7th March, 2021

Week 10/52

😝 Smile

Here is a fun project. Let’s call it Human Emoji.

It’s super easy: just setup a camera on a tripod with a plain background and recreate each of your favourite emojis in turn.

I wonder if these could be turned into an actual font?

Anyway, try it out. Send me your best examples!

🏷 Define (part 3)

Technology is anything that was invented after you were born, everything else is just stuff

Alan Kay

Technology is everything that doesn’t work yet

Danny Hillis

Let's return to our occasional series on building an ecosystem of innovative technology startups in New Zealand.

For previous instalments see Part 1: “Startups” and Part 2: “New Zealand” ...

Next up: technology companies.

One of the things we like to say, when we talk about technology companies in NZ, is that the tech sector is the third largest exporter, behind dairy and tourism.

Actually we say that over and over and over.

This has been repeated by so many people (even the Prime Minister rolled it out recently) and for so long that it's now accepted and believed, mostly without any reference to any source.

I think the original source is Technology Investment Network (TIN), which is a “quantitative study of New Zealand’s leading technology export businesses” published each year since 2005 by a private company, and sponsored by a range of public sector and private sector organisations.

I have a long list of problems with TIN, which I have screamed into the void about on Twitter for many years with little to no impact (surprisingly!) Those who are sadistically inclined can read the footnotes1, but for our purposes here let's just focus on one: their definition of “tech sector”.

Q: Who is the tech sector in this context? What is a tech company?

When you look for it using official Stats NZ data you discover that “tech” is not actually an industry sector at all. Instead they group companies using the Australian and New Zealand Standard Industrial Classification (ANZSIC), where the nineteen top level categories cover Accomodation & Food Services to Wholesale Trade. The closest we get to technology is Information Media & Telecommunications. If you want to really geek out, scan the full list of level four categorisations, which covers everything from Hydro-Electricity Generation to Whiteware Appliance Manufacturing, and ask yourself which of these now use (or should use) technology as an important part of their businesses (spoiler: it's nearly all of them).

So, to get to the “third largest” headline numbers TIN combines companies from across these industry groups, including IT services and support, software development, high-tech manufacturing (we'll talk more about high and low tech in a minute) and biotechnology.

We can argue about what should and shouldn't be included, and different organisations producing reports will choose different combinations2. But, either way, the point is that it's all somewhat arbitrary. Unfortunately it turns out saying the tech sector is the third largest is not that different from saying that NZ is “100% Pure”.

It gets even more complicated when you consider that the sector celebrates together each year at the Hi-Tech Awards. You may ask, how high does tech need to be in order to be high? And is a product that is fundamentally a multi-tenanted database with a web front-end really very high at all? Those are good questions!

Maybe a better way to approach this question is to consider the business model or what is the product or service being offered, how and to whom?

Let's look at some examples...

Consider an organisation that employs a team of engineers and designers etc to develop a software platform that is sold on a subscription basis to other organisations. This is what we call a software company these days, and most of us would consider this to be a tech company. The raw materials are software, the product that is sold is software and even the method of delivery is software. Easy.

Now consider the same team but rather than building a platform that's sold to customers they develop in-house tools to run an online store and behind-the-scenes retail warehousing system. Are they still a tech company or are they now a retailer?

What about the same business but without their own team building the tools. If they use third-party software but run their business entirely online are they still a tech company or still a retailer. And what of the third-party team that's contracted to build those tools? Are they a tech company or a services business?

What about if the third-party doesn't actually build tools at all but instead provides technology services - e.g. processing payroll, running data centres or providing customer support (note: one of the biggest companies in the TIN index is Datacom and this is a simplified description of their business).

It gets even more blurry when you consider companies that don't employ any engineers or designers at all, but who use technology at the core of their product offering - e.g. a business that sells photo and video packages so those who are brave enough to bungee off a bridge can relive the moment and brag about it to their friends. Are they a tech company or a tourism business?

Those are just software examples. We can go down the same slippery slope with manufacturing and biotechnology. If you develop robots to grade and sort your fruit are you a tech company or a horticulture business? If you purchase these robots pre-designed from overseas and then assemble and sell them here are you a tech company or an importer/distributer?

Describing most of these examples as “tech companies” makes about as much sense as saying that your local cafe is an electricity company, because that's how they power their coffee machines.

Start with the customer and work backwards to the technology3

At this point it's tempting to push for a narrow and pure definition (e.g. only those companies that develop software and/or hardware and sell it themselves?) But, what use is that?

I think the correct answer is to use a much broader definition: every company is a tech company. Or, at least, every company could and probably should be a tech company.

Having engineers on your payroll is nothing to celebrate, in and of itself. Although maybe not having any should raise an eyebrow.

