Age of Automation

Thesis:

• We are entering the "age of automation" where software will be hired instead of people. The way businesses view the work being done by their employees is changing. This is an opportunity.

• Many of the high value specialists employed by businesses are valued and paid more simply because of their temporary rarity. Fixing the supply of these skill sets will result in a normalisation, and in many cases a deflation, of pay.

• The truer measure of value is the automatability of a given role. The more interchangeable the people doing the job are, the more likely a role is to be automated. This automation doesn't strictly require AI - many jobs are very automatable through interchangeable human labour and software based configuration and management systems.

• Business owners are beginning to realise that instead of running think tanks, they are running factories. Raw material in, efficient processing with labour, product out. The current labour shortages, and the gap between employee and owner expectations, is already pushing business owners to streamline. No owner wants less productivity in exchange for higher costs. This will escalate. As business owners see their businesses more as factories, they are looking to bring down their fully loaded COGS on a role-by-role basis.

• Purchased or rented software tools (built to be used) will be replaced by "hireable" semi-autonomous software powered services (built to be configured, and then started and stopped, requiring no dedicated employee to "use" them).

• This will compound until it is possible to run a very large business as a single employee "configuring" dozens of software services such that they provide the labour for every aspect of the business. Only the extremely specialised and difficult to automate employees will remain directly employed by businesses. All other employees will end up working for specialised service providers.

• The most valuable employees will be those able to design the webs and flows of work, connecting and configuring these automated services.

• A consumer product company worth $1B that has only 1 employee will be created. This single employee will efficiently command hundreds of employees scattered across dozens of service providing software companies without ever talking to or managing them directly.

At SRTX we're on a mission to solve the problems in apparel with materials, technology, and software.

We started back in 2017 exploring why the apparel industry hadn't figured out how to make a pair of tights (pantyhose) that didn't rip. We discovered a world of material science not being considered for apparel. We found a way to miniaturise our first material, produce it at scale, and to manufacture using it. We've done it all in Montreal, in what used to be the textiles hub of Canada, while simultaneously reducing our cost-to-produce by over 90% in just 3 years of manufacturing. We've scaled our tights business to the point where we make the #1 best selling pair of tights in the US, we've created more novel materials, and applied them to apparel categories beyond tights. I love the fusion of material science, modern manufacturing in the west, and practical everyday apparel products made and priced for everyone.

We've been in growth mode at SRTX for a few years now. Doubling our revenues annually, 4x-ing our production, investing to become increasingly vertical, both to maintain our growth rate, and to protect against the global supply chain. This expansion - both in terms of depth: the number of manufacturing machines, the number of stations, the volume flowing through the factory, the number of orders being processed; and in terms of breadth: making our own tights, then our own raw fibers, in-house packaging manufacturing, in-house fulfilment, in-house creative, branding, ecomm, advertizing, we've built 5 innovation teams making software, data, materials, new product categories, and machine automation; all of this - has required a very flat structure, with roles created daily to deal with whatever the problem of the moment was. These hires would then need the tools required to do their jobs. In many cases we ended up wanting to use certain software and hiring employees to "drive" it. Much of this software costs large percentages of the rates we paid the employees using it. I'm a strong believer in vertical companies - they're incredibly hard to build - but they have staying power, and a problem solving spirit that other businesses simply can't match. But our journey has convinced me of an opportunity.

In the future you will be able to build a vertical company like SRTX - owning every aspect of your operations, your supply chain, your manufacturing, of your marketing & sales, your creative, your brand, your advertisement, your customer journey, of your innovation, and your back office - but instead of buying dozens of software packages, each being operated by dedicated employees, it will be possible to "hire" software powered services in the place of these employees, in place of these software tools.

All of a sudden your 6 person HR team becomes 2 or 1 or in many cases 0 employees, supported and replaced by a service, highly configurable, highly accountable, more trustworthy, trainable, and easier to please than whatever employees you otherwise needed to hire. The same will be true for your finance modelling, your accounting, your bookkeeping, your recruiting, your customer support, your sales, your creative production - everything except the most specialised portions of your business will be replaced by a very different feeling service.

Enterprise software sales will change significantly - the only buyers of legacy software will be legacy businesses. The next generation of software will sell the service instead of the tool - which means they will sell it very differently. The best versions of these tools will be configurable in a day - highly trialable, and will cost less than the employee that you otherwise would have had to hire to run some legacy piece of software. The conversation will change from efficiency - how much easier this tool is going to make someone's job - to how many jobs will be replaced by this new service. The goal will be to run the smallest headcount company possible. We will talk about "hiring" software, instead of buying or subscribing to it.

It will become possible to found and start a business with very few employees and offer the same level of service as a much larger business. Similar to the democratisation that happened as in-house mainframes got replaced by cloud services - any business, no matter how small, will be able to offer a full legal function, a full finance function, a full HR function, etc all without hiring a VP legal or a head of human resources.

These new services are fundamentally different from consultants. They aren't telling you how to do the job. They aren't advising. And these new services aren't tools either - though internal to these new companies will be many, many tools - and they will present you the customer with a software interface for configuring how they work, and providing them with the necessary assets and details - but they won't feel the same. You wont feel like you need to open the tool to do work. You will think about them like you think about an employee - that the job is being done because you hired Susan and gave her the accountability.

Prior to SRTX I ran a startup called Upverter where we made one of the earliest cloud based CAD tools. For most of our journey we were trying to make a better tool. More features, faster, collaborative, automating tasks, checking for errors. But as we grew we realised that the much bigger opportunity was in replacing the very engineers that we were building our tools for. Most CAD software is based on a library of widgets and shapes that get added to a design and connected together. Instead of providing tools to import and create new widgets faster and easier - we started creating them in the background, automatically, as you were designing. We went as far as guaranteeing the widgets would work - we took accountability. It felt very magical. Like you had an assistant helping you as you worked. Doing the worst parts of the design for you - so you could stay focused on the big picture.

