate with the arrow keys. On my last project (a web mapping application) I opted to design and plan the entire thing with MindMeister. MindMeister is an amazing mind mapping tool, with a powerful collaborative twist. By the time my project entered the development phase, our work plan consisted of about 1,000 action-items - nearly all of them estimated in hours, and none longer than two days. By the time development was done, the detailed work plan of this year-long project had about 14,000 revisions and 1,700 items. That's a big mind-map. When fully expanded, this tree would require you to scroll the screen 20 times in order to see all of it. 

The fact that there were so many revisions illustrates how much of the initial plan had to change as the project progressed. That's another thing that's so great about mind maps - editing is so easy. You can turn an entire plan on its head in seconds with a few drag-n-drops. You can delete and throw away the parts that didn't make much sense. You can re-prioritize at will. 
The simple ability to visually rearrange your plan on screen quickly is crucial. Plans that are harder to change often remain unchanged and thus ignored. Ultimately, this makes people not want to plan at all. Why bother, right? If 40% of the plan is going to change after only four weeks then just plan ahead for those four weeks and plan a bit more when you get there. If only that were true. The sad truth is that there is immense value to planning ahead as far as you can think, but people hesitate to do so. The price of redoing your serialized plans is simply too high with existing planning tools. Many people miss out on the value of planning ahead because the process is painful, slow and time-consuming.

In fact, I propose that most work plans nowadays end somewhere I call the "screen horizon". Most people have no problem writing plans that fit into one screen, because that's the limit to ultimate quick editing. You can drag, drop, cut and paste very quickly as long as you don't have to scroll your screen too much. But no serious amount of planning can really fit into one screen. This is a real limitation, if you think about it. Your crappy planning tools dictate how far you allow yourself to think into the future. With a better tool you should be able to push that screen horizon forward and plan comfortably until you decide you've planned enough. With a better tool, you wouldn't mind planning further ahead, knowing very well that your plans are bound to change as your project advances. Changing plans is fine, as long as editing your long-term plan is as easy as dragging around items on a single screen. 

Tuesday, September 21, 2010

Testing, Testing, One, Two...

Hello, World!

Welcome to the Gigantt blog. No, don't bother googling it; nothing interesting will show up; yet... But let me tell you what Gigantt is going to be.

It's my plan to liberate project management from project managers. It's a web application that's going to make planning so much fun you won't be able to resist planning your projects in however great detail you desire.

Planning has become unpopular. Dare I say uncool. Everybody's "agile" these days, and it seems hardly anybody sees the value of planning ahead significantly. Waterfall is for people who work at Initech and Gantts are for Pointy-haired-bosses who grasp at any diagram that might simplify the insane complexity of the projects they manage.
Don't get me wrong. I'm not here to advocate any particular methodology. I'm just saying that if a methodology sacrifices forethought because changing plans is an annoying process, then that's a shame, because it doesn't have to be this way.

See, it's not planning that's a pain in the ass - it's your planning tools. Can you honestly tell me you're in love with your project planning solution? Didn't think so. Most of them suck, and all of them fall short of answering what I see as The Six Requirements of Planning-Software.

In my book, every planning software has to:
  1. Be collaborative (bye, MS Project!).
  2. Allow you to plan at the speed of thought. Think mind-maps. If there's a form to fill, that's a fail.
  3. Let you change plans as quickly and easily as first making them.
  4. Visualize very complex plans in a simple way. Ever created a Gantt with more than a 100 items? Exactly, it's horrible. And how will you ever visualize a plan that has 1,000 items? 10,000 items?
  5. Not try to solve every possible need of your organization. You shouldn't have to ditch email or switch to another bug-tracker because your project planning tool is an all-in-one solution that wants to take over your intranet.
  6. Realize that a plan is not a laundry list of tasks that are going to be performed. Plans change - that's the number one cause of project delays (not poor estimates). Plans are more like algorithms than checklists, and any planning tool has to be able to accomodate this fact.
In upcoming posts I plan to discuss these requirements. I hope to demonstrate why they are, in fact, necessary. More importantly, I want to tell you how this new project - Gigantt - will meet all of them.

Unfortunately, no one can be told what Gigantt is - it has to be seen.

But seriously, it's a very graphical and unique web user interface, and until you see how a 10,000 item plan can be elegantly visualized without having to scroll your screen until you get RSI, you probably won't understand what I'm talking about. For now let's say it's a project planning tool that resembles Google Maps more than it does FogBugz or BaseCamp. A web-based ZUI that's as intuitive and quick as a mind-map but also lets you represent very large, and complex graphs of work items in a clean way (mind maps are basically trees, and plans are essentially DAGs).

So, stay tuned. Subscribe and I'll update on my progress from time to time. If you have thoughts on the subject, comment away.