If you run with that logic, there is no technology sector. There are just companies (and charities, social enterprises etc) all applying technology to real sectors.

Once we can accept that it's mostly nonsense to try and define what is and isn't a tech company, then we can stop putting our energy into trying to draw a line around "the sector" and reaching for participation trophies and instead focus on how technology can be applied to make every organisation better.

A much more interesting exercise is to consider what we mean by “better” - i.e what is it is about those companies that we currently call tech companies that we actually like? This brings us back again to the same question that we have and will continue to weave through this whole series:

Q: What do we want?

We will hopefully get to an answer eventually. But next, we need to innovate, baby!

🏏 Doubt

Photo by Alessandro Bogliari on Unsplash

The intersection of sport and technology is my happy place.

It was, for example, a delight to watch the Australian Open tennis this year and see the Hawkeye technology has advanced to the point where organisers were able to replace the fickle line judges and clunky challenge system entirely with automated line calls and replays that didn’t seem to delay play at all.

But there is one example that irks me, where nostalgia seems to be desperately hanging on: the “umpire’s call” in cricket.

To explain for those who haven’t been paying attention: if there is disagreement over the umpire’s decision about an LBW appeal, either team can now ask for a review. At this point a range of different technologies are used to replay the ball to allow the third umpire to determine what actually happened, culminating in “ball tracking” which used the flight path of the ball to predict if it would have hit the stumps or not. This is a level up from tennis where the technology measures where the ball actually landed. So there is a degree of uncertainty.

Currently if, after all of that, the technology is not able to say definitively yes or no it reverts to an “umpire’s call” at which point the decision goes back to what one likely distracted dude standing 20m away thought happened, having seen it just once in real time.

That’s a nonsense. It means if the batsman was initially given out but when referred the ball tracking isn’t sure then the original decision stands. And that is completely inconsistent with the principle we’re taught as kids:

The benefit of doubt goes to the batter.

Fun Fact: You won’t find the words “benefit of doubt” in the official laws of cricket. This is as close as it gets:

31.6 Consultation by umpires

Each umpire shall answer appeals on matters within his/her own jurisdiction.  If an umpire is doubtful about any point that the other umpire may have been in a better position to see, he/she shall consult the latter on this point of fact and shall then give the decision.  If, after consultation, there is still doubt remaining, the decision shall be Not out.

Emphasis on the not out.

<old man waves fist at clouds />

Top Three is a weekly collection of things I notice in 2021. I’m writing it for myself, and will include a lot of half-formed work-in-progress, but please feel free to follow along and share it if it’s interesting to you.



The TIN report is compiled by a privately owned business, largely funded by public money, asking private companies to contribute their own data for free, so that it can be packaged up into a report that is sold back to them (at $400/copy, more if you prefer a printed version). Ironically TIN have refused to disclose their own revenue, but presumably given that business model it’s pretty lucrative.

Because the data is supplied on trust it's impossible to verify - I know of multiple examples where very round numbers have been supplied and printed, and other examples where no numbers have been provided at all, so they've presumably relied on educated guesses.

Greg Shanahan, who is the founder and managing director of TIN has admitted that the report describes exports only in the "colloquial sense" - i.e. includes revenues that are earned and spent offshore and never touch NZ - see Kate MacNamara’s excellent reporting on this for example.

Even the decision making about which companies are included and highlighted (or not) is opaque - if we think back to the last instalment in this series when we considered what is and isn't a “New Zealand” company and apply that logic to some of the companies highlighted in the TIN index it is ... interesting.

Then, when it's "launched" the price to attend is $55/head, but does include puy lentils.

(╯°□°)╯︵ ┻━┻┻━┻

What's not to like?


For example, in the Digital Nation New Zealand Tech Sector Highlight 2016 report prepared by New Zealand Institute of Economic Research these are the ANZSIC codes used:

The ICT Sector includes the following ANZSIC codes: J542, J580, J591, J592, F349, M700, C241, C2421, C2422 & C2429.
The High-tech Manufacturing Sector includes the following ANZSIC codes: C243, C239, C184, C244, C181, C245, C246, C249, C231, C241, C2421, C2422 & C2429.
The Tech Sector is the sum of the two groups above.

Clear? 🤔

Source: Figure NZ


Full quote:

You've got to start with the customer experience and work backwards to the technology. You can't start with the technology and try to figure out where you're going to try and sell it.....we have tried to come up with a strategy and a vision for Apple, it started with “What incredible benefits can we give to the customer? Where can we take the customer?” Not starting with “Let’s sit down with the engineers and figure out what awesome technology we have and then how are we going to market that?” And I think that’s the right path to take.

— Steve Jobs