Archive for the ‘Kalzumeus’ category

Kalzumeus — The What, The Why, and a bit of The When

June 17, 2007

Picture this: you’re a young doctor, 32, recently married to your med school sweetheart.  Life is pretty hectic, what with being constantly tied to a beeper, carrying a cellphone so you and your wife can murmur sweet nothings in each others ears when not on call, and you having all the distractions of normal every day life.  Also, earlier this year you purchased a nice little house in the city to rent out to folks for some passive income, with eventual designs of using it to build up a little equity and maybe give something to the child you expect to be having one of these years, and then it hits you:

Being a landlord sucks.

Every first of the month you get to have even more thrown at your plate, and it is by and large busywork.  You have to check your mail to see if your renter sent you the check.  If they didn’t send you the check, you have to contact them fairly quickly to request that they send you the check, assess a late fee, field the phone call whining about the late fee, explain to the renter that just because their back door is a wee bit off its hinges that doesn’t mean they get to live rent-free, try to remember to call the handyman to fix that blasted door this week, get the check a few days late, drive to the bank, and wonder why on earth you took out a $200,000 mortgage to buy yourself a minimum wage job. 

There has to be a better way, you think.  Why can’t this be as easy as your wife’s hobby selling those Beanie Babies on eBay?  Deal with any questions or disputes by email, get an email from Paypal when the money arrives, no checks in the mail, no runs to the bank, no forgetting to mail the Beanie because there is a big announcement on your eBay dashboard on your computer saying “Hey, you haven’t mailed the Beanie yet”.

Kalzumeus makes being a landlord as easy as selling Beanies on eBay.

(Note: Kalzumeus is a codename and will eventually be the name of my LLC.  The actual product name is quite boring, was chosen mostly for SEO, and will be announced later when that website has some content on it and, ideally, a functioning demo.)

One of the nice things about Kalzumeus being a web app is that features can be added to it fairly easily.  At release Kalzumeus will support:

  • Landlords and renters logging into their accounts.
  • Landlords adding properties, residences, and renters to the system.
  • Landlords billing renters, with automatic repetition, in a very flexible fashion.
  • Billings for rent, parking fees, what have you.
  • Renters paying their bills online through Paypal.  All the landlord needs is a Paypal account, the computer handles the rest.  No messing with buttons.
  • Email notifications of rent due.
  • Automatic bill payment for most renters.  (My systems are more flexible for automatic rebillings than Paypal’s are.  I can, for example, bill someone on the 1st Monday and 3rd Thursday of every month.  I don’t know why you’d want to do that, but “1st and 3rd Friday”, for example, is a typical paycheck schedule and landlords like being able to bill people immediately after they get their paychecks.)
  • Roommates.  (This was a low priority feature, but it was free with an architectural decision, so it made it into the first release.)

What I eventually want to add:

  • More payment options for landlords to choose from.
  • Automatic paper dunning letters.
  • Expense/work order tracking.
  • Reports.
  • “Click this button to print out the stuff your accountant wants from you.”  (Depreciation calculation and all that jazz.)
  • Export to Quickbooks, CSV, etc.

Is this a big market?

There are fifteen million landlords in the United States, according to Intuit.  The most common number of properties owned is one.  (Cue the “If I could get 1% of that market…” song.  Now all I need is a venture capitalist and I’ll have all the ingridients for a funding round.)

Who are the competitors?

On the one hand, we have property management companies, which charge a figure anywhere between 6 and 10% to stand in for you in all or most interactions with tenants.  There are also services which will do ACH (Automated Clearing House) deductions for landlords, which are a way to automatically withdraw set amounts of money from someone’s bank account, for approximately $10-15 a month plus $2 per renter.  Some of these services also provide online reports, although the functionality is fairly limited.  Finally, there are many, many companies which offer “property management software”, which is basically integrated accounting packages set up for landlords.  One I have a lot of respect for is LandlordMax, run by a fellow uISV.  (Before anyone asks, I don’t see him and I as really being in the same niche.)

Why go into a market that crowded?

