Three years ago this month I started Dev to Agency... whoa, that's a long time.

Originally this was a place for me to decompress... to work through my thoughts and feelings after selling my agency (while taking 12 months off to travel Australia).

But I slowly realised I enjoyed sharing the knowledge I had learned (the hard way!), and I built a nice community around it.

It then turned into a place for me to find consulting gigs - and through this blog (and newsletter) I helped a lot of dev agencies, they helped me, and that was a win.

Those who have been subscribed for a while will probably have noticed the significant lack of recent posts, and those who subscribed recently may have been disappointed about the radio silence. Sorry about this.

My new startup Userdoc is gaining traction and now needs my full attention - as I want my next five to ten years to be about building an amazing SaaS product - that's my next challenge.

Realistically, I won't be posting any more... at least for the foreseeable future.
I feel I've said what I needed to say, and the content will remain up, as the majority of these learnings are timeless, and I hope they can help people along their path.

All the best to everyone, thanks for everything.

Now let's all go out there and smash it 🚀

— Chris

The world has changed. AI is no longer a futuristic concept confined to science fiction movies, it's here, it's powerful, it's accessible, and improving by leaps-and-bounds every month (or day?)

In 2024, millions of businesses - small, medium, and large - will be turning to their software development providers for one thing: AI.

Your Clients Need Something This Year - and it's AI ✨

Businesses are looking to understand how AI can provide them with value, and how to start planning an AI strategy. I know this for a fact, as I have helped multiple businesses with this recently.

They want advice on how AI can help their business run more effectively, efficiently, or to save costs. It’s up to us as development vendors to ensure we are across this, and can advise and implement the tools and practices needed to support this technical revolution.

Don't be scared 😳

I know many software developers and agencies that are reluctant to embrace AI. This may be because they are too busy, it's too fast moving, the simply because they don't know where to start.

But the one thing I know for sure is that we are at a turning point right now... And to be a software professional on the right side of history, I believe you should add AI engineering to your toolbelt, and be able to offer this as a service to your existing clients, and new clients.

AI Consulting: Guiding Your Clients 🤝

As dev agencies and freelancers, we need to understand the fast-moving AI landscape. We need to know what can be achieved, the safety and ethics involved, and how this can positively impact businesses. This means staying up-to-date with the latest advancements - or at least understanding what's now possible, attending AI events, meetups, and conferences, and importantly - actively learning and experimenting with new tools and models.

Businesses (potential clients) are desperate for guidance and advice, they want

1. An expert technologist to help guide them (that can be you)
2. Examples of how AI can be used to improve their specific business
3. What are the potential risks and challenges right now
4. What the future opportunities could be

I strongly recommend you step up, and make it your responsibility to be able to answer these questions, and to help businesses develop a realistic and achievable AI strategy.

AI Augmentation: Work Smarter, Not Harder 🤖

Beyond consulting, we need to ensure we are leveraging the best AI tools to deliver value better and faster. This isn't about replacing humans, it's about augmenting our existing skills and processes.

Some industry-leading AI tools that can help agencies and freelancers include:

1. Github Copilot: This tool uses AI to suggest code completions and entire functions, which can significantly increase developer productivity.
2. Wix Studio: This platform offers AI-powered website creation and management, specifically built for agencies and freelancers
3. Userdoc: This AI-powered requirements management platform helps agencies and freelancers gather and manage project requirements more efficiently and effectively.

These are just a few examples, and new tools are being developed every day. By staying informed and experimenting with these tools, we can find ways to improve our workflows and deliver better results for our clients.

AI Implementation: Bringing AI to Life 🌱

Ultimately, it's up to us as the builders to implement AI-augmented and AI-forward services and products for our clients. There are so many possibilities, but some examples include

• Building AI-powered chatbots and agents to improve customer service and automate tasks.
• Developing custom software that leverages AI for data analysis and reporting.
• Integrating AI into existing systems to improve efficiency and effectiveness.
• Creating new AI-powered products and services for clients to sell or use internally.

The possibilities are endless, and the potential for innovation is immense. By embracing AI and integrating it into our offerings, we can help our clients achieve their goals and stay ahead of the curve.

AI is here to stay, and it's changing the world around us.

As development agencies and freelancers, we have a unique opportunity to leverage our skills and expertise to help businesses navigate this new landscape. By embracing AI consulting, augmentation, and implementation, we can add immense value to our clients and ensure our own businesses thrive in the years to come.

If you want more clients for your dev agency or freelancing business, then I wrote a book for you - it's called Client Bytes - Dev Agency and Freelancer Sales.

My good friend Nat (who runs a very successful dev agency in Calgary) co-authored the book to tackle one of the hardest questions...

1. How do I get more clients?
2. Get more work from existing clients?
3. Find the sales and marketing tactics that work best for me?

Over 12 months of writing, over 22,000 words, 74 pages of highly actionable content with 37 proven strategies for landing more clients and increasing your business's awareness.

Now I'm not a typical salesman type... and it seems super "salesy" to say this... but I'm so confident this book will improve your business that I'm happy to offer a 100% money-back guarantee.

Oh, and because you are my awesome Dev to Agency friends, you can get 10% off if purchased within the next 7 days.

Click the link below to learn more and purchase.

📖 Client Bytes - Dev Agency and Freelancer Sales

Thanks friends, I'd love to know what you think - so please leave a review on Gumroad 🙂

Chris.

There are a million things you need to focus on in an agency, and recently I wrote an article about wearing many hats around how learning to do lots of things is a superpower in your agency.

This is because it’s what worked for me, and helped me grow a 7-figure business in a couple of years. By learning many new skills fast while staying small and lean.

But running your own business needs to be about doing what you want, and I speak to many agency owners who don’t want to end up wearing “all the hats” - and that’s totally fine.

Instead, you can focus on what you do best, where you have the most impact, and what you enjoy the most. This is your superpower.

Whether your superpower is software development, UI/UX design, sales and marketing, or whatever - there are ways to optimize your time on these tasks and leverage them to build your business.

Let’s be honest - there’s no way you can just be a software developer who codes 100% of the time and also runs an agency. That’s an employee.

But you can use strategies to ensure you spend the most possible time doing what you do best -  let’s run through a few of those.

Leverage you niche 🎯

Working on every project that comes along can work well to get started, but being a “jack of all trades” comes with its overheads, whereas having a niche can reduce some of those.

A lot of times an agency revolves around sales, marketing, and building a brand - because this is how you get work.

But what if instead, people came to you because you were the #1 expert in your specific field?

And the more projects you get of that certain type, the better and more efficient you get at it, and the more clients are blown away by your agency's awesomeness, the more they cannot help but come back to you and recommend you to everyone they know.

Building a niche is not something that happens overnight, you can’t avoid putting in the initial work, but long term it can reduce your sales and marketing efforts, increase your referral rate, increase your profits, and optimize the time you spend delivering projects, not searching for them.

Leverage humans 🧍‍♀️

Outsourcing or insourcing is another way to ensure you stay focused on what you do best.

There are hundreds of tasks that other humans will be better at than you, but just make sure you're not delegating your competitive advantage to someone else. For example, if you’re awesome at paid advertising and lead conversation, that’s probably not something you want to outsource (unless of course you very specifically train someone on your methods).

There are hundreds of tasks that other humans will be better at than you, some include:

• Hiring an accountant to do your taxes
• Hiring a sales and marketing team to create and implement paid advertising
• Hiring a tester to help with quality assurance
• Outsource basic website creation to a freelancer (yes, even if you’re a programmer you can focus on the more impactful code!)
• Outsourcing copywriting to a freelancer
• Outsourcing graphic design to a freelancer
• Delegating admin tasks to another team member
• Delegating admin tasks to a virtual assistant
• Delegating all client inquiries, lead followups, and sales funnel tasks to a virtual assistant

Leverage tools 🛠️

A great way to save time on tasks is via automation and optimization - and computers are awesome at both of those.

There are a ridiculous amount of tools at your disposal here, but some that I use include…

• ChatGPT for everything 🙂
• Xero to get your accounts in order, keep track of finances, and cut down the time your accountant needs to lodge your taxes
• Tweethunter and Taplio to automate your social media, and get great content inspiration
• WixStudio to create and manage websites, specifically built for agencies and freelancers with some pretty amazing AI
• Apollo.io for lead generation and email automation
• Calendly allows people to book meetings that fit everyone’s schedule, avoids ten emails trying to find a suitable time
• Loom can easily record and host video for asynchronous and repeatable communication or training
• Zapier and Make.com can automate all the things
• Userdoc for software requirements management and automation with AI (shameless plug!)

Leverage documentation 📚

If you don’t want to do something again, then write down how you did it, and next time you can get someone else to do it. Simple.

It’s amazing how many people don’t do this, but seriously, this is how businesses scale - and if you are a one-man-show documentation will allow you to outsource and onboard a lot ****easier. Plus it’s an investment in stopping a lot of future questions you will be asked such as “What do we do when X happens?”

Examples of what you can document…

• Your lead conversion steps - what you do when you have a new lead (e.g. tell them about Y, ask about X etc.)
• Standard client contract
• Your client onboarding steps - what to do when your lead becomes a customer (e.g. send them through the client contract to sign via DocuSign, set them up in Notion, Jira, Xero - whatever)
• New hire onboarding steps - this is how we work, this is where we keep our code, this is where we keep our requirements, this is where we keep our shared passwords, etc.
• Coding standards - we use tabs (or spaces), we leverage this coding standard (e.g. AirBnB JS), we use semantic commit messages, we do PRs using this template etc.

Documentation can seem like a ridiculously large task - but you shouldn’t write it the first time you do a task. Write it the second or third time, so you have time to optimize your process, plus it doesn’t seem like you’re spending all your time documenting.

An important note with documentation is it’s often easy to write, but hard to maintain.

You need to embrace continual, living documentation, that just becomes part of your business foundations and creates a living, breathing resource that evolves over time.

Whilst the tactics covered above still take a bunch of work to implement, the payoff is an increase in the time spent on your thing - your superpower.

And ultimately, you need to build a business you want to work in.

Like a lot of developers, for the last few months all I have been thinking about is AI - and yes - this explains why my posts on Dev to Agency have become fewer in frequency 🙂

I’ve immersed myself in learning all I can about the new frontier of generative AI, and playing with every tool possible. I pivoted my agency consulting role into AI engineer, and started building tech demos of how gen AI can be used to offer new and improved services to clients and enhance their products way beyond what was previously possible.

I’m integrating more advanced and richer AI experiences into my requirements management startup Userdoc, and every week I am blown away by what is possible today.

It’s a freaking exciting time to be a software developer.

Why dev agencies need to pay attention to AI ✨

I ran my software agency when Blockchain was getting big, and when Web3 was going to change the world. But I largely ignored it, and never thought of pitching to my clients how Web3 could help them make money or save money. I felt that when the mass hype died down - it wouldn’t in fact change the world like people thought, and on that point I feel I was correct.

But with AI and large language models (such as GPT-4), it feels different.

AI can literally allow us to do things we NEVER could before. It’s not a new way of doing X, or a faster way of doing Y, it’s an entirely new platform that unlocks completely new opportunities.

Similar to when modern smartphones came out around 2007 - a whole new market segment has just opened up to people who can code and create software.

If 2 years ago a client asked my agency “Hey Chris, we need to chat to our database using natural language, oh and I also want to see graphs and charts!”

I guess I would have known *someone out there could probably do it *-  but unless we could find them or a product/API we could leverage, we would have to say “Sorry, this is out of our wheelhouse!”.

But I’ve just built this exact thing in a couple of days, allowing someone to talk to their database (any database) using natural language sentences. Oh, and it supports replying with dynamic graphs and charts - just like magic.

And this is just a drop in the ocean of what you can offer to your clients…

What dev agencies can do today with AI, that they couldn’t do yesterday 💡

OpenAI released API public access to GPT3 way back in November 2021, and I know that specialised AI peeps could weave magic for a long time. But I’m focusing on the generalist developer or agency - and what AI and LLMs can allow us to do a lot simpler than ever before.

The answer to what we can do is essentially endless, but focusing on areas where software development agencies could add value to their clients, here are a few high-level Ideas.

• Data mining & semantic search across unstructured datasets - whether it’s ad-hoc Word documents or PDFs, emails, or speech-to-text-recorded telephone calls. AI can add structure and summarisation to this “data”, and also look through it to pull out specific details (e.g. “Which of these recorded telephone calls are complaints about products or services?”
• Quick and easy sentiment analysis - this could be used for classifying comments, reviews, questionnaires, forms etc. as Positive, Negative, or Neutral. And even providing more context, e.g. “Samantha G’s telephone call was negative, expressing anger and dissatisfaction about the Tesla she purchased in 2022”
• AI Agents - Chatbots have been around for ages (with varying degrees of “AI”) - but whereas chatbots “talk” - agents “do”. Agents perform tasks, and actually get shit done. Agents can actually cancel a customer's subscription, not just tell them how to do it themselves. Agents can actually order you groceries, not just talk about it. Agents are the new generation of chatbots, and agents will be huge.
• Automated reporting - managers want a weekly report email about X,Y,Z - often just to see if everything looks ok. But what if AI sent automated report emails when anomalies happened, not using hardcoded boolean logic, but more natural decisions. “Does this week's orders seem similar to the last 10 weeks’ orders”, if not perhaps “Generate a small summary of the differences”, and let’s get a manager to check that out (all automated).
• Systems integration - like my database chatting example previously (that generated valid SQL from natural language), LLMs are amazing at understanding code and syntax. You can give them an API spec and they can generate example integration code. You can give them code, and they can generate a detailed spec. This opens a lot of opportunities for writing dynamic integrations and allowing the LLM to be the middleman.
• Content generation - I’m not talking about blog posts or product taglines. But boring business content that needs producing. Forms, reports, documents, and even emails. Things businesses spend time on that now could be automated.

So what should I do about it? 🤔

I believe development agencies are in a great position to learn, master, and offer AI services.

Even though it seems we are at “peak AI excitement” at the moment, we are not.

As technologists, we see it coming first. Give it a year, maybe two years - and every single business we will want to harness the power of AI for two core reasons that are as old as time.

1. To save money
2. To make money

Sure, there will be AI-assisted products and SaaS services to help them, but just like always, organisations have endless business rules and complexity, and many will still need custom software to work exactly the way they need it to.

How do I get started? 🏃‍♂️

Start learning by playing with AI development tools, AI is moving crazy fast, and every day there are new tools, models, and companies that provide opportunities and inspiration.

I don’t just mean ChatGPT, I mean tools like LangChain and BabyAGI that let you talk to many different models in many different ways. This is what opens open the power to developers.

Build small demos, build them for fun, build them to explore what’s possible. It will inspire other ideas and help you get a grasp on the true power available.

For example, recently I got beta access Anthropic's Claude API, which allows you to send 100k worth of tokens in your prompt. I didn’t get the power of this at first, but I now see it. And it’s huge. Being able to send around 75,000 words in your LLM prompt is a game changer, and it’s filled my head with new ideas.

Find a way that works for you to stay up to date. I listen to AI Podcasts and read AI newsletters - and follow people in the AI space on Twitter such as swyx.

Have fun, exciting times ✨

As an agency owner, one of the biggest challenges is having to juggle multiple tasks and responsibilities - wearing many hats.

Unlike larger companies with specialized departments, as a small business owner, you are responsible for everything from development, sales and marketing, accounting, and customer service - you have to wear many hats.

Some people hate this.

Some people love this.

I see it as a necessary part of getting your business off the ground, and if you do it well, you may not need to wear all those hats forever (if you don’t want to).

