So, I knew some bloggers make lots of money from AdSense (and other ad systems) and thought this was a business model for blogs and content sites.

Then I read the blog of PlentyOfFish.com owner. PlentyOfFish is (yet another) dating site, only it’s completely free. There are no paid memberships! All revenue is based on AdSense: check this post for a photo of a $900,000 (nine hundred thousand) check for 2 months and this post on how the business grew (in general it’s a very interesting blog).

I find it amazing you can make $5.5 mil / year from AdSense alone (and I guess PlentyOfFish has other advertising programs). I would still hesitate to do something similar, but it certainly shows it can lead to success.

Akismet is a service by the team behind Wordpress, that analyzes a comment and tells you if it’s spam or not.

The corresponding Wordpress plug-in (included in Wordpress 2.0) works great: I have disabled every other anti-spam plug-in I was previously using, and Akismet has caught ~500 spam comments (with only 1 going through).

The nice thing is you can use it for your application (just get an API key and a library for your language of choice). There are 3 licenses, including a commercial one.

Being a service has one big advantage: the system learns from many sources (thousands of blogs), so if a new kind of spam message is used (e.g. one resembling natural language and with few links) the system will learn quickly (because you don’t have to do all of the training yourself).

I’m wondering if the same system or one like it could be used for emails and whether it would be better to use that instead of building a custom solution…

Lately I see some articles on Google’s increased competition. Here are my thoughts:

Leaving Yahoo / Altavista / whatever for Google was easy. Google consistently provided much better search results on all queries. There was no question as to which engine was better.

Now, even if some other search engines (MSN, Yahoo) have surpassed Google in search quality, it’s difficult to verify it for a fact. Search quality among engines seems equal and it would take a complicated scientific experiment to sort them correctly by quality. And since I am already used to Google, I don’t have a reason to leave it based on quality.

The only thing that could make me switch is the various add on services (e.g. image, video, blog, local, map, news, Usenet and desktop searches), but even in these areas Google seems ahead of or at least as good as the competition.

Someone must come up with a new killer feature that will take search engines to the next level, in order for me (and probably lots more) to switch from Google.

I get asked why I’m not going to offer a hosted solution for MagnaCRM. So here are all the reasons that made me decide against offering a hosted solution:

• Technical background. I am a developer, not an administrator. I know how to set up a server but I also understand I would not be up to the task of handling multiple servers, load balancing them, keeping them updated and running.
• Information shortage. While there is a wealth of information (articles / blogs) regarding the process of creating a stand alone app (desktop or web), I can’t say the same for hosted applications.
• Human resources. Although I intend to offer solid support, I have to face the reality: it’s just me. If a client sends an email on Saturday, he may receive an answer on Monday. No big deal. If the DB server would go down on Saturday it would be expected to have it fixed ASAP. By Monday everyone would be rightfully angry and probably looking at the competition.
• Up front costs. You can’t start offering a hosted solution using a shared or even a low end dedicated server. A high end dedicated server is quite expensive and I’m not even sure if one would suffice.
• Pricing. Not sure if this is indeed a problem but I think it as such. A monthly fee is more expensive than a one time fee. Even a low $15 / user / month is $180 / year. As I target small businesses, I believe a $180 one time fee would seem more fair.
• Host it yourself. Just because I won’t host it, it doesn’t mean no-one else will :-) There are hundreds of hosts offering PHP & MySQL, with prices from $5 / month (and maybe even less). I certainly intend to research and suggest some of them, even offering to do the initial setup.

With the release of the Google Maps API (followed by Yahoo), I started thinking how online services will be used in future products and specifically a CRM app.

