David Kellam

So, what are we in the Blogosphere going to do about it next time?
http://www.rollingstone.com/news/story/10432334/was_the_2004_election_stole

Read this book:

Get angry. Be prepared.

Steve Chan on Javascript tags:

… put simply, they are fuzzy dice in a car made of php

I liked that one. More on AJAX soon, once The Gatekeeper is up and suitable for public consumption.

An enlightening read: The Tao of Backup

Waterfront IT offers backup & disaster recovery systems consulting.

I wish I knew more about Psychology… then I’d be able to explain why this site is currently sitting on >95% accuracy. It’s only got one guess wrong (out of more than 20) - it thought I was over 50. That probably isn’t too far from the truth, thought pattern-wise

Update:
Here is the list of its recent predictions (since I started copying & pasting them):

If given a choice of boxers or briefs (on yourself or on a guy), you’d choose briefs.

You are very interested in matters of technology.

You prefer instructions when they’re fuzzy.

You are fairly wealthy.

You probably like to take things as they come.

Your desk is a complete mess right now.

The only one there I’d dispute is fairly wealthy, but I’m willing to accept it based on:

• relative wealth
• family wealth
• likely future wealth
• non-financial wealth

Further Update:

It is obviously so accurate because it checks its internal consistency: it reshows each image from a “confused pair” in combination with different images to produce a more accurate comparison. As the images undoubtedly have multi-variable attributes, when used in differing combinations, this probably results in a high probability of successful determination. At a guess.

John says it got less accurate for him over time. Maybe I’m more subconsciously aware of the techniques they use and am modifying my answers to suit??? (e.g. prior exposure to mensa tests, IQ tests, psychometric testing, personality type profiling, tickle.com, etc. etc.)

Kevin’s post (and the links therein) is most timely. John and I were today discussing our strategic approach for our new software (the codename keeps changing) in light of the recent Leopard announcements. Mail is broken and still will be from the looks of things. We’re going to fix it. Anyway, back to the subject of Kevin’s post - .Mac

Aside from the usual stuff about Apple “not getting it” and the silo mentality evident therein (i.e. the fact that certain apps ‘get it’ and others don’t; certain apps have acceptable UIs and others don’t, etc.), Apple could take a leaf out of Google’s book that most people don’t focus on. Google isn’t (solely) about new-fangled management techniques which only work given their cultural context (so don’t try to copy them), dictatorial benevolence and providing a bunch of free shit. Google, whilst being a ‘new media’ company (or is that ‘nu media’? I’m not hip with it), is relying on elementary economics, specifically the cross-price elasticity of demand

In Google’s case, the “good” is advertising. Its complements are many. I read it elsewhere first, but this quote puts it nicely:

Innovation in complements lies at the heart of Google’s strategy. Because Google makes its money by selling internet advertising, anything that promotes people’s use of the internet - from software applications to online auctions to telephone service - is simply a complement to its ad business. It’s in Google’s interest, therefore, to develop and give away as many of those products as possible, or at least to keep their prices low…

From The Sydney Morning Herald (Fairfax).

Apple is doing a fantastic job creating complements for their hardware; it’s what they’ve always done best. Historically, the Mac OS can be seen as a complement for Apple hardware. Sure, you could run BeOS or YellowDogLinux instead, but the value proposition reall ychanged when you threw in the Mac OS. You got a whole lot of stuff, seemingly for NOTHING. For some users, the price really was zero, but for others it was the Mac hardware premium they weren’t otherwise willing to pay. Recently, they’ve been doing it with their iApps. You used to buy a Mac with NOTHING installed on it except a shitty incompatible word processor (AppleWorks - sorry John, anything that didn’t speak .doc was useless even back then) and some trialware. Now, you buy a Mac and the ONLY thing you need to add to it is Microsoft Office (or OpenOffice if you can put up with X11 or Java) and optionally a bunch of free software from versiontracker. Now you have iEverything. Heck, thanks to Bootcamp, Windows XP is now a complement for Apple hardware!