Because the vast majority of my competitors sell chainsaws and there are a lot of landlords who really need butter knives.  Much of the property management software is geared at professional property managers, including folks who have strong accounting backgrounds and are managing hundreds or thousands of units simultaneously.  I think there is a whole lot of overkill going on for folks who are at the “I’m not so much a landlord as I am a teacher who happens to own a house which I rent to people”  side of the market.  Additionally, the pricing models for many of the existing solutions totally ignore the needs of small landlords, probably because they are not nearly as lucrative as large landlords. 

Pricing Model

I am thinking of doing one month of free trial and then billing folks $X per month.  I have not decided on a final X yet, but am thinking $10 or $15 puts it in line with many of the successful small business web applications.  If folks want to prepay for a year I’ll give them two months free.  The wild card is Paypal costs, which for many of my customers are going to run at essentially 2.9% of their monthly rents.  That is a lot of money for just processing payments (comparable with running credit cards yourself, a heck of a lot more than using an ACH service).  At a $650 average rent in the US, that works out to about $19 in Paypal charges on top of whatever I charge.  I am cautiously optimistic that if I make it easy to use I can justify a price premium from some landlords over ACH payments, which are not exactly easy to set up (contracts need to be signed, faxed, etc), and am extraordinarily confident that for folks who are looking to accept credit cards I can do a bang-up job.

I am also thinking, eventually, of offering a separate $99 a month account type targetted at professionals (management companies and the like).  They’ll be able to create logins for the property owners, who can check the website at any time and see “Ooh, yay, I’m making money”.  That will require a bit of rearchitecturing so it is slightly down the road.

The Shoestring Factor

I have already booked most or all of the prelaunch costs.  While I don’t have my tab in front of me at the moment, they came to a little less than $250 last time I checked.  That includes a year worth of hosting with Textdrive prepayed.  (Just over $10 a month, even counting the setup fee.)  If my hosting plan is inadequate for keeping a Rails application running, which I have been hearing conflicting reports about, I’ll get one of their medium accelerators (basically, a slice of a server) for about $65 a month.  Currently, my break even number of customers is one.  If I end up buying the accelerator, that will put the break even number at between 5 and 7.  (Bingo Card Creator, by comparison, is profitable from the first customer every month to the last.)

Where is the end-game for this?  I don’t know.  My goal for the intermediate term is 300 (paying) customers, and I’m hoping to have say 80 to 100 by the end of the year.   (I generally like to set small and achievable goals and scale up from there.  My first goal for Bingo Card Creator was $200 a month in sales.  Hit that and kept going…  working on $1,000 a month now.)  As scary as this is to say, 300 paying customers would be enough for me to quit my dayjob.  (I love being a uISV.  If you’re doing some sort of advertising funded social networking site, you can have hundreds of thousands of users and still be losing money every month to your massive hosting and infrastructure costs.)

Marketing

Blog, SEO, and AdWords to start out, more as time permits.  I’ve been looking at the market for a couple of months and while “property management software” is an absolute bloodbath I am fairly confident that I can SEO very well for other queries, like “pay rent with credit cards”, “online rent collection”, and the like.  I don’t need or want to compete with the largest players on their own turf — for the moment I’ll be quite happily picking and choosing crumbs dropped from the table.

Demo / Ease of Use

I’m working at having a mostly full featured online demo for the software, part of my usual quest to get folks to the shiny bits as quickly as possible.  I also have no-hassle, no download, no credit card, “Give your email address and get started” account creation.  There are plenty of examples you can find of Web 2.0 companies who do Actual Business Processes without requiring 30 minutes of forms to get started.  I am practicing the most sincere form of flattery with regards to the design of my own signup process. 

In my timer tests (something I often did with Bingo Card Creator — time how long it takes from hitting the download free trial button to when the cards come off the printer), I can sign up and have automatic billing working for a renter in 90 seconds.  We’ll see if I can’t shave some more off of that when the interface is more complete and AJAXified.  For example, I want you to be able to do the most common types of rent (monthly on the 1st, monthly on the 1st and 15th, etc) with about two to three clicks while adding a renter.

