Thursday, February 27, 2014

(Technology) Road Maps

When I was a kid we took a lot of road trips. Usually they were to see family and occasionally for vacation, but the part I liked was holding the atlas. I loved keeping track of where we were along the way, estimating how long until we reached the next big city (which often meant a food stop), and seeing all the other places along the way we could possibly go.

Yes, I said "atlas". Before GPS and online maps, even before mapping software existed, we carried an atlas in the car. We knew our starting point and our ending point, and we used the atlas to decide how to navigate from home to our destination and back again.  I remember my parents poring over the atlas with relatives: "If you take 568 back this way I think it will be a shorter route than the way you came." "Look, we'll be close to Chicago - we can stop at the science museum." "The last time Uncle Joe sent us that way we all got carsick - I'm not taking that road again!"

The atlas was wonderful. While we drove along I wasn't constrained to just our trip - I could flip the pages to other states and imagine all the other trips we could take and places we could go. As we passed big cities I would study the inset map of the local streets and find the landmarks we could visit someday.  When we took the inevitable detour or got caught in traffic, the atlas was our guide to a route around the blockage.

The atlas had its limits. If we didn't have the current year's atlas, we might find ourselves on a new highway that wasn't on the map.  Construction barrels could appear unexpected, any time, adding an hour to the trip. If we were off the highway in a small town there was no inset street map to guide us - and heaven forbid we stop and ask directions.

At some point there was a leap of technological advancement. The auto club provided a service where you met with an agent before your trip, told them your starting point and destination, and they would generate a customized map of your trip, on the spot.  It was a small flip book assembled from individual segments of the journey, each on their own page,  with easy-to-follow directions.  The proper highways and exits were noted.  Major construction projects were indicated, with possible detours.  Sometimes alternative routes were included on the map so you could decide whether to take the shortest mileage and save gas, the fastest speed limit and save time, or the rural route and enjoy the scenery.

This was a wonderful achievement.  No more wrestling with 6 square feet of atlas in the passenger seat.  No more fighting a state map to unfold and refold (I could solve a Rubik's Cube faster than I could properly refold a state map).  When you reach the end of one segment of the trip map, you just turn the page and focus on the next leg of your journey.  Most importantly, it was simple.  You no longer needed an experienced navigator to figure out the next exit. Any child in the car could direct the driver accurately.  Attention was focused on just the highways of interest and major exits that affect our route.  Possible delays and road blocks were anticipated with a strategy for how to deal with them.  Time estimates for each leg were printed on the page for that leg which settled all the questions of "are we there yet" and "how much longer, Daddy?"


Fast forward to the days of turn-by-turn voice navigation.  I don't carry the atlas or the auto club trip plan any more, but I find myself using the same concepts in technology projects.  We often define a "road map" for large-scale technology projects, such as major system replacements or database upgrades.  We use terms like "current state" and "future state" rather than "home" and "Grandma's", but the ideas are the same.

What made a good map for a trip in my youth still makes a good road map for technology projects today.

Is your technology road map like the atlas?  Sure, it includes pictures of your old platform and what you are going to build, but how clear is the path to get from current state to the goal? Is any of the approach explained? Alternatively, is so much explained that the path lost in all the approaches and spider-web-like possibilities?  Are your road map review sessions like those conversations at my relative's dinner table about whose route is best?  Does the team get distracted by all the possible side outcomes of the project?  (This is all necessary to get to a concise project road map, but it doesn't need to become part of the everlasting documentation.)

Is your road map like the auto club map?  Is it focused on the effort at hand, rather than all possible projects?  Is it clear and concise (can a new person on your team explain it)? Does it list the main alternatives, determined after pre-filtering many of the possible options?  Does it include expectations about where the project may slow down and hit challenges?  Is it broken into manageable segments?  Are those segments time-bound with a rough estimate based on known constraints (think simple estimates like the trip's "distance" and "speed limits", not yet calculating specific tasks or detailed effort)?

In large organizations, road map efforts can be a large project in themselves, and may span months.  If this is the case, are you updating the road maps (like we replaced our atlas) to verify the original plan is still viable in the current environment? Can it be modified to use a more efficient "highway"?

I should mention the GPS.  The GPS with turn-by-turn navigation is great, but it's not really a map any more.  You only see a small segment compared to the big picture of the atlas or even the focused picture of the trip plan.  The GPS excels at providing just-in-time directions rather than just listing possibilities. If we carry the analogy through, the GPS is not a technology road map, it is a project-level plan. It is easily viewed as a modern, agile-style, just-in-time stream of activity, directing what you need to do next, just in time to do it.  If you get off course, the GPS recalculates, similar to how an agile team can be responsive to road blocks or detours that come up mid-project.  This is quickly becoming an analogy for another topic of its own, so I'll say no more.


Clear, concise road maps are great for technology projects. Apart from the office this summer, I think the family needs a road trip. I might just turn off the GPS, hand my kids an atlas, and see where we end up.