SaaSkatoon

The death of Coghead generated a lot more news than I had expected; perhaps the relative silence in the press about the company cutting most of its staff nearly three months ago after failing to secure late round funding implied that few people really cared about the outcome of Coghead.  But the media attention since Coghead announced that it was fading away, from the Wall Street Journal to TechCrunch, has been significant…so obviously people did care.

The great irony for me in writing this post is that I had actually sat down to write it yesterday.  As a charter partner and ardent evangelist of the Coghead platform, I was tired of sitting back and waiting for the other shoe to fall after  having not received any “official” communication from the company since December 21.  A few partners had read between the lines and moved their apps to new platforms, but I kept receiving phone calls and emails week after week asking the same question: “Do you know anything?”  The fact is that I did know something…I knew a lot of things, in fact.  But I kept them to myself because I had a good relationship with the Coghead folks and was quietly rooting for them to pull out of the nosedive and find a miracle outcome.  Yesterday at 6:30PM, I decided that the miracle wasn’t coming, and that I was going to start strongly recommending that folks start thinking about porting their apps.  The silence from Coghead was deafening, and they were stringing too many people along.

It turns out my gut feeling was fairly prescient: a quick scan of my Inbox uncovered a message – sent just 7 minutes earlier at 6:23PM – with the ominous title, “Coghead service termination notification.”

…

I first came across Coghead in mid-2007.  After having connected with a former Siebel Systems colleague over beers one night, we came away with the idea of turning a piece of intellectual property (what eventually became the Marketing Lucidity Lead Model) into a SaaS application.  I thought it was a fun and challenging project, and because I can’t say no to fun and challenging projects I started working on The Lead Model.  It’s important to note here that the extent of my programming experience was writing BASIC math games for the Commodore 64 for my second grade class, and a couple of programming logic and Java classes in college.  I was not a coder.  Yet I took it upon myself to write The Lead Model in Java and attempt to publish it on a Tomcat server that I threw together over a weekend.  It was OK, but it would never work commercially.  So I started down the path of exploring this relatively new concept that salesforce.com was calling Platform-as-a-Service, or “PaaS.”  The Lead Model got half built in Zoho Creator before I realized that it just wasn’t going to handle the complex logic requirements; I started tinkering with DabbleDB before I realized that the nearly 400 mathematical functions would have to be built in the most unorthodox fashion; Force.com was just emerging and didn’t support the dynamic references in the data model at that time.  So The Lead Model just wasn’t happening…until we happened to stumble across a blog posting extolling the virtues of Coghead.

Coghead was love at first sight.  It did everything I needed it to and more…and it was easy.

Thus began what would be an intense, albeit extremely short lived, relationship with Coghead.  After six months of using the platform, I decided to become a Coghead partner and completely abandon the next-generation IT service delivery consulting practice that I had started in 2006.  The Mikan Group became Delivered Innovation, and we were all about Coghead…all Coghead, all the time.  We released Marketing Lucidity Lead Model in late April of 2008 in conjuntion with the Web 2.0 Expo in San Francisco, and had a full head of steam built up.  But the honeymoon didn’t last long, and by July we were noticing fatal flaws with Coghead the platform and Coghead the company.  Just as the country is quickly realizing that putting the collective hope of the nation on the shoulders of an unproven politician isn’t such a great idea, I quickly realized that putting all of my eggs in one basket – while never a great idea in any circumstance – was a shortsighted strategy.

I loved Coghead.  I believed in the technology and the team behind it.  I was Coghead’s biggest cheerleader outside of those collecting Coghead paychecks.  You couldn’t talk to me for a period of nearly 6 months without hearing at least something about Coghead.  Coghead, Coghead, Coghead.  I built a company around Coghead, investing nearly 4,500 hours of my time and personally going 6-figures into debt bootstrapping the operation.  I hired a small army of interns with a seed investment, and I set out to conquer the world with Coghead.  Unfortunately, by early summer it became painfully clear that things just weren’t going to work out.  By the end of summer it was clear that Coghead was entering a death spiral.  By Christmas it was circling the drain.  In hindsight it was an incredibly short period of time, but in my mind it felt like 10 years.

Delivered Innovation is still kicking, and today we are a salesforce.com partner, and we have ported our flagship application to the Force.com platform.  A second application, Marketing Budget Management, is nearing completion and will be submitted for certification on the AppExchange by the end of March.  But we’re in this position because we saw the writing on the wall with Coghead; many others didn’t, and in some cases I blame folks for holding out for an outcome that just wasn’t probable, and in other cases I blame Coghead for stringing partners along for the past few months.  In any case, there are extremely important and valuable lessons to be learned from the collapse of Coghead, and my intention is to capture these lessons thoroughly and objectively.  What follows is an analysis of events and observations related to the shutdown of the Coghead platform, from the perspective of a fiercely dedicated partner that obviously backed the wrong horse.

