The Curse of Transparency

It’s sometimes called “Kennygate” but some who are paying attention know it as that time EllisLab got called to the mat, acted on it and is now paying the price.

For the uninitated, “Kennygate” started with a blog post on October 13, 2010 by Kenny Meyers wherein he laid out several complaints about and suggestions for EllisLab and ExpressionEngine. Most of them were valid points about ExpressionEngine 2 and how they communicated. I once described it as a “ranticle” because it was a rant and perhaps, in my personal view, a little more incendiary than was warranted.

In the end, it didn’t matter how I described it because the community responded and raised their digital hands and pixelated voices in agreement. The outcry from the community was only tamped down when EllisLab responded. Not only did they respond with a blog post but it triggered a series of decisions and announcements whereby they promised improved communication about ExpressionEngine bugs, releases and plans. You could hardly find anything wrong with this response. It was swift, decisive and came directly from Leslie Camacho, the CEO of EllisLab.

But that’s where it began.

Quietly Growing

I started using ExpressionEngine in early 2006, which is relatively recent compared to others who have been around since the pMachine days. My recollection of working on EE in my early days in the community was that EllisLab (then called pMachine) was notoriously quiet and private. Maybe this was just the company inherting what I see as EllisLab founder Rick Ellis’ personality: outside of the limelight and just interested in creating cool things that help people. EllisLab weren’t interested in bragging about what they created. They knew it was cool. Their users loved it. I loved it.

Feature for feature, it is tough to argue that ExpressionEngine wasn’t a giant leap forward for how people created and managed websites. In a time when blogging tools were surging, ExpressionEngine was the affordable CMS that could do more than just set up a blog. It was the prim and proper, powerful CMS that would help designers and developers out of the awful situation of having to cram a website into a buggy, unsecure blogging tool.

March 2008

In March 2008 at SXSW, EllisLab held a special session to demo and talk about their upcoming release: ExpressionEngine 2.0. It featured a complete rewrite of the code using the open source PHP framework CodeIgniter, a colorful new interface designed by Veerle Pieters and lots of eye candy. The group in attendance was excited (I was not at SXSW that year) and buzz online palpable. This was the big coming out party for a company and product that had always lived quietly in its own corner of the CMS world. I was excited, you were excited, the nerds were excited.

The release date for EE 2.0 was set at “Summer 2008.”

21 Months Later

One of the highlights of Kenny’s blog post for me was this part where he addressed the nearly 2 year delay in releasing EE2:

Stop licking your wounds over the EE2 release date fiasco. We get it. Nobody won. ExpressionEngine 2’s release caused a lot of internal and external strife.

That’s right. Nobody won. Not EllisLab and the staff, not the community, not the add-on developers, not the people writing (and rewriting) books and other training materials. That’s why this article isn’t about rehashing that particluar piece of ExpressionEngine history.

In December 2009, EllisLab released the EE2 public beta. This was the first time the public could purchase and get our hands on and use ExpressionEngine 2. Up until that time you had to be part of the private beta or developer preview, the latter of which started earlier in the year (in February or March, if I recall correctly).

EE2 beta wasn’t always pretty but it was a beta.: feature complete and in need of a lot of testing in the field. There were major bugs, some unnecessary use of jQuery effects and a lot of gnashing of teeth in the community over the Control Panel design.

The beta was far from perfect but it was released. That’s a huge milestone. Within days of its release there was a book available on ExpressionEngine 2 and some developers had already migrated their add-ons to work with the new release. It seemed like we turned a corner.

Two Point One

The following Summer, in July 2010, EllisLab released ExpressionEngine 2.1 as the first non-beta version of the software. It still had some issues, but again, it was an improvement and between then and this week—with the release of EE 2.3—ExpressionEngine has slowly gotten better and more reliable.

The response by the community to the EE 2.1 release was to embrace the new version and run with it. Quicker than I thought would happen, people stopped building sites in EE 1.x and moved to EE 2.1. EllisLab, did you notice that? I know you did. That’s a huge achievement.

Today we have EE 2.3, a thriving add-on community with more than 800 EE2 add-ons listed at Devot:ee and a bunch of community websites and services.

The Response

If A equals success, then the formula is A equals X plus Y and Z, with X being work, Y play, and Z keeping your mouth shut. –Albert Einstein

After “Kennygate,” Leslie Camacho responded with a blog post. He acknowleged Kenny’s post and talked about what they plan to do and the problems they’ve faced.

Historically its no secret that we’ve been bad at communicating with the Community during times of growth and the stress that goes with it. That’s precisely why I hired Leslie Doherty.

There’s no arguing (not even from EllisLab I’m sure) that ExpressionEngine 2 was a messy situation. But the community survived thanks to the relentless efforts of the EllisLab team. I don’t write that glibly or as a backhanded compliment. We ship websites. They built the software that let us ship those websites. Let’s not downplay or forget the difference.