Sadly, it seems Apple doesn’t quite get this. Perhaps they’re caught up in all their iPod success and figure “we can start charging people for everything again! The world LOVES US MUWAHAHAHAHhahha…” (runs all the way to the bank). Now they’re charging people for iLife and iWork and soon iThis and iThat (iThis goes with iThat at iSusan?). There’s nothing wrong with charging for your complements… except when people aren’t willing to pay for them, as is evident with .Mac, and where the increase in demand for your main product more than offsets the loss in revenue from making your complement free. I think .Mac and most likely iLife itself are prime contenders here.

So complements need not be free, but when you have a choice between keeping 3.5 million (at 2002 levels) iTools users or extorting US$99 subscription fees from 1 million complaining Mac users (2006 levels), perhaps it’s time for their business model to change. Think about any other organisation on the planet and what they would pay for even a statistically significant percentage of their client base to advertise their company in every single e-mail message they send. That @mac.com is significantly undervalued by Apple.

In any case, Apple could easily justify forgoing US$100M in revenues each year just based on the lock-in effects of providing @mac.com e-mail addresses for all their users. At least until our new Mac software is launched

N.B. Bootcamp also solves another of Apple’s market share problems: fear. But that’s a topic for another day.

N.B. This argument is a lot more tenuous if you accept that Apple is becoming a software company, in which case you’d argue that the Mac hardware lock-in (Intel chips haven’t changed that… yet) is simply acts as a mandatory complement for Mac OS X. The biggest problem with this approach is that it’s a juggling act between too little and too much functionality in the “included Apps”. Coupled with the evidently low conversion rates between bundled and premium (i.e. iLife, .Mac) software and the low unit price of such software anyway (i.e. in contrast to the “Pro” apps - FCP, Aperture, etc), Apple should make these products free and let the complements (sic) speak for themselves.

Just read Christopher Hawkins’ post on One-Stop tech shops and had to comment. I think he’s mainly right - there’s a serious argument (backed by economic theory and history) for specialisation. There’s also serious contemporary practice for it as companies are divesting their non-core assets and concentrating on what they do best. In a society increasingly intolerant of bullshit (e.g. conventional advertising), excellence in your area is mandatory. Specialisation and focus are the methods.

That said, at the smaller end of town search and transaction costs are significant and change the game.
Waterfront IT touts itself as an IT Consulting company. This is not an accident, nor is it misleading given most of our work is in the ’support’ department. We are aiming to act as a one-stop IT shop for small companies, just as Christopher claims is inefficient. But, if you look at it another way, we are a specialist IT Consulting company for small non-IT-savvy businesses. We just happen to in-source (the make component of the make-or-buy decision) certain functions as a convenience for our clients. I’m sure our web hosting service isn’t (technically) as good as ICO’s, but that doesn’t mean we shouldn’t do it. In fact, there is no tangible difference between a $30 Waterfront IT plan and a $500 ICO plan for someone who just wants a basic “contact me” website and e-mail hosting. In fact, when operating as Consultants (and billing by the hour), the search costs associated with finding another appropriate vendor, negotiating with them, setting it up, liaising with the clients, etc. are significant. We do it, but only after we’ve made an assessment that it’s worth it (to the client).

I started Waterfront IT with the goal of having no inventory, billing close to 100% of my time (it’s a part-time business, so this is more attainable) to clients and not expanding beyond about $50,000/yr in income. There are plenty of good reasons for this: efficiency of billing keeps rates down for clients & profit per hour up for the business; no inventory reduces costs; artificially low capacity allows you to be more selective about clients &/or increase your rate when your capacity is being reached; not focussing (and hence incurring opportunity cost) on internal systems development and business expansion, thus leaving more time for my other businesses and not diluting my focus. I do the work for the client, go home and don’t have to think about it. Not so.