Lets take the Account screen for example. Apart from the expected things, which consist of static data (like the address, employees, todo lists etc), here are some things that could be offered (probably not yet, but soon):

• Map displaying the location of the company.
• If I am within driving distance, directions to get there (suppose the app knows my position, e.g. I carry a GPS) or public transportation methods available (with timetables). Real time traffic data and nearby parking lots.
• Nearby hotels and restaurants.
• The map should also display other nearby clients (with info windows showing contact info), in case I want to drop by.
• Latest news regarding the company. Information should come from “official” news sites, the company’s web site or blog, other blogs.
• Stock information
• Popularity stats (think g-metrics.com)

Taking this a step further, here is a more advanced scenario:

1. A salesman plans to make a trip to another city, in order to visit a specific client.
2. He starts the appropriate wizard and enters some data. The wizard based on previous similar trips, recommends a few more clients and does an initial schedule (including flight, visiting times and durations, accommodation, eating etc).
3. The salesman calls these clients and makes arrangements.
4. Back to the wizard he enters the final data.
5. The wizard does the rest: flight tickets booking, hotel reservation, car rental etc. Then it prints a nice report containing all info about the trip and about the clients (as in the previous scenario)

We are not there yet, but especially the first example is not that far away.

To tell the truth, the salesman scenario keeps reminding me of clippy ;-).

The bottom line is that available information and online services will start being embedded in software more and more. And that will certainly lead to much cooler products!

For the last 2 years I am using a notebook for all my computer related activities. I still have an up to date desktop computer (or two), mostly for backup reasons (in case something happens to my notebook), but it doesn’t see much use.

Having everything on my notebook is really nice, mostly because I can have all my data and applications with me all the time. For some time I tried using both the desktop and notebook, but synchronizing was difficult and frustrating.

Although the notebook is quite powerful (3GHz, 1GB ram) and has a large display (1400×1050) there is one thing I would love to have: dual monitor setup, for when I am at the office (which is quite often, since we are talking about a home-office here).

I know there are some add-on solutions for this problem (e.g. SideCar) but they are expensive and I am not sure it does exactly what I want. And some notebooks (mine included) support an external monitor functioning as 2nd monitor.

But it’s really awful having dual monitor setup with different monitors! What I would really prefer is this: two identical external monitors. When they are plugged in, the notebook’s monitor should be deactivated. I wonder if any notebooks have this feature (although I doubt it).

While reading “Running your company on web apps”, which is about using ASPs for almost everything (Project management, Bug tracking, Information management, Help desk and things like DNS, Subversion etc), I had some uneasy thoughts about some aspects of this whole thing.

Note, that these things apply to “pure” ASP products, like Basecamp and Jotspot, where you don’t have access to the source code or DB:

• Do I own my data? I guess I do, but can I have it locally whenever I want? Do all these services provide solid, full exports in a convenient format (XML, CSV or maybe database dumps)? Otherwise I am locked with these companies for life, which makes me more than a little uncomfortable.

  Checking Basecamp left me a little disappointed. I can export messages and comments but not my projects, todo lists, milestones (which is the core information stored in a Project management system).

  I also checked Salesforce, a well known CRM provider. I couldn’t find any Export functions, although you can get some of your data through reports (which is difficult and leaves out lots of information, like links between accounts and contacts).

• Usually you pay a per month fee to use these solutions. Which is fine. Until the prices change. And when that happens, you have no option but to continue paying or search for an alternative service (which can be tough - data export/import, learning curve etc). With one-time fee products, you always have the option not to pay for an upgrade or new version and your product continues to work.
• ASP solutions offer the same features for everyone. New versions / features affect all customers at once (which is basically a good thing) but there is the occasional unwanted feature, which you have to accept no matter how much you hate it. Again you don’t have the option to continue using the old version.
• ASPs must be good at two different things: software development and hosting / administration services. Which is not easy.

Of course, ASP solutions have a lot of advantages too, like no installation and system administration, online access (since we are talking about web apps), better prices etc. But I wouldn’t feel comfortable in using one, unless these concerns were somehow addressed.

I think shrinkwrapped web apps combine the best of both worlds. You pay a one-time fee and install the app either on your servers (if you have the know-how) or at a hosting company (which you can easily change at any time). You buy upgrades / new versions if and when you want and sometimes the vendor is willing to do custom work for you, to accommodate a special need.