I've been transfixed by this idea ever since. I think we should have taken it even further. Instead of just a library - why not make the entire design for them? Why even offer a design tool? Instead, provide an interface where the manager of the team can somehow talk about, or describe the job - and the system automagically just does it for them. They wouldn't need a whole team of engineers anymore. They could just configure the work - and it would happen. As if they had dozens of engineers, without any of the frustrations or costs of managing and supporting employees, and without any of the back-and-forth, anxiety or sliminess of contractors. The most valuable thing we built at Upverter was really a tool for us - a tool that would have let us delete entire engineering teams from our customers - we thought the right thing to sell was a software tool - we should have sold an accountable design service.

I think many software tools makers are going to realise this exact thing over the coming years. There are literally hundreds of customer support platforms. Many with little to no differentiation whatsoever. At SRTX we switch to a different one every 6 months. They are all the same. And we've paid as much on these platforms as we do on the team that uses them. I don't want a better tool - I don't really care about the efficiency - I don't really care how it auto tags, or integrates with my other software - deep down, all I really want is a service that does CX for me, better than my people do. A service that actually takes accountability. Presenting my metrics to me. Costing less than the people I employ, and doing a better job. I want to hire software, which I get to configure, without a conversation, to run part of my business, under close supervision, with deep accountability.

My sense is, like we did at Upverter, most toolmakers feel above the nitty gritty of using their own tools. It's messy providing a service. Accountability is much harder to honour than availability.

I think the first wave of these accountable software powered services will mostly focus on the perimeter of the business - the back office functions like legal, HR and finance, followed by the customer functions like CX. Then I think commercial will come - I love the idea of a service which does full creative and digital advertising, they take your brand direction, they take control of your marketing accounts, and they deliver with full accountability delivering against a certain spend, conversion rate, ROAS, ratio, etc. Maybe they take control of your shopify too. You log into a dashboard and see progress and poke parts of the configuration, which automatically trickles out into the work the service is performing. Fulfilment and manufacturing are already ahead of the curve - the interfaces just suck - it needs to be much easier to "hire" manufacturing and fulfilment. I think engineering will come later - it will start with things like what Upverter could have been - and it will evolve into rarefied and unobtainable software developers being obfuscated into the business equivalent of setting up a rails project. True innovation will be last to move. But by then a business could exist with nearly every aspect of the front and back office replaced by services.

• I can imagine a service that takes complete accountability for getting your business grant money. They do 99% of the work. They don't require you to answer millions of questions. They don't feel like a review service. They actually go out and get the grants, they get paid some base rate plus more for success. They have a software that you log into and provide them with information. That's where you can configure how aggressive they should be, which grants to prioritise, etc.

• I can imagine a financial modelling service. Instead of a tool like Abacum or Planful - a service that actively re-forecasts your business weekly. Actively debugging and improving the model. Replacing your whole FP&A team. They have a software that you log into - that's where you can configure certain aspects of the model.

• I can imagine a service like Acquire Agency but on steroids - not only do they run the digital marketing accounts, and make their own content, and constantly monitor and iterate on spend - but they take full accountability for sales targets. They "own" the top line, and the ratios - so much so that they don't get paid if they miss.

• I can imagine a service that runs netsuite for you, doing all your bookkeeping and accounting, without requiring in house accountants to support the out of house ones, without generating a million back and forth questions, without all the deniability of a big accounting firm. Digital bookkeeping as a service, with full accountability, fully customizable through a software interface, responsive within hours of changes.

• I can imagine People Operations & Performance becoming services. Everyone gets paid, everyone gets reviewed, pulse-checks happen and get reviewed and responded to, pay is balanced and fair throughout the org, all without dozens of apps and 6 HR staff managing all these tools.

The tricky bit is going to be around how a service gets "configured" and how they honour the accountability. It's no good if a tool pretends to be accountable, hires the same people I can hire, uses the same software, manages them the same, gets the same from them, and delivers less than they were accountable for. The concept doesn't and shouldn't require AI - people are key - but the service needs to get more from their staff than I could ever. They need to craft their software for their workers. They need to specialise and provide massive leverage. They need to own the accountability - communicating progress, hitting their numbers, charging their monthly rates based on performance. Probably they don't get paid if they don't deliver - certainly they would get paid a lot less. It all becomes a little bit like paying software a commission.

I wonder if Agencies will merge with Software Tools companies. I wonder why that hasn't happened more already. I assume the tools companies value their revenue disproportionately highly, and the Agencies don't know how to make tools.

I don't have a strong stance on the perfect version of accountability based subscription revenue. That said, I imagine a user interface showing metrics as sliders - where moving the difficulty of a given target, moves the commission should the service provider succeed. Some way of de-risking services missing their numbers - I imagine below a performance threshold the service is free. A way to allow rapid configuration changes, and a way to respond to those changes equally fast. A way to scale up as a service provider when every one of your customers moves the slider up a notch - it takes time to hire.

I've been thinking about the right way to talk about these new style software powered services, these "hireable" softwares. Right now I like the idea of something like "fully accountable software" - I like the idea that its still software, maybe only a slider of it gets exposed to the purchaser, but its software all the same - and its accountable, it's not a tool that needs a driver, it is the tool and the driver all rolled into one, it provides reports and metrics on the work its doing, it hits targets for you. I wonder if some company will start using "no tools" like the salesforce "no software" slogan.

If you're building a tool - consider pivoting towards selling a service that's better and cheaper to operate because of the software you're building. If you're working on a software service that replaces entire jobs and functions please reach out, I'd love to invest, to bring you into SRTX, or to help however else I can.