I/Waterfront IT now:

• Has its own testing, internal & production hosting servers
• Is accumulating inventory (networking equipment, optical & magnetic drives, anti-virus software (Trend Micro PC-Cillin), etc.
• Has its own Accusys 76510 backup unit stuffed with 3 x 320GB SATA II hard drives for testing (for a push into small business backup solutions)
• Is a company, with MYOB Accounting Plus and MYOB PocketAccounts for billing (increasing complexity/time but improving financial reporting and accuracy)
• Spends considerable time on quotations (even though we have a no free quotes philosophy
• Involved in business negotiations with several entities for part/deferred payment deals (thus increasing up-front costs/decreasing up-front revenues for the possibility of future risk-adjusted income
• Spends considerable time in the car. Now with operating closer to the production possibilities frontier this represents a problem - time spent in the car is compensated at somewhere between $0 and $40 (minus car costs, which economically eat up almost all of that), whereas on-site time is compensated closer to $100.
• Investing in marketing systems
• About to invest in additional casual staff to cover booking timetabling inefficiences (multiple appointments at once), risk reduction (multiple appointments of variable duration close to each other), service enhancement (holidays, unavailability, busy-ness), timeliness (quicker turnaround time due to task decentralisation)

And more.

These efforts can be broken into two goals: service enhancement and business improvement. My initial goal of an asset-poor consulting approach turned out not to be very realistic for smaller clients. It’s easier for me and cheaper for clients to set up a hosting server and offer people e-mail for $100/year than to set up a separate (and possibly different) hosting account for each client. By doing it myself, I also have full control over what goes on, so when a client is on BigPond ADSL and suddenly finds outgoing port 25 connections blocked on the border network, I simply set up my server to also work on port 1025 (which isn’t blocked). This flexibility allows me to deliver better service at a better cost, even though it’s not my primary business nor an area I’m actively pursuing. In economic terms, hosting is a saturated market, where the expected (market-based) economic profit hovers around 0. Not so in this instance, where the economic microsystem of client–>consultant–>service provider is so inefficient that backwards integrating many of their functions makes sense.

For instance, I now stock Anti-Virus software. Partly because I get volume discounts (but I’d rather not, as cashflow is more important to me: not having my cash tied up in inventory) but mainly because it’s more convenient for the client. I can be on-site and know that if I need it, I can install anti-virus without having to:

• use a trial or NFR serial number (which needs to be changed later, assuming that is even possible)
• ringing my supplier and ordering a serial number (which admittedly is a great service, but there’s still a 15-minute wait period assuming they’re contactable
• go to a store and purchase a physical boxed copy
• download the software and deal with credit cards on a potentially dangerously infected PC

There are issues with stocking it:

• I have to carry it around (or open up each container and note down the serial number and keep that up-to-date
• I am incentivised to sell Trend Micro, rather than any other solution (by virtue of the above efficiencies and the desire to turnover stock)

The second is my main concern. I happen to be satisfied with Trend Micro, above and beyond NOD32, MacAfee, VET, Avast/Antivir and particular Norton. It has excellent anti-virus, OK spyware (I use Spybot anyway) and a customer-friendly firewall. So in 90% of situations, I’ll recommend it. That makes it worth stocking and also unbiased and good value to the client, but the product analysis is not necessarily correct for other products (e.g. wireless routers).

Some equipment, such as the Accusys backup unit, has been purchased to improve knowledge on how things work and to be able to adequately test it in a non-live environment. So I have to wear the cost for that, but it should be paid back pretty quickly on the next backup job. It could be argued this creates an incentive to sell that product. True, but there’s more to it than that. It enhances my knowledge of a particular product, thus decreasing the cost (search costs) of implementing that product in the future. I can and do still recommend different solutions, but these require the consulting approach (requirements –> analysis –> design/implementation/testing/maintenance, etc.) which costs relatively more. So, if I follow my approach of only internalising products & services I would recommend anyway and ensuring I still provide impartial advice, there is no harm done and in fact the client receives better service at a lower cost.

