What data are you not taking advantage of?

July 14, 2010
I just did two postings on how you need good metrics to help guide your management of systems.  They can be found here:

But what metrics should you be looking at?  Busy and productive are two different things. I don't know your business, but I can provide some suggestions I have used at multiple companies.  With today's economy all business are trying to do more with less.  Everyone is running as fast as possible.  But with all that running are we missing out on what is right in front of us?

  • Who is looking at your web logs?  What are your users doing on the site? How do your users really use the site?
  • If it is a SaaS solution you can still build in the functionality to track what users are doing.  
  • Are you using A|B testing to test the changes / updates you are working on?  Today there is no need to guess what works, you can measure it. Using real users.
  • Are you accurately tracking development hours so you know how long features are taking?  Are you comparing actuals back to the estimates? Don't tell me you are not balancing the investment (cost to build the feature) to the return!  No one has too much development capacity!
  • How about bugs?  Are you tracking bugs back to features? How about back to the development team?  Are you catching bugs early before the code is released to the users?
  • What about help desk tickets?  Are you seeing trends?  Something more systematic that should be addressed?
  • How about sales? Who is selling?  What are they selling? Where are they selling it?  Where do you see unexpected success?
  • How is your forecast?  Do you know what to expect next month or 3 months from now?  How are you making informed business decisions without a good forecast?
  • Are you measuring response rate?  Are your marketing messages resonating with you customers?  
  • How about your search engine rankings?  How well are you doing SEO?
  • What are your customers telling you?  Are you listening?  I wrote a whole post on these customer gifts.  And another on using Customer Support to help drive sales.

These are just a few areas where you could be looking at real numbers and making better decisions based on real data.  Not guesses…  So slow down for a second and think "what data do you have today that you are taking full advantage of?"   And what data should you be looking at?  
 


Part 2 – If you can’t measure it, you can’t manage it

July 13, 2010
I just did a post on "If you can’t measure it, you can’t manage it."

I was thinking, I should have brought up what to do with these metrics. Often these metrics are used as part of a Continuous Improvement Plan like in a Shewhart cycle (or sometimes known as the Deming Cycle).  A Shewhart cycle is Plan, Do, Check, Act (or PDCA).  In many ways it is just common sense.  If you want to improve a process, you should: 
  • Select the metrics you are going to measure
  • Benchmark where you are now
  • Align group and individual goals to what you are measuring
  • Clearly and transparently publish how the numbers are trending over time
  • Then use the metrics to improve processes and procedures
  • And start all over again (thus the "cycle" in "Shewhart cycle")

To me the most important part of the process is deciding "what should you measure?"  Be careful here, because this will effect behavior.  If you measure it and align goals to it, employees will work to improve it.  
  • So don't just measure and pay sales people based on total value of the deal.  They may sell a lot, but maybe not a lot of what is important to the business!  Set the bogey based on the value of deals of the products the company wants to invest in going forward. Or even better based on the profit of the deal (but I have never seen this done successfully).  
  • Don't measure just lines of code per week, but instead lines of code per week and the bug rate (bugs per 1,000 lines of code).  Or even better base it on functionality (function or feature points) and the bug rate.
  • The most important metrics for a consultant are probably billable hours worked in a year and customer satisfaction.  
So pick smart metrics that are truly important to the business.  Not just because "we've used them in the past."

So maybe the cycle should be: 
  • Very carefully select the metrics you are going to measure
  • Benchmark where you are now
  • Align group and individual goals to what you are measuring
  • Clearly and transparently publish how the numbers are trending over time
  • Then use the numbers to improve the processes and procedures
  • Validate you are measuring the correct metrics still
  • And start all over again (thus the "cycle" in "Shewhart cycle")
Note: And, before someone else brings it up, I will agree that not all things that are important to a business can be easily measured or, at the very least, the perceived return on the measurement may not be worth the investment…  Like I said common sense should rule…


If you can’t measure it, you can’t manage it

July 12, 2010
"If you can’t measure it, you can’t manage it." (source unknown)

Too often I have seen Management run the business based on gut feel and hunches instead of real data.  In this day and age, with the technology we have, that just seems slightly lazy and very risky to me.  

While working at a large automotive information site I learned the value of data.  We measured everything!  It took work to gather the data and to manage all that data, but decisions were based on real data, not on guesses.  We were able to tune and improve even the best idea based on real customer data.  For example – make a change to the page flow, throw it up on an A|B test, measure the results, better? push it up to everyone!  Or worse? tweak based on the results and try again. It took the ego out of the equation.  Decisions were based on real data not someones hunch. 

So are you basing your decisions on data or guesses?

Note:  I tried to find the source of "If you can’t measure it, you can’t manage it."   But I couldn't find a clear source.  I was sure it was Deming, but… now I am reading it is often incorrectly attributed to him.  But it seems he did say "In God we trust, all others bring data."  And I found that Wikipedia has a Drucker quote of "You can only manage what you can measure."  So pretty good company and close enough for me.  I would think they would both, within some common sense limits, agree with me!  🙂


%d bloggers like this: