On 30th June 2015, I officially left my job as Software Support Officer in the Department of Computer Science at the University of Manchester, and started freelancing full time the next day. I’m still going ten years later, and this post is a summary of how it’s played out (you can also read my annual reviews for 2020, 2021, 2022, 2023 and 2024).
Achievements
Starting off on a positive note, my achievements include:
Average income in line with employment: I earn a ‘salary’ similar to what I would if I went back to full time employment, and significantly more than at the University (which underpaid technical staff).
Fully booked all year round: Whilst it took about 12 months to gradually build up my business, after the first year I was fully booked all year round and have continued to be so. This is through a combination of regular work, which accounts for about two thirds of my time, and one-off projects (these usually have a 4-12 week lead time). Unlike some businesses, my work isn’t seasonal – there are no quiet periods in August or December (or any other month!).
Diverse portfolio of clients: I’ve worked with clients in insurance, paint manufacturing, consumer credit, mountain biking, chemical manufacturing, aviation, training, publishing, electrical installation, equipment hire and more.
More time off: I take more time off than I would if I was employed, and I can take it when I want (the University was good for holidays but I couldn’t take time off during certain times of the year). If I want to attend groups such as Warrington U3A History (where I also give talks), I can do so without having to ask permission.
Fewer hours: I often work about 5-10 hours per week less than I would in a full time role. This is a personal preference rather than not having enough work.
Fully remote working: I don’t have to commute to work every day, nor do I have to travel up and down the country attending meetings and staying overnight. This is much better for my health and wallet.
No unpaid invoices: Whilst I’ve had to occasionally chase for payment, most clients pay me on time and I have never had anyone fail to pay an invoice. I’ve achieved this through screening potential clients, a strict ‘no contract, no work’ policy, payment terms of 14 days, and a general vibe of being the sort of person who knows how to and is prepared to take you to court if you refuse to pay.
Challenges and lessons
It hasn’t all been plain sailing – I’ve made some mistakes and had to overcome challenges along the way.
Losing a regular client who provided a third of my income: This was due to the client not having any more regular work for me, rather than any dissatisfaction with my service (they’ve hired me for one-off projects since), but it was still very painful. To prevent such a big loss of income in future, I’ve introduced a financial exposure limit, which is the maximum amount of work I allow to be outstanding at any one time.
Fluctuating income: As is to be expected, my income fluctuates depending on workload. A good year can be 50% above my average earnings, and a bad year can be 30% lower. That’s a wide band of income to be working in, so I have to make sure that my outgoings are either lower than my lowest earnings, or keep a cash buffer to see me through quieter periods.
Pricing: I still find it hard to price my services, and I always feel stuck in that area between not wanting to lose business by charging too much and underselling myself (other people have told me I don’t charge enough as well). I am trying to get into the habit of annual price increases to take account of inflation if nothing else.
Timewasters: Unfortunately I get a lot of people who waste my time – often with completely unrealistic expectations such as wanting a bespoke website built for £100 (and by 9:00 tomorrow morning). I’ve also been asked to attend meetings in person to discuss freelance work, only to have someone try and pitch me a full time role. I’ve managed this by being up front about how I work (especially rates and payment terms), and gradually tightening up the rules on how much time I provide for free. The initial consultation I offer is now limited to 30 minutes and has to take place remotely.
AI and LLMs: The idea that you can get a service like Copilot to write most of your code for you has taken root with some decision makers, and this makes it harder to justify my rates in some cases – why pay me when an AI can do it for free? The answer is that I will do a proper job and you’ll own the rights to the work, but when I started freelancing full time I didn’t have to make this argument. The other challenge is that a growing number of users are starting their web searches with the likes of ChatGPT instead of Google, which means I’ve seen a drop-off in enquiries from people who have found me through a traditional search engine.
Making Tax Difficult: Sorry, that should be Making Tax Digital. This is a grand scheme from people who have zero expertise in running a small business or building IT platforms, who have decided that it would be good for me to buy proprietary software (instead of the open source solutions I have at present) and file information with HMRC quarterly and annually (instead of annually), for no benefit to me. The date has been pushed back repeatedly – more evidence that it isn’t ready or useful – but I have to start using MTD-compatible software from April 2026.
Where to from here?
My immediate plans are to continue as I am, working with a group of regular clients, and one-off projects filling in the gaps. In the medium-term, I want to:
Launch side projects: I have a huge list of side projects that I want to work on, some of which would be useful in my business and for others in the same situation. In particular, my niche is building web applications which:
- Respect the privacy of users.
- Don’t share or sell data.
- Are accessible – including using semantic HTML and minimal JavaScript.
- Don’t feed your data to AI / LLMs.
- Do one thing and do it well.
I have become increasingly frustrated with the degradation of user experience online over the last 5-10 years, and I want to provide an alternative which harks back to times when web applications were simpler and didn’t track you. I’ve heard other people express similar sentiments, so I think there is a gap in the market which isn’t being met.
I also want to eventually charge for some of the side projects, so that I have a revenue stream that isn’t directly related to the amount of hours I work.
Move more into consultancy: Although I like writing code, there’s a cap on how much I can earn doing so due to the level of competition, and it doesn’t fully utilise my unique skill set. I would like to move more into a consultancy role, where I help clients design the right solution but don’t necessarily build everything for them.
Non-executive directorships: I’ve always wanted to be a non-executive director, which is someone who provides expertise at board level but isn’t involved in the day to day running of the company. I have struggled to find opportunities though – I suspect a lot of them come up within professional networks and particularly in London.