The reaction by EllisLab—to somehow atone for their sins of shipping EE2 two years later than originally announced and communicating poorly along the way—was to open up and communicate more but, a lot of the time, overcommunicate. The overcommunication led to speaking freely about ideas that weren’t solid yet, talking off the cuff during podcasts and other public interviews (I’ve had several off the record conversations with EllisLab where project information was shared with me that was still in progress or still in the idea phase. I am not including those conversations as part of my analysis).

As part of an effort to be more open about what the team was working on, they took Kenny Meyer’s great suggestion to set up a status tracker, which listed the current work on EE2 and who was working on it. It was widely lauded and appreciated. Now we’d know if they were working on an oft requested feature or a particular annoying bug we’ve run into. It was our little window into their daily work on EE2.

But then they stopped updating it. It is stuck in an unclear time and lists people that no longer work at EllisLab. The problem isn’t that they stopped updating it, it’s that they create the tracker in the first place. I have never worked for EllisLab, so take this as pure speculation, but I don’t think it’s in their company culture to have public work trackers like that. Isn’t that okay? Yes, I think so.

This is the curse of transparency.

There was the time CCO James Mathias was chatting on the EE Podcast and took a completely inaccurate and inappropriate swipe at a popular ExpressionEngine add-on developer and add-on that undoubtedly made EE a better product. It came just after Pixel & Tonic released Assets, completely overshadowing the updated File Manager (which is now a solid feature in EE2). It came off very poorly but I would guess that the CCO was just trying to have a candid, frank conversation with his hosts.

I wrote extensively about the odd state of the announced yet unannounced EE Reactor project. In what was most likely a gesture of openness by tweeting progress and mentioning it in a blog post (see my posts for more information) it created genuine confusion. And now I’ve written twice about and helped publicize a project that might not ever happen or go past the experimental stage.

This is the curse of transparency.

EllisLab was very public about their hiring of a designer (which turned into hiring James Mathias as CCO). It was a much needed position for a company that didn’t employ a single designer. James is a great hire for EllisLab. In the blog post announcing James’ hiring, Leslie Camacho noted:

After a long in person interview in Seattle, he persuaded me to take the risk and bring him on in such a way that he’d have the authority and leeway to do what needs doing to make EllisLab a user experience focused company, something we started out being but drifted away from.

This note about refocusing EllisLab on UX was lost on most of the community (including myself until I reread the blog post while researching this article), because everyone thought James was coming in to fix the Control Panel issues. In fact, that was the focus initially and announced as such. As far as the community is concerned, the only new design we’ve seen is a community page for EllisLab, new forum badges and some other designs unrelated to the Control Panel. There have been some EE Control Panel tweaks but just not the overhaul that perhaps some expected.

This isn’t a bad thing at all. I’m not arguing that EllisLab is making wrong business moves. EllisLab has an internal agenda that fits their business and product goals. You can’t argue it should be anything except what they want it to be. But after “Kennygate” they’ve been giving the perception of being transparent about their plans and now the community expects it.

This is the curse of transparency.

A New Tack

Almost two thousand words later, we’re finally here. What is EllisLab to do today, right now to break the curse of transparency? This is total backseat CEO advice, but I’ve come this far, so why not go all ten toes in?

First, I’d like to see them go back to only talking about stuff that is solid and ready to launch. In his blog post Kenny Meyers wrote:

The Apple silence strategy works when you release high quality, excellent products that surprise everyone. You released ExpressionEngine 2 beta. Start talking.

EllisLab, go back to the Apple silence strategy. No, your products aren’t as beautiful as what Apple delivers. But few achieve that. In my personal projects, I rarely pre-announce anything. What happens is that excitement of announcing saps the energy to complete the project. Yes, annonucing is exciting but save it!

A favorite article of mine is Joel Spolsky’s Mouth Wide Shut. It opens like this:

When Apple releases a new product, they tend to surprise the heck out of people, even the devoted Apple-watchers who have spent the last few months riffling through garbage dumpsters at One Infinite Loop.

Microsoft, on the other hand, can’t stop talking about products that are mere glimmers in someone’s eye. Testers outside the company were using .NET for years before it finally shipped.

EllisLab, please announce products and services when they’re ready. Not when you think of them and not because you think announcing early will make the company seem like it communicates more. That’s just playing into the hands of the curse.

Being transparent about major bug fixes is, of course, important. And those active in the forums and the bug tracker know that EllisLab is responsive and transparent about bugs, bug fixes and when they will be rolled out into releases. You’ve even updated us on your release schedule so those people who maintain a lot of sites can plan and schedule updates.

As your customers that’s what you owe us. You don’t, however, owe us to expose your entire annual master plan 12 months before you want to see it come to fruition. If you ever feel like you did, I’m sorry. I think that sucks.

The secrecy and announcement of MojoMotor was a great example of not saying anything until it was ready. That was a great unveiling of a new product—despite the awkward timing with EE 2 still in beta—and it generated a ton of excitement.

If you do go radio silent on everything that is unfinished people will still complain. Yes, Twitter will be full of bitching and moaning and plenty of flapping jaws. That’s okay.

Stick with it. Show us your stuff when it’s polished and ready. I can’t wait to see what is coming next.