Legal

I am not decided on whether I will have a lawyer draft the terms of use or not.  I’m leaning towards “not” — the application can’t kill anybody, and a quick glance around the world at my various competitors shows that companies with Serious Money On The Line are quite happy to just have generic “If you bill your renter for a gazillion dollars and get taken to court, that is Your Problem” disclaimers.  The privacy policy is already drawn up, and looks very similar to the Bingo Card Creator one: we use cookies to track X Y and Z, we don’t sell your information, we won’t spam you, any questions feel happy to ask.

So when will this be publicly released?

I will happily show this as soon as I can without embarassing myself.  Since it is currently black text on a white background with the default Rails stylesheets for errors and much navigation is still accomplished by manually keying in the proper URLs, the app is not quite ready yet.

At the moment, the business logic for the first shipping version is about 80% complete.  The billing system still has yet to be implemented, and I have yet to write the systems for sending reminder emails, etc.  The interface, on the other hand, is only about 15% complete — things link to the right pages most of the time, but I will have to redo most of the views so that they look nice and pretty in Multiflex-3.  I also want to hand-edit the Multiflex-3 WordPress theme so that clicking from the product blog to the product site is a totally seemless experience, which by necessity involves a bit of headache since WordPress is PHP, the product site is flat HTML, and the application is Ruby on Rails.  Yay for getting to tweak the same template three times.  (I might eventually make the static pages served by Rails, but for the time being I know that new Rails apps can be a bit tempermental and if I should cause the app to die I would like people to be able to access the front page and send me a letter about it.)

Of course, since I’m currently winding down my employment contract with the day job and actively searching for a new one, the schedule could get changed at any time.  That is another reason why I don’t want to have a site in front of potential customers until I can have reasonable assurances that I have time to act on things that potential customers tell me.

My Capsule Impressions of Ruby on Rails

How am I finding Rails?  Elegant, but not easy.  I will need to do a refactor or three over the codebase to standardize the way I do some things before I release.  The amount of existing code which can be leveraged for my project is pretty low but when it does exist its amazing.  (You should see the graph libraries.  Wow.)  On the other hand, the community is growing rapidly and as a result the average level of skill in the community (and hence on forums, etc) is not quite so high yet. 

Test driven development has proven to be a lot like going to the gym — I didn’t enjoy starting it but the results are easily visible at the annual physical.

Since my interface is not in anywhere close to its final form yet I haven’t been AJAXing everything in sight, but I have identified a few places where there are major, major wins for user experience using it.  I was also able to replace graphs generated on the server with graphs generated on the client (using a clever bit of Javascript whose name eludes me at the moment), which moves the performance of the application from “fast relative to my projected needs” to “stupidly fast relative to my projected needs”.

Programming Productivity Up By 10%… Yowza

June 7, 2007

Direct AccessI have tried expressing my love for Direct Access before.  Really, its hard to describe until you’ve tried it.  However, Andrea was smart enough to include an automatic logging feature in the newest version, which demonstrates exactly how much time it saves you.  (Memo to self: scrape user accounts for Kalzumeus, generate similar ROI-proving eyecandy, and put at the top of the Upgrade Your Account screen.  Its brilliantly simple salesmanship.)  For me, that number is scary.  Lets focus on yesterday, shall we.  Here is a shot of Direct Access on my home computer, which was on yesterday from 7 PM when I got home to 2 AM.  Of that, 3 solid hours was programming time, after doing battle with SVN fruitlessly for far, far too long. 

During programming time, I’m typically opening folders and programs left, right, and center.  Copy this to there, grab the backup of that, nuke those sources, where the heck was my Rails API reference, need to open Thunderbird to retrieve my SVN password, that sort of thing.  Apparently I do this more commonly than I had thought:

Direct Access Stats

I generally estimate 20 seconds to accomplish any random opening from staring at my IDE window.  For example, to get to the Rails API, that would generally be Start -> Internet Explorer  -> Google (my home page) -> Rails API -> I’m Feeling Lucky.  (Why type as opposed to using my bookmark?  Faster, less requirement for right hand, gets autocomplete.)  Given the need to take my hands off the keyboard and the lingering pain in my right hand, thats a good twenty seconds.

