I spend countless hours talking to people in the startup ecosystem. If anyone is willing to have a coffee, schooner or a video call with me to talk about technology, it’s usually an hour well spent. It’s a great way to understand what other people are working on and the challenges in their respective industries. So without further ado, let’s look at some insights I’ve gleaned from conversations I’ve had in the last six months.
People are using COVID19 as a chance to re-evaluate
We’ve been fortunate in Australia and New Zealand with COVID19. Even so, it’s changed us. It’s given us time to step back and evaluate what is meaningful in our lives. Self-reflection was a common theme in conversations I’ve had. People are prioritising quality of life over CBD living and impactful jobs over higher wages. Remote/hybrid working practices have allowed tech workers to leave the cities and focus on things they care about rather than spending endless hours commuting.
People are also thinking harder about the companies they’re considering joining and whether their missions align with personal values and goals. I don’t remember this outlook being as prevalent in pre-COVID19 times. Still, people I spoke to are giving significantly more consideration to the question, “is my labour contributing to a better society?”.
For many reasons, highly skilled Aussies are also leaving SF/Seattle/NYC/London and returning home, adding some experienced folk to the Australian market. There are also highly skilled tech workers from other nationalities leaving Australia and returning home. These ebbs and flows have always existed, but COVID19 brought a focus on being closer to family. Whether this is a long term trend remains to be seen.
Funding is abundant
Surprisingly, considering we’re amidst a global pandemic, there seems to be plenty of funding available. Founders are raising significant amounts of capital. VC funds themselves are growing and hiring. Australian startups are completing IPOs. It feels like as good a time as any to raise or join a startup.
On the other side of the coin, it seems like some early-stage founders are more willing to avoid this whole situation and opt to bootstrap. High accelerator percentages and unfavourable VC terms are putting people off. Some of the prevailing attitudes about the necessity of venture capital aside, it’s a viable option for founders who have angel funds and want to get to profitability fast.
It’s also a sign that some founders don’t necessarily want to achieve unicorn status. They want a stable, sustainable business that grows organically on their terms. Bootstrapping is a privileged position to be in and is not for everyone, but Australian success stories like BuildKite have shown the payoff if you can wait it out to raise capital.
The hiring market is tight
The hiring market in Australia for startups has always been pretty competitive, and it’s even more so in 2021. COVID19 has meant a lack of skilled visa applicants to boost the workforce, and the current cohort is pickier about remote-first and hybrid companies.
We’ve always had to compete with the Atlassian’s, Canva’s and Google’s of Australia regarding salary and benefits. But startups are a unique environment to work in, and there are many ways to make your roles enticing to prospective employees.
If you run your startup like an enterprise or an SME, you’re going to have a hard time attracting people. Cut the bureaucracy, leverage every benefit within your means, give people clear ownership of their work and the power to grow their careers.
Also, think carefully about diversity. If you’re excluding anyone, you’re potentially missing out on a great hire. Rewrite your job descriptions to have more inclusive language and point out that you have made significant efforts in this space. Tell your recruiters that you want to see a broad selection of CVs. It still amazes me that companies use the terms ninja, rockstar and gun in their JDs in 2021. If you see a JD like that, bin your application.
Hiring aside, diverse teams build better products, communicate better and have less internal friction. If someone tells you diversity isn’t a priority for them, they’re creating a suboptimal company.
Finding the first technical hire is still challenging for founders
I have a lot of conversations with first-time founders, and the most common question from the non-technical ones is, “how do I get my platform off the ground?”.
Do they cobble together a couple of SaaS platforms and evaluate their business model? Do they outsource an MVP to an offshore contractor? Do they go all in and hire a CTO or similar to build out their team and IP? The answer, as always is, it depends.
When a non-technical founder is ready to build, they often struggle to understand where to start. Do you need someone with startup experience, or will a senior dev from a more prominent company do the job? Where do you find someone willing to work for less than industry rates and a handful of equity to realise the dream? The likelihood is that you can’t afford a specialised recruiter, so it’s down to each individual’s networking ability.
I’ve met great founders on every platform, from Angel and LinkedIn to Twitter and Gumtree. Yes, that Gumtree. I once met a founder while trying to sell a sofa and worked for them for years. Startups are weird, and anyone who tells you otherwise is probably full of shit.
It also primarily comes down to your pitch. Not only do you need to convince investors your idea is viable, but you also need to persuade early hires that the financial hit will be short-lived and that your business has clear, achievable goals.
So, where do you start? Search for people who have some experience and ask for their advice. Startup people are more willing than not to give you some of their time. We’ve all been in this position before, flailing about in the dark and trying to figure things out, and we remember the people who helped us along the way. Buy people coffees, have Zoom chats, ask for referrals and introductions. Build up a network of people who will give you advice and introduce you to the right people. It’s hard work, but it’s the only way to do it.
Typescript & Go are the languages of choice
Two programming languages that came up repeatedly in my conversations with technical founders were Typescript and Go. These two picks came as no surprise as both are startup-friendly languages that prioritise rapid development, low learning curves and relatively solid code.
Typescript brings a level of stability and type-safety to already ubiquitous React/Vue/Node stacks. While I’m not a fan of JS in the backend, the ability to write full-stack code in a single language removes a lot of cognitive complexity. When I say full-stack, I mean it, you can have a Typescript CDK for your cloud infrastructure, Typescript in your backend and Typescript in your current-gen frontend frameworks such as Next.js. While this might not be great at scale, it’s a big win for early-stage startups regarding shipping quickly with a small team.
