Tag Archives: Leading software teams

The Top 5 Rookie Mistakes in Software – #3 Techno Hype

technology.jpg

The Technology Dream

It’s just a tool, not a Strategy

It’s amazing how often technology gets confused and spun as a marketable feature. It’s not uncommon for companies and governments, for instance, to state that they will “Leverage Technology” as a business strategy. In the same breath, they’re also going to reduce unemployment, clean up the environment and fix the economy. Sure they will.

There’s an exciting appeal to ‘technology’.  There’s still a mystery and newness to even the word that elicits a dreamy promise of endless possibility.  It’s impressive apps like Google Earth, Halo 3 or the iPhone that people likely think of when talking about technology; apps that look good, sell well, and entice people to get lost and obsessed in them.

Sexy apps, however, are built by clever people with a focused vision around capability, content and a dreamy interface. The fact that they leverage technology is almost incidental. It’s easy to get caught up in the excitment of the dream and forget that business is still business regardless of the tool you use.  The truth is, it’s not uncommon for both business people and technical people to miss the point that technology is only a tool. Without a purpose, that tool is just a means to an undefined end.

Without establishing a direction and purpose for a software product that is focused on boring old things like: User, Market and your Business; your technology dream will simply deflate into another expensive mistake rather than a hot app.

The Framework Dilemma

As technology professionals, we get caught up in the dream too.  I’ve been guilty many times of approaching business problems with a technical solution without stepping back and wondering if it’s truly the right thing to do. It’s easy to get wrapped-up in the world you know and use that as the solution to all problems. “Have you considered building an Application Integration Framework?” – Those were my words. It seemed like a reasonable thing to suggest. This guy’s media company had multiple products that couldn’t talk to each other and ‘Integration’ was high on his agenda.

“NO! NO FRAMEWORKS!” – Was his response.  He even stood up out of his chair and leaned forward towards me pretty aggressively when he said it.

Wo, that guy was jaded. But I understand why now.  He quickly calmed himself and explained to me that a consulting company had sold him the same line a couple years ago. They spent those 2 years in, what I call, “a Dark Place”, just building a Framework. They didn’t consult with them about what the products do, or what the user needs, or business drivers, or the data or how this could tie into their overall technical strategy – nothing like that. They just built a framework.  It took 2 years of nothing tangible, but these consultants were finally kicked off the property. All there was to show for the $1.2 million spent was a bunch of code that had no business value.  I’m sure it was a brilliantly designed framework, but it didn’t do anything. I’m also sure there are thousands of skillfully composed frameworks out there sitting lonely on subversion servers around the world. I’ve personally participated in a framework or two that went nowhere. It’s a common thing for developers to want to do.

targeted.jpg

I don’t want to dismiss the importance of a proper architecture and framework; all software applications need a solid foundation. But they have to be built to fit the business need, and with a specific user-driven purpose.  Frameworks should be small, lightweight and grow with the product. Key to making sure that happens is: every aspect of a framework has to have a consumer. If someone on your team is building a component or framework or utility in anticipation of the potential of it being useful, that effort should be denied.

Technology For Technology Sake

My old boss used to have this tendency to wander into my office and recount stories about his past. He didn’t have all that much material, so his stories got repeated a fair bit – I didn’t have the heart to tell him that of course.  Anyway, one of his more riveting stories I got to know well, was about the specifics of a product idea of his that failed. It was essentially like this: a data service architecture that could connect to multiple databases; a domain layer that sat on top of that which reconciled data from these different locations; on top of that was a communication layer based on web services and finally there was a forms builder to create custom interfaces as the top layer of the product.  He thought it was such a great idea and would shake his head in bewilderment as to why it never went anywhere. “Well, what did it do?” I asked him one brave day. “It could connect to multiple databases.” Was his answer.

“To do what though?” I asked.

“Well, to do anything.” He said.

“Anything that needs multiple databases…” I said, to be agreeable; but really I just wanted to get out of the conversation.

So, who’s going to buy that? It’s a thing that doesn’t have any real purpose, except to say it serves a purpose behind an unknown purpose. A body without a head. I didn’t have the heart to tell him, of course, because he thought it was so awesome … and it was kind of his baby. But like a lot of people, he chose to build technology for technology sake, and didn’t appear to understand that that isn’t a product or strategy, it’s only a tool.

Chris Ronak

The Top 5 Rookie Mistakes in Software – #4 The Team

How do I know if my Development Team is any good?


Good Question.
Actually, it’s a critical thing: the importance of The Team can’t be understated. It’s one of the single biggest factors influencing the success of your software.
Those who aren’t seasoned with a) how to hire a good development team and b) know what to expect from them, will likely be faced with a frustrating time. Without more to go on, they’ll use something like the following as their criteria:
  1. They don’t dress very well
  2. They play Team Fortress II at lunch
  3. They talk techie all the time
  4. They cling to each other at the Christmas party
  5. They babble like idiots when a girl walks into the room
If all these are true, they MUST know what they’re doing.