Well, thats a bit shorter, isn’t it. 

rrrTAB (used to be railsTAB, but I found myself typing it a lot) gets me a command shell opened to my Rails application directory.  sshTAB gets me a Putty window set to automatically log into the server I’m most frequently using.  All the power of the command line and autocompletion… from anywhere.  Bwahaha.  You can keep your OS X, vim, and bash, I’m puttering around in a deliciously iconoclastic Windows XP… at Unix sysop speeds.

I also love the autotext for hammering out boilerplate code… that turns out to be a heck of a lot less useful in Ruby than it was in Java, though, since a) I write less boilerplate and b) I’m not sufficiently well versed in the boilerplate to have prepared it as a macro yet.  In Java, on the other hand,  I know I will have to do intconvertTAB again, writing out a try-catch block to surround a very prosaic String to int conversion. 

try {

replaceMe = Integer.parseInt(cursorStartsHereAfterIHitTab);

} catch (NumberFormatException e) {

// Handle if string was passed by user, safe to ignore if string came internally.

}

The third time I type the same freaking snippet out it goes in Direct Access with a memorable abbreviation.   I have twenty of them for Java now, and a few for Rails, mostly for migrations and to act as quick references for syntax. 

You wouldn’t think shaving a few seconds here and a few seconds there would really matter that much, but the stats don’t lie.  A 165 minute programming session was a 180 minute session purely thanks to Direct Access.  Not bad.  I really wish they hadn’t forced me to uninstall my copy at work — “Well, Patrick, you see, we have a new security policy…”  Suit yourself, boss.  I get paid regardless of whether I’m being productive or not at the day job.  When I’m the boss, heck, time spent clicking is time spent not doing more important things.

If you don’t have something like this… get it.  Really.  Its a no brainer. 

As usual, product placements on this blog are 100% uncompensated.  I like it, I use it, it saves me time and money, I blog about it.  I do not solicit, accept, or envision folks giving me money for anything other than the products I sell.

Spending Money, Incurring Headaches

June 7, 2007

Kalzumeus is getting to the point where I want to start periodic deployments to the Internets to test it (yes, yes, that DOES mean I’ll be able to announce it publicly, sometime this month if development doesn’t hit snags), so I went shopping for a hosting service.  I eventually settled on TextDrive, which appears to have a decent reputation for Ruby on Rails hosting.  $124 later I was the proud owner of a new “startup” hosting package for a year.  I’ll be spending another $120 or so later to get a certificate for the domain but for now I don’t have any data that needs https so its a waste.

Sidenote: Its a wonderful age we live in where you can buy 95% of the tools you need to run a business for less than $10 a month.  The founder of Textdrive, a guy named Jason who clearly has some chops in the provisioning enterprise class server deploments department, reckons you should budget 10% of sales for infrastructure.  So, for example, if you plan on getting $10,000 a month in sales your monthly server/colo fees/electricity/etc bill should be about $1,000.  This may well be true for larger installations, but I’m pretty sure you can get by on about $30~$50 a month for $10,000 in sales as a uISV with some care given to app selection.  I’ll break down the math for you some other day, after I have an app and benchmark numbers I can point you to, to demonstrate that I’m not shooting smoke out of my hindquarters.

Anyhow, yesterday was a comedy of errors.  These are 100% my doing — I bit off a lot more than I could chew with technologies I do not have a full conceptual understanding of yet.  Example: I spent three hours trying to manhandle my local SVN repository onto the server.  The textdrive docs got me set up with a new repository in 3 minutes, but actually importing the information to it was a constant battle with Netbeans.  I eventually lost my most recent work to corrupted settings (“What do you MEAN I only have 3 class files?!”) after twiddling stuff with text editors, and had to cleansweep my local copy of everything, then restore from the local repository.  Then I got out Cygwin (bless you, Cygwin), killed all the .svn directories, deleted and remade the remote repository, and imported the newly-checked-out-locally repository into the remote repository.  One more open of Netbeans later and I was done.  Blaaaaah.  On the plus side, both Kalzumeus and Bingo Card Creator now exist safely outside of my hard disk.