You also avoid some of the issues with having a distinct frontend/backend/ops split. What if your only backend dev wants to take two weeks off? You want the other team members to jump in and patch an API or deploy a new endpoint to unblock themselves.
Go is my language of choice, and I’m so happy to see it growing and people hiring for it.
The people I spoke to using Go are mainly focusing on ops platforms and data processing. Go has long been the ops language of choice and is embedded into a lot of modern toolchains.
One of the overlooked uses of Go is that it’s a complementary language to other, more established slower languages. If you have a monolith in Ruby, Node or PHP, Go can slot in around it as a series of data processing workers. Do you have a machine learning platform written in Python? Go can handle the grunt work and leave the data science work to Python. Combine that with serverless pipelines, and you can quickly build up a reasonably sophisticated MLOps system.
Go is also really simple to learn. It has a lower learning curve than Rust; it’s less prone to bugs than C and is more concise than Java. You also don’t need to hire Go engineers, you need to employ solid engineers, and they’ll pick up Go in a week or two.
There are still plenty of startups using the usual web framework suspects of Ruby, C#, PHP, Python and Java, and more specialised systems languages such as Rust, Scala and C++, and many esoteric functional languages in the blockchain space. So no matter what your stack of choice is, there’s plenty of startup jobs out there for you. But it feels, in Australia, at least Typescript and Go are in high growth areas with interesting, impactful projects.
Voice and AR are the technologies with the most potential
After the last 12 months of existing on video, I’m so glad to see voice taking off. It has all of the human qualities of video but a lot less of the cognitive overhead. Voice can be anything from 15-minute daily podcast news updates, Twitter Spaces/Clubhouse and home automation to companies providing audio aspects to their products that you can consume on the go. Voice is now my go-to way of consuming information.
Augmented Reality has had a slow start and has existed for years without really catching on. However, the latest technology improvements help the move from fun social media applications and e-commerce to more traditional businesses such as agriculture and construction. Once someone figures out how to build a contact lens or a pair of glasses that don’t weigh a ton, I can see it being a daily part of our lives.
There’s a lot of exciting things happening in the Ops space
DevOps-as-a-culture has a long way to go, and there are many companies out there still not quite managing to break out of the sysadmins-rebranded-as-DevOps culture. It’s heartening, though, to see some early startups leveraging the latest technologies released in 2021 to help them build their products.
I’ve spoken to people fully embracing Amazons new CDK as a replacement for Cloudformation/Terraform and teams rolling out Backstage to improve their engineering processes. Teams are trialling integrated production platforms such as Encore to get off the ground quicker, and people are building sophisticated MLOps pipelines to empower their data teams.
For the first time, I’ve also seen an uptake in true observability in early-stage startups. After years of seeing teams (myself included) install APMs, log aggregators and profilers and essentially forget about them, it’s refreshing to see people care about real-time application data.
A big focus for new startups also seems to be PeopleOps. This industry has grown over the last couple of years and probably has more of a spotlight on it due to remote working since COVID19. If the hiring market is tight, you want to keep employees and make sure they’re happy once you get people through the door.
I’ve been using PeopleOps platforms for the last five years for hiring, team management, 1-on-1s, reviews, etc. HR (or People & Culture) usually doesn’t exist in early-stage startups, leading to an inconsistent experience and misunderstandings between managers and employees. Suppose you’ve been around the startup block before. In that case, you’ll know that it’s better to bed down people processes incrementally rather than have to make a People & Culture emergency hire when you finally figure out there’s a lot of problems and your staff are unhappy. PeopleOps can help you mitigate some of the most significant risks until you have the budget to hire HR.
There is also uptake in DesignOps and DataOps processes in smaller teams. These were traditionally the preserve of more established companies that have the time to build out these systems. Improvements in design tooling and cloud pipeline services have meant that both of these are more attainable for small teams.
DevOps platforms have traditionally appeared in the open-source space, and PeopleOps have usually been in the closed-source space. This split is essentially still the case, but the industry building around open-source as a service continues to grow. Whether this is an open-source core with paid additional functionality such as hosting or a central hub that consolidates and standardises a set of public APIs, this space seems to be growing.
Blockchain technologies still don’t have a clear use-case
Every 12 months or so, I revisit blockchain technologies, talk to people involved, and try and understand what companies could use them for aside from speculative assets. I always come up with the same answer: nothing that can’t be solved a lot more effectively using an RDBMS and with a significantly lower environmental cost. It is overkill to need cryptographic certainty and distributed ledgers for the majority of business challenges. Sure it’s cool to build but unnecessary.
The ideas about democratisation are lovely in theory, but we’ve already seen consolidation in this area. I’m old enough to remember when the HTTP protocol and the world wide web was hailed as the democratisation of information, and look how that turned out.
My best prediction for blockchains, the Ethereum smart contract platform and NFTs especially, is that they certainly have a use, but we’re in the wild west early days. Things will stabilise, consolidate, hopefully, become carbon neutral and eventually become another option in the developer’s toolbelt. There is no utopian dream here, just another pile of zeros and ones that the majority of commercial companies haven’t figured out how to monetise yet. It’s also somewhat interesting that Microsoft is shutting down its blockchain offering and IBM’s is essentially shuttered.
The whole space is still eluding me, but I’m not writing it off. I also haven’t invested a lot of my time trying to understand the individual use cases, primarily because … well, blockchain people. Oh well, maybe in 2022, it will click.
Thanks for reading, and I hope it’s been insightful. Please correct these largely anecdotal observations with some data-driven evidence in the comments.