Leaving it to chance is a pretty common mistake.
I had a funny-but-tragic experience working with a guy who’d been given the responsibility to integrate a recent acquisition his company had made (he was the VP of something). He was a super-smart guy, but not that experienced in the world of software. He figured out pretty quickly that to accomplish the IT side of this integration, all he needed to have built was a Map.  A web-based map that a user could log into and find out key land ownership information by zooming around and clicking on land parcels.  Really, that easy.  But he needed it quickly – this was key.  So, rather than taking his requirement to the IT (dev) department, who were notorious for having all kinds of barriers and rules and taking forever to turn something around, he instead gobbled up a couple developers who happened to be kicking around after some acquisition or other. Two guys who fit his profile: they were programmers and at that moment had a fairly aimless list of things to do.
They’ll do, he thought.  A developer’s a developer, right?

time

So he gave them his map vision, and to his amazement, they had something up and running within days.  Days! This was gold!  He couldn’t wait to give his status update to senior management.
Flip to the developer’s perspective now: they wanted to secure their jobs.  So to get some quick sizzle to the VP, all they did was download an evaluation copy of ArcGIS Server and ran a tutorial with some sample data and sample shapefiles.  They loaded the server part, clicked next->next->next on all the various wizards; and splammo, things were pretty much up and running.  They couldn’t wait to show him their status update: they’d basically got the nuts & bolts of his Map working.  He, of course, was over the moon.

I’m going to skip a few details and flash forward a bit…
A year later, my VP friend still didn’t have his map. A simple map to enable users to click on a land parcel and get a quick report.  What!? Why would that be so hard?   I had lunch with him around that point and, the poor guy, he couldn’t contain his frustration, “All I want is a MAP!” he echoed over and over while pumping the air with his fists.
Those two developers? They were re-distributed into the marketplace, and a few new ones had been brought on to take their place. There’s now a full team working this project.

So what happened there?

Self-Organization and The Product Team

Well, lots of things happened. I could tear into this problem – unfortunately, a pretty common problem – from a variety of angles, but I’ll summarize it this way: the developers didn’t organize themselves into a product team.  They didn’t take technical ownership of the situation and instead reverted into being day-to-day programmers, just coding whatever the VP asked for.  And ya, HE was totally at fault too for completely misunderstanding the magnitude of the effort required, and for dismissing the need for proper development splat 2infrastructure. However, in this world we live in, there is an obligation for the technology professionals to self-organize and assume a higher level of responsibility. In their case, they completely failed to effectively communicate, and act on, of the true scope of his simple Map.  They just performed tasks and went home at 5:00.  No Team. No Product.

This type of thing happens all the time.  All. The. Time.  Complete disconnect in communication and understanding between technical and non-technical people:
  • Person needing dev help clutches onto a programmer and asks for stuff
  • Programmer goes “Sure”
  • Person is pacified
  • Programmer is thinking, “Task”, but Person is thinking, “Commercial Product”
  • Ok, person is really thinking more like, “Product that makes serious money or makes me a star.”  Programmer is thinking, “Maybe I can finally use some Windows Workflow for this.”
  • Disaster. Everyone is sacked
The best designs and successful software products emerge from high-functioning teams.  The developers themselves don’t necessarily have to be superstars, of course.  As in most anything, even a group of strong players can fail badly if they don’t work as a team.  Similarly, a team of average programmers who communicate well and build trust & respect and a team atmosphere, are more likely to succeed than a group of aces who fail to interact as a team.

The point I’m trying to make is; to determine if your dev team is any good, I’d worry less about whether the individuals are champion programmers, but more about how they organize as a team.  This goes for small teams, large teams, small companies and big companies.  It’s really all about the team and the fact that they recognize that they’re not simply completing tasks, but building a commercial software product that has a customer at the other end of it. There’s no question you do have to have good people who know what they’re doing and know how to get software to market, but it’s the team that builds great software.  To that point, the aim in this article is to present some ideas on how to evaluate a software product dev team.  So, the discipline of ‘Hiring Good People’ is not really the topic being addressed – that is an art and a science, and a different conversation.

What to Expect From The Team

Still not sure how to tell if your dev shop is any good?  Below is a handy list of 6 quick tips to help you figure that out.

As an outsider looking in, you might struggle to know what they’re supposed to be doing on a day-to-day, week-by-week basis.  Don’t just cross your fingers and hope for the best – you’re better off knowing what should be happening.Emergence

I’d encourage you to have a look at an earlier posting on Software Failure that contains key recommendations for processes your team should be engaged in.  The following points – which are aimed more at someone overseeing a dev group – will provide a quick indication of whether or not a development team is working effectively.

  1. The software should always be working as an integrated product. You should be able to see continuous measureable progress in the product. It should move forward every day and every week.  You shouldn’t have to go to a developer’s machine to look at software in progress.  In other words, they should be ‘releasing’ it on a daily or weekly basis.
  2. The team should be transparent amongst themselves and to all outside stakeholders. No secrets, no dark places.
  3. The team should be measuring their own progress.  They should metric themselves by estimating on sprint tasks then, after the sprint is complete, reflecting on what was accomplished. Each team member should present/describe what they’re going to do, and then explain what they’ve done
  4. They should be continuously forthcoming about risks, scope, timing and any issues that affect the product or release
  5. They should be taking advantage of productivity tools and methodologies to ensure a rapid turnaround
  6. They should have an obvious desire to innovate

Let me know if you have any questions about any of these.

This isn’t a one-sided relationship – Obviously the team has many expectations from outside stakeholders as well. That is coming in another posting on rookie mistakes.  Stay Tuned!

Chris Ronak