Fixing Perverse Incentives in Software Development

-

I read with interest an article about picking the right metric to incentivize desired behavior and thought I would add a little of my own personal insight. One problem with many (maybe most) software development organizations is that they inadvertently create perverse incentives, rewarding undesireable behavior and creating confusing and chaotic environments that, despite best efforts of all involved, seem to only on a hit or miss basis produce the desired result. Just as important, often the rewards are implicit and it isn't obvious that developers are actually being rewarded for the errant behavior.

Some short examples of widely used, but poor metrics I've observed as well as some simple and arguably better alternatives follow. For example:

  • Rewarding developers for "count of bugs fixed". Without accountability for who created the bug, this simply incents developers to release buggy half finished software that they will then subsequently be rewarded for "fixing". It's probably better to reward developers for "bug-free features delivered" or "lowest bug count".
  • Rewarding developers for "hours worked". Do we really want to incent someone to take 20 hours to do something they could actually accomplish in 20 minutes? It would be better to measure features delivered per unit time than simply the raw number of hours in chair.
  • Rewarding folks for being the "expert" in something. Inevitably this leads to situations where folks WANT to be the sole expert in something... hiding or obfuscating things or creating proprietary boondoggles that no outsider ever has a hope to figure out. Better to reward someone for creating something that is transparent and usable by all ... or better yet reward someone for teaching others how to do something than be the "expert".

Some examples of inadvertent rewards that crop up over and over again and reinforce the negative behavior:

  • We're going to pay you to your own mess. Yep, consultants love this one.. pay to build a mess AND fix it, you'll be gainfully employed for life! Maybe it would be better if we let you fix your bugs for free! on your own time! I wonder what would happen to the bug count in this case?
  • Look at the superhero. Wow! load balancer crashed and we had to call you on vacation because you're the only one who knows the root password/knows where the server is. What an ego boost! Perhaps instead we should celebrate the sysadmin who has NEVER had a server outage instead of the one who can't take a vacation because they're the only one who knows how to reboot the load balancer.
  • You're fired! This is a tricky one, and actually not a reward, but if there is a absolutely no reward in pointing out problems and ONLY risk, who's going to stand up and risk the axe for pointing out an honest error? It's much better to reward honest and constructive criticism of problems with appreciation for the effort and concern than creating a hide and go seek culture of "it's not my fault". I realize this is actually a risk, but I've put myself on a tight timeline so you'll have to live with it :)

While not comprehensive, hopefully we can use this list as a lens to view our software delivery teams and look beyond the superficial problems. In doing this, I believe, we can start getting better at constructing better metrics and rewards. This should lead to build better software, of a higher quality, delivered more reliably and quickly.

Comments

Post a Comment

Popular posts from this blog

Please use ANSI-92 SQL Join Syntax

-
Image
Way back in 1992, a standard was published for SQL. In it, the syntax for specifying joins in SQL Database was finally standardized. This was a good thing because (for folks who've been doing this a while), trying to decode the vagaries of how various Databases handled outer joins could drive you bonkers yielding different results depending on the parse order of things in the from and where clauses. Unfortunately, some DBMS vendors took a while to adopt this standard (Oracle didn't support it until 2001), but at this point if you're using a major RDBMS and NOT using it, you're probably mired in a tradition or habit that bears changing. As an example of what I'm describing, here are some examples of what I mean. Old School (no standard syntax + implementation and evaluation of predicates varied by vendor [and query]): select * from customer, order where customer.customer = order.customerid ANSI way (agreed to standard, predicate evaluation has some rules
Read more

the myth of asynchronous JDBC

-
Image
I keep seeing people (especially in the scala/typesafe world) posting about async jdbc libraries. STOP IT! Under the current APIs, async JDBC belongs in a realm with Unicorns, Tiger Squirrels, and 8' spiders. While you might be able to move the blocking operations and queue requests and keep your "main" worker threads from blocking, jdbc is synchronous. At some point, somewhere, there's going to be a thread blocking waiting for a response. It's frustrating to see so many folks hyping this and muddying the waters. Unless you write your own client for a dbms and have a dbms that can multiplex calls over a single connection (or using some other strategy to enable this capability) db access is going to block. It's not impossible to make the calls completely async, but nobody's built it yet. Yes, I know ajdbc is taking a stab at this capability, but even IT uses a thread pool for the blocking calls (be default). Someday we'll have async databa
Read more

The difference between Scalability, Performance, Efficiency, and Concurrency explained

-
I thought I'd give some explanation of the difference between scalability, performance, Efficiency, and concurrency. To make things more accessible (and because I'm watching a cooking show on TV) I'm going to use a baking metaphor to help explain the differences. In our case, the system in question is baking cookies for consumers. Scalability, is how well something responds to adding more resources. If we can double our capacity by doubling the number of resources, we've got linear scalability. In our baking example, if we add more ingredients, we can scale to a larger number of cookies. To contrast with our other metrics, this doesn't actually increase the performance of creating a single cookie (they still take 10 minutes to bake), and the efficiency is the same (we just have more capacity). We also cannot bake more cookies in a fixed time period, but over a larger time period we can produce more cookies. Performance is how fast something can get done, u
Read more