Yes, there are apparent inefficiencies in the above approach. The time and $ I spend on internal systems/testing these products should be captured in future jobs either through charging the same as if I had to investigate from scratch every time or increasing my hourly rate. I tend to go for the latter, but this is difficult to achieve in the early stages. My rate has increased from $50/hr when I was still working casually for $10/hr in the computer shop, to $100/hr when the opportunity cost on my time is probably $40-50/hr for doing something similar. So relatively I’m not doing as well (%age-wise), but I’ve also hit the price points of the competition. So how do I stand out from them? Word of mouth. So far, my rates are lower than I would like (particularly for consulting/architectural design), but the goodwill and word-of-mouth is necessary to create both the demand and reputation to enable me to charge more. I also need to differentiate my services along more economic grounds - market rates for support (e.g. removing viruses) just don’t leave that much room for leeway or leverage. Economic value is higher for networking, Infrastructure setup, data recovery, systems integration, custom development, etc. so rates can be higher. The trouble is convincing people of that. I see three ways:

1. Switch to providing fixed price quotes (where efficiency results in higher effective hourly rates)
2. Wait for word-of-mouth referral and demand to catch up
3. Using viral marketing, blogs, etc.

I guess this post is an instance of the third

Feel free to comment with further suggestions for improvement of the above!

This is how job applications should be. I can show you that one. I can’t show you the applications I received for this position, but suffice it to say most came nowhere near Guy’s second version, let alone his first.

I did lament this fact to a few people, but quickly realised I should not try to ‘help’ applicants who don’t bother (or are unable) to proof-read their applications. Doing so may get them past the first round (The Great Resume Cull), but then they’d simply fail in the interview. So it’s probably better that they quit whilst ahead. Don’t proof-read. Don’t tailor your application to the position. Just SPAM every available position that has some tangential relevance to whatever you think the world owes you as a career. Of course, they may have more luck with an HR drone than a small business owner/operator (and budding entrepreneur), particularly when you introduce political reasons for hiring (as opposed to straight ‘Can they do the job?’ issues).

I did receive one e-mail from an evidently competent person (those who know me realise I am more than quick to use the word ‘incompetent’ on nearly anyone, so take this as something of a big deal) who termed it an “absolute classic”. I was surprised, because I distinctly remember taking out the geek humour (such as, at a bare minimum, requiring applicants to provide the answer to the Ultimate Question - you know, Life, The Universe and Everything, but preferably submitting a proof to P=NP or, failing that, to the Poincaré Conjecture as an attachment) due to character restrictions. Anyway, this guy is busy (understandably - competent people are a) hard to find and b) self-motivated to succeed and hence –> busy), but I do have his CV on hand. I’ll just have to remember to keep it in a separate pile to the ‘wood-yu-laikto-outsorcedat?’ applications. i.e. OUT of my Junk Mail box.

Anyway, I’m currently debating whether to:
a) start a few more blogs and separate by content (www.lifewithmediocrity.com for whinges, blog.hiveservices.com for management & keeping www.drkellam.com for personal stuff)
b) the above, with a metablog containing all posts, categories determine sub-blogs
c) keep ranting & raving here on anything that grabs my fancy
d) all of the above
e) none of the above
f) some of the above
g) logically incongruent sequence of above
h) all of the above excluding the ‘x of the above’ questions
i) all of the above INcluding the ‘x of the above’ questions
[mandatory multi-guess formatting applied]

Homework for Strategic Cost Management this week is (partly) The Lemonade Game! w00t!

I finally got around to fixing my C610’s wandering cursor problem without having to break anything in the process, thanks to these (brief) instructions and these more detailed ones.

Now I just have to convince Windows 2000 it’s still itself and doesn’t really need to reinstall all those drivers…

DRK

Update: Had a client (actually one of John’s that I’m looking after atm) that had the exact same problem! So the ‘free labour’ internally actually paid off

http://www.apple.com/getamac/drivers.html

they REALLY need to advise to look for “mac compatability”, at least (or particularly) on printers.

Even most mac-compatible printers I’ve worked with had something fucked up about their driver. e.g. My current Brother setup utility will only detect the printer when only one network interface is present.
Some MFD devices work fine as printers, but not scanners. Lexmark are terrible offenders - even when they do work on OS X, it’s usually thanks to some bug-ridden driver supplied by the manufacturer’s website (because nobody in their right mind would use the one supplied on CD!)

Their other reasons (on the RHS) are valid, but I do wonder how many people, even Mac users, really use all that multimedia software. Maybe kids these days are being taught skills we weren’t! Still, I bet I could outrun them in an Apple // typing speed contest… ah, the good old days