…

1. It’s the economy, stupid. [This analysis has been amended to reflect the content of a conversation with Coghead CEO Paul McNamara on February 21]  When I first received the anonymous email announcing Coghead’s shutdown on April 30 blaming the economy for the shutdown, I cried foul because of statements that were made to a select group of Coghead partners at an October meeting at the company’s HQ in Redwood City.  CEO Paul McNamara had indicated that Coghead had commitments from investors to provide enough capital to last through 2009 at a minimum.  Not two months later we were hearing that the company was searching for a buyer.  The connection that I failed to make at the time of this post mortem was that American Capital, an existing Coghead investor, shuttered their venture wing in early December, which contributed to a series of events that left Coghead searching for other sources of capital or a possible sale of the company to a bigger fish that could ride out the Nuclear Winter.  Paul would not have had foreknowledge of this event at the time he addressed us at the Coghead partner meeting.
2. A blurred line between development tool and platform. The Coghead development interface was a sight to behold.  It was beautiful.  You could visually develop applications by dragging form input “widgets” onto a canvas; the widgets in turn propagated the data model, and the visual representation of the form that you saw as a developer was the same interface that was presented to the application user.  Application logic was designed in the same drag and drop manner; common branching and looping operations were represented visually by widgets, and conditions were entered directly into a graphical overlay using dropdown menus and buttons.  The platform that delivered these rapidly developed applications was anything but beautiful; in early July, Coghead experienced a number of lengthy outages that nearly led Delivered Innovation to abandon the platform completely.  The outages followed a major platform upgrade, and it had become apparent that Coghead was sacrificing platform stability for features and functions.  After enough partners complained, that practice was changed and there was more focus made on stabilizing the platform and slowing down the pace of upgrades.  Platform performance was always an issue for me as well, and one of the most frustrating aspects of the Coghead infrastructure was that performance varied from node to node; I could literally execute the same process simultaneously from two separate accounts in two different browsers, and one process would complete in less than five seconds while the other would time out after two minutes.  The lesson learned is that if you offer best in class development tools, make that your focus; if you also want to be in the PaaS business, make sure the platform is on par with the development tools…if it’s not, get out of that business and leave it to someone more capable.
3. He said, she said. Coghead’s messaging was schizophrenic.  To their credit, they finally realized this, but long after we had lost many deals as partners because our customers had completely unrealistic expectations of what was involved in developing Coghead applications.  If you believed what Coghead told you on their website, the development tools were so intuitive that you could just dive right in and build apps in a matter of hours; while this may have been the case for a single-form data entry tool, we were getting requests for proposals for highly complex “run the business” apps that admittedly could be turned around much faster than if we were building the underlying infrastructure from the ground up instead of a leveraging a rapid application development tool, but by no means were these trivial projects.  I have been responsible for the delivery of literally hundreds of  IT projects throughout the course of my career, and I know what’s involved in turning around IT services.  I encountered customers with such completely unrealistic expectations of the costs involved in application development that in one extreme case I quoted a dollar amount that was below cost just to get the business, and the customer was “shocked” by the cost…when I pressed, he was expecting something “in the $500 – $600 range.”  This was an application with 10-12 forms, each one with a series of business rules that were triggered based on a number of conditions specified at the field level.  $500 – $600 for a custom app to run an entire business is not realistic, but somehow this was the impression that the customer was given based on Coghead’s “it’s so easy” messaging.  I finally called the customer to get to the bottom of his expectations, and I was told that he had seen a slide in a Coghead webinar that stated that a partner had built an app for $100 on Coghead after having spent tens of thousands of dollars developing it on .NET.  I was floored.  I remember the exact slide, and I know the partner that was profiled fairly well; what Coghead failed to mention is that the gentleman that built this app is exceptionally intelligent and driven, and spent nearly a month of his own time dedicated to building the app.  To somehow imply that the app only cost $100 ignores the opportunity cost of the developer’s time; if his time was only worth about $0.50 an hour, then that figure might have been accurate.  But his time was worth far more than $0.50 an hour, and that slide ended up costing me countless hours of my own time producing proposals for customers that had similarly skewed expectations based on how easy they though Coghead made the development process based on what they read on the website.
4. Who is our customer? Coghead’s inability to find a successful monetization strategy is not unique; what is perplexing is the way partners were brought in to help drive adoption and revenue, only to be left high and dry in the end.  Coghead marketed to individual developers and ‘Solution Providers’, but never really knew which one to focus on.  And even within the partner ecosystem, it was unclear what Coghead’s idea of an ideal partner looked like…was it a consultant that could bring 20-30 customers onto the platform to use an application specific to an area of domain expertise?  Sure!  Was it a SaaS provider selling application subscriptions to potentially large swaths of customers?  OK!  Was it a development shop building custom situational applications for customers that were savvy enough to specify use of the Coghead platform?  We’ll take them, too!  It’s one thing to be opportunistic, which is perfectly reasonable for an early-stage company.  It’s another thing to have no cohesive marketing strategy at all.  I won’t even go into the pricing issues.
