I was in a meeting a couple of days ago where we were dicsussing best practices for some corporate initiatives, and one approach that came up a couple of times was, "find out what Google does and do that."

Makes sense, right?  Google is so wildly successful on so many levels it's staggering--stock price is sky high, revenue and profits are outstanding, employee satisifaction is high, and their basic product (text search) is still notably superior to their competitors.

Of course, me being me, I think blindly doing what Google does is probably the worst thing you can do.  Not because of the standard, fluffy "copying can only bring us to their level, we need to exceed them" response, but for far more pragmatic reasons--here's the underlying thought:

Success on the macro level obscures problems and failures on the micro level, whereas failure exposes problems.

Certainly Google is doing some things right to stay ahead of the competition.  Personally, I think the main reasons behind their success are 1) a fundamentally superior ranking algorithm and 2) a workable advertising revenue model.  But take something like their free food perk--is that contributing to their success?  Or is it just a drain on profits?  As long as the company is successful, you can't tell.

Contrast that with some of my favorite internet meltdowns: Pets.com demonstrated how shipping costs will kill you when you're selling a relatively heavy, low cost product.  Jasmin.com may be my all-time favorite, though; trying to sell a product (perfume) where the sole attribute (scent) can't be communicated over the internet.  Yeah, that one didn't last long :-p

I think this is really demonstrated when you get a company in transition, like Wal-Mart. For years they were the ones who were doing everything right and should be emulated.  Now that sales growth has slowed, you can see them struggling to try to figure out what needs to be changed to get growth back on track.  Personally, I see this as a sign that they don't really understand what was contributing to their success, and what was doing the opposite.

So what's it all mean at the end of the day?  View your failed projects as huge learning resources.  Figure out the root cause(s) of the failure, and redesign your processes to eliminate  the root cause in the first place.

Ben

In general, abstraction in software design is warranted if it leads to a net reduction of lines of code.

Me: So, you use C++ for the switching software, right?

Interviewee:  No, we use C.  C++ is too slow.

Me:  Hahahahaha

interface Renderer
{
   void RenderCharacter(char value);
}

 

class ConsoleRenderer : Renderer
{
   public void RenderCharacter(char value)
   {
      Console.Write(value);
   }
}

 

interface RenderableData
{
   char[] GetAsCharArray();
}

 

class StringData : RenderableData
{
   private string text;

 

   public StringData(string value)
   {
       text = value;
   }

 

   public char[] GetAsCharArray()
   {
      char[] result = new char[text.Length];

      for (int i = 0; i < text.Length; i++)
         result[i] = text[text.Length - i - 1];

      return result;
    }
}

class RenderingAgent
{
   public static void RenderData(Renderer renderer, RenderableData data)
   {
      char[] dataArray = data.GetAsCharArray();

 

      foreach (char value in dataArray)
         renderer.RenderCharacter(value);
   }
}

class Program
{
   static void Main(string[] args)
   {
      Renderer renderer = new ConsoleRenderer();
      RenderableData data = new StringData("Hello, world!");

      RenderingAgent.RenderData(renderer, data);
    }
}

class Program
{
   static void Main(string[] args)
   {
      Console.WriteLine("!dlrow ,olleH");

   }
}