Incidentally: its far, far simpler to set up hosting with GoDaddy than it is with TextDrive.  The difference in the level of control you get for TextDrive is night and day though.  I’m eventually going to have both Apache and Lighttpd running on the server, Apache proxying everything and making sure the /blog/ folder goes to WordPress, with Lighttpd putting things to the Rails application.  Try that on Godaddy… if you dare.

Then, despite the fact that it was running rather late, I wanted to get some actual coding done rather than losing a night just to wrangling with tools.  Again, 100% due to my own negligence, I lost a lot of time debugging why a transaction wouldn’t work.  Well, OK, 98% to my own negligence.  You see, Rails takes an awful lot of syntatic shortcuts which are deeply meaningful but documented in, I kid you not, exactly one throw away line in The Book.  And, unfortunately, the emphasis on “create code whose intent is crystal clear” gets lost in some of these shortcuts.

Example: build vs new vs create.  One throwaway line will teach you that create actually creates the object in your database while build/new just create it in memory, a sort of Key Distinction when you’re deciding “What do I need to guard with a transaction block?”  And, for more fun, build and new are interchangeable… except when they’re not.

@user = User.build #OK

@user = User.new #OK, same as above

@new_friend = @user.friends.build #OK

@new_friend = @user.friends.new # No new method on friends!  Whoops!

@new_friend = @users.friends.create # Works, but note different database behavior!

Anyhow, I will soon be establishing a blog at kalzumeus.com (you could visit there now, but its a fairly boring “coming soon” page that was autogenerated).  That will be getting most of my future programming/business oriented and professional development posts.  This way I can keep the blog for the actual product site focused on the customers rather than on myself or an interest the customers don’t share (“helping uISVS make money”).  I won’t be abandoning this blog in the near future, though.

Ranking For An Arbitrary Organic Search Query

May 30, 2007

This was posted on May 30th, 2007, Japan time. If it is after June 5th where you are, I predict that I’m pretty high on Google for the search arbitrary organic search query. I know this mostly by construction — I looked at the results before writing this post, observed that the competition is very weak for that query, used those words in my title, and will easily leapfrog over the weak competition.

What does that have to do with microISVs? Well, indulge me in a little meandering around the world of SEO for a moment. Roughly a fifth of my traffic at Bingo Card Creator comes from obvious and highly competitive search terms like “bingo cards” or “bingo card maker”. You know how many people have a substantial monetary interest at being #1 for bingo cards? Plenty of them, and most aren’t selling to elementary school English teachers, if you catch my drift. (Ironically, most of us who actually rank highly are. Go figure.)

Then I get another chunk of traffic from less obvious search terms, which I know because I know my niche well. Dolch Sight Word Bingo, for example. The amount of people searching for that won’t make me rich but they easily justify writing a page about it, which pays off month after month.

Then I get a huge percentage of my organic search traffic, about 60-70%, from arbitrary organic search queries. The majority of them are never repeated by another person, which gells with Google’s observation that 50% of search queries are unique (that is a “remembered factoid”, treat it with a grain of salt). Some are simple typos, many of them are natural language searches (“how can i make a bingo card for a third grader”), and then the rest are just unique because they’re… unique (“kasmir pulaski day spanish bingo” — yes, by the way, Bingo Card Creator will meet your needs). I like to mentally think of these as snowflake queries — every one is unique but if you look at a lot of them at once they certainly look a good deal similar.

I have actually been looking at snowflake queries and doing some work based on them. I’ve been doing some minor optimizations to my website for months, gradually including more content (which has a tendency to grab snowflake queries just because educated writers use synonyms and from Google’s perspective third grader != third grader != grade three != beginning English student) and adding in specific vocabulary which I wouldn’t use naturally but which my searchers do for whatever reason. For instance, some people call bingo cards “bingo boards”. Who knew? Certainly not me, as I went through my entire life without hearing that usage, but my search logs do not lie. This is the reason “bingo boards” is now bolded on the front page of my website and sprinkled on a few of the sub pages.

However, I had a bit of a brainstorm recently: this sort of optimization is nice and demonstrably effective, but what would happen if you took it to the next level. The trigger for this was when I wrote the blog post Increase Your Software Sales, I mentioned that it would rank pretty highly for “increase software sales”, which would be a nice thing if I cared about that keyword. When I said that, it was mostly a minor boast which I thought of little importance in the scheme of things. But it sent me to thinking:

1) Hey, wait a second, I can rank for a snowflake query with a really trivial amount of work. Put query in title, use in body text, don’t spam, done.

2) I have pages and pages of snowflake queries. Many of them have strong commonality in either words or theme.

3) These queries make me money. Snowflakes account for more than half of my organic search conversions.

And this got the wheels of my head turning. What if, instead of doing the haphazard optimization to grab some of the words in these queries that I wasn’t targetting already (like “bingo boards”), I just data-mined the bejeesus out of the suckers. Say I found 100 strings from there that were reasonably close to each other, distilled that down into 5 main words and 5 supporting words or variations, and then wrote my next resource page or blog post about them. Why, that page or post would probably rocket to near the top of 100 queries. That is worth pure gold, since people will write dozens of minor variations of each of those minor little snowflake queries. And my page or post would suck them all up in one big snowball of goodness.

I was briefly very, very excited about the idea, and started working on a gawk script to start clustering my snowflakes. (Incidentally: by training, I’m a natural language researcher. I know this to be a hard problem and yet hacky solutions to hard problems are fun for me — thats why I got into natural language research in the first place.) Then I slapped myself silly, figuring that somewhere on the Internet somebody smarter than me has already had this brainstorm and developed the same tool. I should really pay them the money for the tool and spend my time actually writing the text which will clump up the snowflakes (which only I can do, since I’m the guy who presumably has the domain knowledge) rather than reinventing a solution to a Certified Hard Problem and then using it to squeeze out an extra sale or two of a $24.95 app a month.

Anyhow, after a bit of searching, it turns out that the guy who already solved this problem made a webapp called HitTail. It has the broad thrust of the features I wanted: tracks what queries get people to find you (unnecessary, I can do that already), and then selectively picks queries out which (the site claims) hit a cluster of snowflakes and are not currently very competitive. I’ll be taking it for a test drive this week.

This is of particular interest for me for my next project (Kalzumeus, for regular readers of this blog). It is adjacent to a market space which is extraordinarily competitive and has many established firms with Big Budget$ To $quash uISVs. I don’t see them as particularly competitive for my niche but I do see them camping on some major keywords (both for organic search and AdWords). Time to go around the obstruction rather than running straight into it. I think I’ll see how far I can get with optimizing for snowflakes, well, once I have something to optimize for at any rate.

April Stats

May 9, 2007

(Images in this post may be truncated by Wordpress.  Click to see in a new window.) 

Oh, yeah, April happened.  The capsule summary: week-long holidays like Easter hurt like crazy — I had exactly one sale during Holy Week.  We’re also fast approaching the end of the school year and I can feel orders slowing down quite a bit from their usual clip in May, so I wouldn’t expect awesome times ahead.  On the plus side, I’m guaranteed to continue covering costs and will probably eek out enough profits to continue bankrolling Kalzumeus.

Sales: 26 (no refunds, 4 CDs — memo to self, find out why those aren’t selling well anymore)

Gross Income: $668.70

Expenses:

GoDaddy: $7

E-junkie: $5

CrazyEgg: $9  (technically only a sliver of that is in April but, eh, who’s counting)

Paypal: $12.50

SwiftCD: $30.56

AdWords: $61.01 

Total Expenses: $125

Total Profit: $544

Traffic-wise, things were all over the map.  My numbers on Holy Week declined almost as low as they did during the Christmas vacation.  I’ve been trying a few things with AdWords (you might notice from the higher expenditures), some of which worked and some of which did not.  (Advertising directly on a competing site did not.  Constant improvement on the main campaign did.  Content network did, to my surprise.)

Rather than spending time copying numbers by hand from Analytics I’m going to be lazy and just paste what it shows me for the month.  For those of you unfamiliar with the interface: G1, G2, and G3 refer to configurable goals and the percentage of visitors from a particular source who complete them.  In my case, G1 is purchasing something from me (note this understates my actual number of customers by 50%, due to some issues people seem to be having with reaching the page that tracks this), G2 is downloading my trial, and G3 is confirming the installation of the trial (by requesting an update for it). 

Conversion Rates by Source for Bingo Card Creator

I generally make most of my decisions based on the G2 column (i.e. free trials downloaded).  The reason for this is subtle: if a person comes to my site, downloads the trial, closes out of their browser, then runs the trial, and later conversions (such as a purchase or check for updates) typically get filed under “direct”.  They typically use the link in the application to arrive at my site, and from Analytic’s limited view of the world thats the same as typing www.bingocardcreator.com “direct”ly into your browser.

As you can see, my most promising prospects are the ones I get from AdWords.  Organic traffic from Google is the largest segment of my traffic, accounting for about half of it, but their conversion rates are pretty poor.  That is unsuprising if you look at what people are searching for when they get to me — typically I perform fairly poorly in terms of conversion rates on the high traffic queries like “bingo cards” and very well on long tail queries like, well, why don’t I show you some:

Popular Organic Search Terms for Bingo Card Creator

In a very Long Tail-esque moment, I get about 30% of my Google hits and trial downloads from the top 10 search terms.  Its 3:30 AM so I won’t do too much math on what the breakdown is for midlist search terms versus the Long Tail shown above, but my naiive guesstimate is that probably one third of my downloaders are from the top 10 search terms, one third are from less common but still fairly popular search terms, and one third are looking for a phrase that only exists for them.  I could show you pages and pages and pages that look like the above image.

Out With The Old, In With The New (Buttons)

May 4, 2007

I have been thinking of how to create application buttons for Kalzumeus in a pretty Web 2.0-ish style and, on the recommendation of several BoSers, bought Axialis Icon Workshop.  To test it out, I decided to redo my sort of frumpy Buy Now! and Download Free Trial buttons for Bingo Card Creator, and see if the new versions aren’t any more successful.  While you can of course see them at the best place to get bingo cards for teachers on the Internet, I’ve isolated the section of the interface so you won’t even have to click off this site:

The Old Buttons (Created in Paint.NET using an open source button as base)

The Old Buttons At Bingo Card Creator

The New Buttons (Created in Axialis Icon Workshop, using their Web 2.0 Objects Pack, with cropping and resizing in Paint.Net)

New Buttons

(Edit: GoDaddy seems to be undergoing maintenance or something.  If you’re getting slow as heck performance grabbing those images, that is also affecting my website proper.  Yay.)

I Love It When People Fix My Problems

May 1, 2007

One feature that I had in mind for the 1.1 release of Kalzumeus (does it even make sense to have a 1.1 version of a webapp?  Well, you get the general drift) was the ability to send a real, honest-to-God postal mail in an automated fashion.  It is a requirement with legal significance for some of my customers, and all of my customers would see the ability and go “Oooh, that would save me so much hassle”.  I have been searching for a good way to do this for a while.  Obviously, doing it myself would transform me from a highly-paid software developer to a low-paid postal clerk, which is frankly stupid when there are plenty of businesses out there which offer ways to do it without ever having to lick a stamp.

I was originally planning on using the Post Office’s API.  They have a service called NetPost which is aimed at direct marketers.  The prices are pretty high for one piece ($2.25 or so) but reasonable if you’re sending hundreds, I guess.  I used to use this any time I needed to send a letter to America which didn’t need a physical signature on it, since it costs about the same as sending a letter internationally and gets there a week earlier.

However, cost issues aside ($2.25 is dwarfed by the number of sales this will get me, and I can probably charge extra for it for customers who use it on a regular basis), I really hate having to learn another API.  It is one more thing that can go wrong with my application, and the consequences of a letter not being sent when my app reports it sent are rather severe.  Then today I ran into a site called Postful, which has a much more convenient interface — send them an email, they send out a letter.  You preload your account in advance and after that its $.99 a letter for one page, which is where all the letters I’d be sending are.

Composing an email to my requirements is trivial, since I already have the exact same functionality elsewhere in the program to deliver this notification via email.   As a result, this will probably get into the 1.0 release of Kalzumeus.

As an extra bonus, I can now save myself $1.25 the next time I have to send out a letter to the States.  Yay.