5. Whatever happened with the Coghead Gallery? This was supposed to be the greatest thing since sliced bread, and it went nowhere.  There was going to be this AppExchange-busting marketplace of partner-developed applications, and again the strategy was all over the place.  When the idea was first introduced, the Gallery was sold to partners as the centerpiece of the Coghead experience; within a couple of months the Gallery was being downplayed and the executive that was brought in to spearhead the effort mysteriously disappeared (which became a recurring theme that accelerated in frequency as time progressed).
6. The case of the mysterious disappearing employees. Coghead had its share of communication issues, which could reasonably be blamed on being a young company.  But when key contacts that I spent many hours on the phone with just stopped returning emails and phone calls, or would be conspicuously missing from developer and partner meetings, red flags went up.  Obviously not every departure at a company warrants an email blast, but the way Coghead handled transitions was nothing less than suspicious.  People move on, it’s a fact of life; but when the event is concealed from those close to the company to the point of conspiracy, what are we supposed to think?  It’s not like I’m not going to notice when the person assigned to me as an executive sponsor disappears off the face of the planet, or the marketing person that coordinates an event that I fly 1,800 miles to attend isn’t even present at the event.
7. Don’t volunteer your customers and partners to do your QA. To Coghead’s credit, they did eventually introduce the VIP environment so that partners could conduct application testing on a new platform version prior to production release, but it was long after I had given up on them.  I was finding so many bugs after platform releases that I finally asked bluntly, “Do you even test this stuff before you release it?”  I hate saying things like that because Coghead had so many great folks that I enjoyed working closely with, but after spending so much time getting issues fixed following system upgrades, I stopped giving the benefit of the doubt and started expressing my frustrations.  I will be the first one to admit that a percentage of issues with the platform were because of my own ignorance, but the pattern of instability far eclipsed the stupid user issues.
8. Make Money With Coghead. There never really was much hope of achieving what the now-infamous webinars promised.  I’ll take as much blame for this as anyone, but I thought that my assumptions regarding the underlying business processes of the partner program would be reasonable enough to justify taking the risk in trying to make money with Coghead.  Provisioning new applications and accounts was a tedious manual process, and the “Hey, we’ll bill your customers for you, take our cut, and pass the rest on to you” promise was supplanted by a hastily thrown together Pay Pal solution that created significant costs for the partner and never actually worked.
9. You hang me out to dry. It all ended like the title of a Cold War Kids song…partners were strung along for months after Coghead failed to secure financing a few weeks before Christmas.  On December 6 we received an upbeat and slightly detailed message from founder Greg Olsen regarding Coghead’s status.  CEO Paul McNamara had been rumored to have been let go during the mass layoffs that followed the failure to secure funding, and hearing from Greg directly all but confirmed that (until the conspicuously absent Mr. McNamara resurfaces earlier today in the From line of a follow-on message about the Coghead shutdown sporting the title of “Chairman” – that apparently was enough to fool TechCrunch into thinking he was still actively involved with the company despite what I was told by a number of ex-employees).  This is followed by a December 21 email that contains the highly cryptic message from Greg that stated, “I wanted to give you a quick update on our progress.  I am unable to yet discuss details, but we’ve made substantial progress  toward a positive resolution of our recent financing challenges.  I anticipate that we will be able to announce something next month.”  That was the last email blast from Coghead until the shutdown notice that went out Wednesday – nearly two months later.  Many of us knew about SAP’s involvement, and the thought (hope) was that SAP would keep Coghead alive, but alas they decided that it would be better to just dump the hundreds of developers and partners using the system rather than keep it alive.  So is life.  The least Coghead could have done was be more upfront about their struggles; I had seen the writing on the wall long before the funding catastrophe and had shifted the direction of Delivered Innovation to focus on SaaS applications developed and delivered on the Force.com platform, but as I’ve seen over the past 24 hours, there are a lot of people that were caught completely offguard by this, and that is unacceptable.

…

It is sad to see Coghead go.  It was such an innovative concept.  In the end, execution could not keep up with vision, and financial challenges cut Coghead’s promising life short long before it had a chance to be the game changer it was destined to be.  There were many talented people at Coghead, and we hope they land somewhere that can put their talent to use.  I hope that this assessment of key issues that I observed with Coghead will prove to be of value to others in our industry as we lead SaaS, PaaS, and other important cloud computing technologies towards maturity and widespread adoption.  It is too important for this movement to fail, and to succeed will require near-flawless execution moving forward.

Michael Topalovich
Founder and CTO
Delivered Innovation