Eight to Late

Sensemaking and Analytics for Organizations

Archive for October 2007

The fat accomplice and other sneaky ways to sell ideas

leave a comment »

In many organisations, innovative ideas have a much better chance of acceptance if they are proposed by  (external) consultants  rather than (internal) employees.  In the course of assorted stints at various companies, I’ve seen several employees frustrated by the rejection of their ideas by management. Chances are this has happened to you at one time or another. In this post  I’d like to discuss a few strategies for getting those “way out” ideas off the ground. I’ll leave out the most obvious strategy- which is to attempt to sell the idea yourself. I’m assuming you’ve tried this and it doesn’t work or isn’t going to.  OK, so here are a few other things to try:

  1. Invoke the “fat accomplice”: If you are convinced of the value of your idea, you could just do it, much as the popular purveyor of footwear exhorts. Implement the idea or get moving on it without waiting for official sanction. Once the wheels are set in motion, present management with a fait accompli, or “fat accomplice” as my four year old calls it. I quite like the imagery evoked by the aforementioned cross-lingual confusion (or False Friend) –  a fat partner in crime who can take the rap if something goes wrong.  One caveat though, if you are going to call on a fat accomplice, you’d better be sure your idea is going to work. Else be prepared for consequences.
  2. Sell by proxy: Discuss the idea informally with people who have influence in the organisation. These folks aren’t necessarily managers. They could, for instance, be people who have built up personal credibility through their contribution to the business. Although not managers themselves, they generally have more access to decision makers, and thus might be able to do a better job selling the idea.
  3. Give it away:  If you discuss the idea with enough people, someone may appropriate it and sell it as their own. They might have better luck selling it to management. This is a viable option if you don’t mind someone else getting the bouquets (or brickbats) for the outcomes. A common variant of this case is when a manager appropriates the idea as his or her own. Although it is natural to feel a little cheated if the boss takes the credit, you should really be quite pleased. The sneaky strategy has worked –  the idea made it through without people being aware of its real origin.

I’ve tried each of the above at various times with varying degrees of success. You’ll need to figure out which is the best one for your particular situation, factoring in things such as work environment, politics and corporate culture. There’s no guarantee that any of them will work for you. The only sure thing is that if you don’t take the responsibility for selling your ideas, they’ll never to get off the ground.

Written by K

October 26, 2007 at 4:40 pm

Posted in Consulting

Friendly “Fire and Motion”

with one comment

One of my favourite essays on software development is Joel Spolsky’s article on Fire and Motion.  In the piece he discusses how large software companies “pin down” their competitors by continually “firing volleys” of new technologies-  OSs, databases or whatever – into the market.  Competitors are thus compelled to spend time enabling their existing software to work with the new technology, which forces them to take time off from developing new products. With the competitors immobilised playing catch-up, the software behemoths are able to cover new ground unhindered, much as an infantry group advances under covering fire.  

The most successful small companies, according to Joel,  ignore the distractions of new technologies (as interesting as they may be) and, instead, focus on their customer’s needs. 

The key point is: the new technology is used only if it makes business sense to do so.

The Fire and Motion analogy holds a lesson for those who manage technical teams in corporate environments –  project managers,  development managers et. al.  It is all too easy for such folks to distract their teams by involving them in tasks that are, at best, peripheral to the primary goals of the group. An example might be a manager who reads about some cool new technology and then, with no regard for actual need, insists that a key team member evaluate it for future use.  Such managerial behaviour is akin to “friendly fire”, where a manager “pins down” his own team by “firing volleys” of initiatives which  do not add any value to the organization.

As in the case of friendly fire, the net result is an unfortunate one.

Written by K

October 19, 2007 at 7:10 pm

Posted in People Management

Manage multitasking

with 2 comments

The costs of multitasking are well known. Firstly,  it takes time for the human brain to switch from one task to another. This is analogous to the overhead of context switching in computers (the term “multitasking” itself originated in computer science).   Secondly,  multitasking  delays the delivery of all tasks  simply because a given task is not finished in a single go.   

The evils of multitasking have been commented on in the popular press, with CNN, Time and even the occasional newspaper article warning against it.  Despite the general awareness that it is counterproductive,  developers working in corporate IT departments are routinely expected to juggle several tasks in the course of their day-to-day work.  Furthermore, I often find that the  “ability to multitask” is stated as a requirement in job advertisements. For example, here are excerpts from a few IT development job ads that appeared recently in seek.com:

  • Software engineer in Auckland: …Must have the ability to multi task and prioritise…

  • Web designer / Applications engineer in Sydney: … Must be able to multi-task…

  • Systems implementation specialist in Sydney: …Should be able to multitask and work on multiple projects…

The expectation is that employees must be able to juggle  tasks and even projects (!)  as required.  Given this, are there things a frontline technical / project manager can do to alleviate multitasking-stress on their teams?  Yes there are, and here are a few that’ve worked for me:

  • Prioritise work: You’re never going to get everything done concurrently. Given the workload on your team, some things may never get done at all.  It is, therefore, paramount that work be done in order of business priority. Obviously the prioritisation needs to be decided in consultation with the business. Under no circumstance should it be based on technology  (I’ve seen this happen) or developer preference (seriously – I’ve seen this too!).

  • Get a high-level view of everyone’s workloadAs a part of the weekly routine,  get each team member to send you a rolling workplan for the next month or two (I go with 4 weeks). The plan should list tasks that the person anticipates doing each week, along with a time estimate per task. The estimate should be at a granularity of no less than 0.5 days or, ideally, 1 day  (see next point). This gives you a high level view of each person’s workload, and alerts you to potential multitasking problems. 

  • Align task switches with the end of the day or week: Encourage people to work on a single task per day,  or whole multiple of a day (unless it’s a shorter duration task, of course).  This way, task switches happen at the end of a work day, thereby allowing time for the mind to orient itself to working on another task. 

  • Plan on a four day week for development work: Assume that programmers have only four dev days – i.e. days for development work. Keep a whole day aside  for other stuff such as ad-hoc requests, support, meetings and administrative work. Depending on the situation in your shop, you may even have to reduce the number of dev days to three. 

  • Insist on decent work hours: It is OK to work overtime once in a way – when a project deadline approaches, for instance. However, it is definitely not OK to expect the team to sustain unreasonable work hours over an indefinite period.

  • Minimise Distractors : Don’t clutter developer  work weeks with distractors like unnecessary (or unnecessarily long!) meetings. Your job is to remove distractors, not add to them. 

  • Provide Support :As Joel Spolsky mentions in an interview“I always think of it” (i.e. management) as more of a support role – like moving furniture out of the way so they can get things done – than an actual leadership role.” Specifics on how one might do this will vary from situation to situation. In some cases you may need to act as a gatekeeper – filtering requests from the business. In others, you may need to answer high-level questions and provide guidance. In a nutshell – do whatever it takes to help your team do their job.

  • …But don’t interfere: I don’t think I need to elaborate on this. You don’t want to give people the impression that you aren’t confident in their ability to handle their work. Such behaviour is tantamount to micromanagement which, as I’ve mentioned in an earlier post, is a definite no-no.

  • Know when to push back on your managers: Take a stand when your management makes unreasonable demands. Explain why it won’t work, and suggest alternatives. This is a an opportunity to  earn your stripes as a manager (and your team’s respect).

Given the increasing workload and resource constraints that corporate IT works under, there’s often no option but to multitask. This can have negative consequences on team performance and morale.  Some of the techniques discussed here may help avoid and/or manage these consequences.

Written by K

October 16, 2007 at 4:14 pm

%d bloggers like this: