It's probably not a secret - I love 37signals . Not only do they build great products, but they have a way of cutting through the clutter and bringing refreshing clarity to often complex issues. They never fail to inspire me and their recent blog post titled The Secret of Making Money Online was no exception. This is a video of a talk that David Heinemeier (of 37signals) gave at Statup School.

David tackled one of the more pressing dilemmas in modern software development...something like:

1. Build great application

2. ???

3. Profit

What miracle happens at #2??? Good question. Some of David's more salient points:

• Having a price helps create profit (it's almost too simple to work).

• This notion has been lost in the web world.

• There are many ways to have a price - not just one.

• If people like your product, they pay you for it and if they continue to like it, they continue to pay.

• This has been a successful business model for thousands of years.

• Building the next Facebook and selling it for 3 billion dollars is an extremely remote possibility. The chance that you might build a decent niche product that a few thousand people like and will make you a decent living - much better odds.

So, David's talk got me to thinking about my original intent for CPO and it's future. My intent from the beginning was for CPO to become profitable somehow - the question was always the HOW. In my perfect world, CPO would always be a free product and earn revenue in other ways.

As you might have noticed, we have been trying some of these options out for the past month. Since early March, we've solicited paid contributions, paid ads and non-financial contributions like monitoring the forum, translating or writing documentation. So far this experiment has had limited...ok, NO success. Now, I'm not expecting miracles here but I genuinely believed that CPO had enough active and loyal users to spin this plan up a bit by now. I'm a little skeptical - looking around I find that others are in the same boat with contribution/ad supported business models. But it's ok - I've decided to give this plan a total of three months to prove itself one way or the other and then I'll move to Plan B.

So I'll level with you and try to be brief. CPO needs to begin earning its own living - pure and simple. I see 2 possible scenarios playing out over the next 6 months:

1. CPO begins generating enough revenue to justify better support, a more robust hosting infrastructure, regular bug fixes and more features. This doesn't have to be a lot of money at this stage - just enough to entice a little help and to keep motivation up enough to get things to the next level.

2. Or, CPO generates no revenue and things stay as they are indefinitely. I can always afford the hosting account but there will be limited support, few if any new features and bug fixes when I get around to it.

I'm not satisfied with scenario #2 mostly because I know it doesn't have to occur. I know that there are some problems with CPO - I know that support has been virtually non-existent for the past several months. I also know that I'm still passionate about turning CPO into a truly great product - but not for free. The key word here is sustainable - we don't need a $5,000,000 infusion of cash but we do need something that will generate revenue next month, and the month after that and on and so on and so forth. Otherwise, everything will stop.

So, how can we remedy this? I'm coming to you, the CPO user community, before making any firm decisions because I feel it's fair that you have a say in how this all goes down. I'll throw out the first pitch...here are some ideas I have for generating revenue:

1. Created a tiered hosted account model. A free version with no support, no file uploads and limits on projects, clients and invoices. A paid ($5-$10 per month) premium account that would include responsive support, file uploads and unlimited number of clients, projects, etc.

2. Offer the source code for download and installation on any server for $149. This would include 3 months of support and free upgrades forever.

3. Offer a combination of 1 and 2.

4. Charge only for the source code and keep the hosted accounts free.

That's about all I can think of at the moment but I'm open to suggestions. I just got done reading some blog posts over at ActiveCollab and it looks like the community really hit them hard for springing a pay only revenue model on a previously free product without any warning. I'm trying to avoid making their mistake so let's all be constructive and try to find a win-win. I trust my users.

One request though...let's stay away from comments like 'Yeah but Joomla/WordPress/Drupal/eGroupWare [insert favorite open source poster child here] give their stuff away for free and still make money.' unless you can provide explicit ways in which they bring in revenue. Otherwise it's just not very helpful. I know there are a number of quality open source initiatives that make money in ways other than selling software - but I think in the grand scheme of things, they tend to be the exception and not the rule. There are, however, numerous examples of shitty OSS that is completely free and worth every penny...I don't want CPO to fall into that camp if I can avoid it.

Thanks for reading and for your thoughts.

Jeff

I’ll admit it…I’m embarrassed that my first blog post in over 6 months is to explain a site outage. I throw myself on the mercy of the court, beg your forgiveness and vow to do better. Now, on to our feature presentation.

So, What Happened?

As many of you no doubt noticed, CPO went offline yesterday somewhere between 12:00 and 1:00 pm PST. It came back to life this morning around 4:30 am. First, I wish to apologize to all loyal CPO users for this inconvenience. I know how earth shattering it can be to rely upon a web service and then one day find out it’s GONE! So, I’m very sorry this outage occurred and I will explore ways to keep this from happening in the future.

Most of you will know this already but I’ll say it anyhow…No data was lost – none, zero, zilch. The database is intact and functioning fine. One of my other web properties, StuffSafe.com , was picked up by Lifehacker on Tuesday afternoon and traffic has spiked through the roof since then. I’m working with my hosting company to determine the exact cause of the outage but my suspicion is that there were just too many simultaneous authentication requests to the database server attached to my cheap-ass hosting account.

“Wait!”, you say. “What does StuffSafe have to do with a CPO outage?”

Well, both sites (and a few others) are hosted on the same account so the traffic spike to StuffSafe brought down everything. (Gasp from the audience and then…pin drop silence.)

“Hmmm, not very good infrastructure planning is it, Jeff.”

Yes, you would be right about that – given unlimited resources or at least a profitable business model, neither of which CPO has right now. What CPO does have in the way of resources is time and the good will of its users, but not cash.

“Ok, enough excuses. How will you keep this from happening in the future?”

I have one idea. Move StuffSafe.com to its own cheap-ass hosting account and if it spikes again, at least it won’t bring down everything else. But consider this, StuffSafe has averaged less than 500 page views per day for almost 2 years. CPO averages about 5,000 page views per day. My other 3 sites combined barely even register since I’m not actively promoting them right now. On a typical day, my hosting account deals with around 5,500 page views which is not taxing in the least.

Short of offloading a couple of sites to other hosting accounts, I’m not sure what else can be done at this point. I’m open to suggestions if some of you have dealt with this problem before – specifically, large traffic spikes on cheap hosting accounts coupled with the inability to afford a more robust infrastructure.

Again, I apologize for the downtime and the heart failure. Thanks for reading.

Jeff

So, the job timer went live on Tuesday night - late Tuesday night - but a full three days ahead of my anticipated delivery date. I'm hoping it's functionality is fairly straightforward so I won't bore you with a lot of details here. There are controls for timer Start, Stop and Reset. You can select a project and a task, add comments and log the hours to your timesheet once the timer is stopped.

Job timer started.

Job timer stopped.

Known Issues

There are a couple of bugs worth noting from the outset. If you stop th timer, the elapsed time will show in the Hours text field. However, if you then navigate to another page or reload the page you're on, the elapsed time will vanish. You can quickly start and stop the timer to bring the elapsed time back but your hours really just need to stay there until you reset the timer. The same thing happens with your selected project/task. Those selections are lost if the page is reloaded.

These are relatively quick fixes but I just don't think I'm going to have time to address them before leaving on vacation for the weekend.

UPDATE: These known issues have been addressed. 

Future Enhancements

1. Integration with the dashboard as a widget - this will also provide integration of the job timer with the Netvibes and Google modules.

2. Port the timer as desktop widgets for Mac and PC and as a Vista sidebar gadget. The gadget will log your time online or offline and synch with your timesheet when an internet connection becomes available.

3. Allow multiple jobs to be timed at once. A couple of users have requested this feature but I'm having trouble figuring out how useful this could be. I mean, I can multi-task with the best of 'em but I can't fully do 3 tasks at a time and feel good about billing the entire time for all three. Anyone seeing something I'm missing here? Please let me know.

I'm also fielding input concerning widget technology. I have to admit that the Yahoo! Widgets platform sounds tempting because it's (in theory) completely cross-platform. The idea of building one widget and deploying on Mac, PC and Vista sounds too good. My concern is the adoption rate of this platform. I don't want to force anyone into downloading Yahoo's widget platform just for a job timer. Any thoughts here?

Another idea I had was building something with Adobe's Apollo platform - but again, there's that whole adoption thing to contend with. So, I'm open to ideas - I'm giving myself one week to field comments and do some research before I decide on a technology so let your voice be heard.

Thanks for reading.

Jeff

First, allow me to apologize for the long delay between posts. I've been trying to be responsive to the forum posts that have been coming in while also picking away at some new features.

Anyway, the whole point of this post is to simply say that invoice editing now works as it should. A bug that prevented an accurate invoice total while editing an invoice has been haunting CPO for the past 2 weeks. I believe this problem has been fixed but, of course don't hesitate to let me know if you experience problems.

I appreciate everyone's patience as this problem was sorted out.

Thanks for reading.

CPO user SebyM was kind enough to alert me to some pretty significant problems with the invoice editing features. Here's what's happening...

1. A user creates a new invoice - everything works just fine at this step.
2. The user goes back to edit the invoice.
3. Invoice edits work fine unless you remove invoice items while simultaneously editing the tax rate.
4. If you remove items from an invoice while editing, the invoice total is no longer accurate and really can't be changed by the user. Even removing all items and starting from scratch won't give an accurate invoice total.

I've looked into this bug today and found that it's a bit of a monster. There is so much code involved and not all of it is really clean. So, my thought right now is that this will take me through the weekend to fix properly. Until next week, don't plan on doing any serious edits to invoices or you may wind up having to delete them and start over.

Ok, now that the confession is out of the way, I'd like to express a tremendous amount of appreciation to a few bug watchers. Yanick Rochon, Maurer Sebastian and Greg Manset - you guys are terrific! Thanks a million for sticking with CPO and taking time out of your busy day to help me identify problems and suggest great improvements.

CPO has grown much faster than I ever could have imagined and, right now, I'm scrambling a bit to keep up. It's a pleasure to have some users who will graciously point out problems without just giving up and leaving.

Thanks so much, guys. I hope I can do something nice for you one of these days soon.

Thanks for reading.

Jeff