This is not SEO targeted AI generated clickbait article, where I tell you how to manage your time, juggle tasks, and stay focused - there are a billion of those articles on the net if you want to google them.

Wearing many hats is your superpower

In the early years of my agency, half of my clients were solo founders trying to build their startups - by paying others quite a lot of money to do so.

They could drop anywhere from $15k to $500k just getting their idea developed, and then another $XYZ on sales and marketing when they realise building it’s the easy part - the hard part is getting a steady stream of paying customers.

If they were successful in getting some initial traction - and hopefully they were - best case scenario they had to keep paying my agency to add new features and evolve as they pivot and try and find product-market fit. And this could still cost tens of thousands of dollars every year, for many years.

But If you can personally build something, attract customers, sell to those customers, deliver to those customers, and build a good ongoing relationship for future work - you are in the 0.01% of people in the world who can successfully do this.

Wear more of the right hats

If you don’t have any idea on how to do SEO, but you’re building websites for clients and confident this is a skill that will add value, and add to your bottom line - then make a point of filling your knowledge gap and go and learn that shit yourself.

Investing in yourself (and your superpowers) will compound over the years as you become an unstoppable force, and add more and more skills to your arsenal.

New technology can feel endless…

You don’t need to learn every new framework or buzzword javascript library - be strategic about what skills will allow you to go deeper, and serve your customers better.

It’s possible that this is web3 or AI.

But realistically, unless you wear these hats already - it’s probably more like sales and marketing.

The hats I wear

I started as a developer. Didn’t care about anything else but code, it’s all I wanted to do - night and day.

I went from junior to mid, to senior, to technical lead.

I built wider skills in client relations, team relations, management, communication, estimation, automated testing -  everything from frontend and backend to middleware and DevOps.

I left 9-5 work to found a startup with some friends. I learnt about entrepreneurship, business, finance, product-market-fit (or trying to at least!), pitching to investors, working super fucking hard, and eventually - burnout.

I walked away from my startup and started a development agency. That’s when I really learn about sales, marketing, building a network (professional and personal), long-term client relationships, the balance of quality and cost, building a team, growing a team, partner disputes, acquiring other businesses, and selling my business.

I then travelled around Australia for a year in an RV, and started Dev to Agency - where I learnt about writing, teaching what you know, idea generation, newsletters, consistency, distribution, Twitter, and building online relationships.

I’m now working on Userdoc - a startup around AI requirements management, where I’m wearing all the hats I’ve worn over the last 20 years of being in the tech industry.

But I’m still learning - SaaS pricing, product onboarding, enterprise sales and about 62,000 other things.

I’m grateful I’ve always pushed myself to keep wearing more hats.

I haven’t done it all myself, as I’ve built some amazing teams - but only once I’ve proven that role needs to exist (e.g. I did sales and marketing until it was time to hire a team etc).

It’s about not having to solely rely on others, and being able to get shit done yourself.

You can build an amazing business wearing many hats - and that can be your superpower.

Parkinson's Law tells us that “work expands to fill the time allotted for its completion” - meaning the more time you have the more time things take.

And the origin story is about a woman who had to send a postcard.

For many people, that may be a 3-minute task. We are busy, so we don’t want to draw it out.

But this woman spends an hour looking for the exact right postcard, another half hour looking for her glasses, 1.5 hours actually writing the card, 20 minutes deciding whether or not to take an umbrella along on her walk to the mailbox etc. In the end, sending a postcard takes her a full day, instead of 3 minutes.

And I’ve often caught myself falling victim to Parkinson's Law.

Give me three days to write a proposal for a client, I can do it.

Give me one day to write a proposal for a client, I can do it.

Give me 3 hours to write a proposal for a client, it’s tight.. but you know what? I think I can do it.

An example of Parkinson's Law

So you have to choose a technology framework for a new project, a relatively simple project, an online inventory system with an administration interface.

You’ve previously used NextJS with much success, but lately, you’ve been hearing a lot of talk about Laravel, and are interested in giving it a try. You have 6 weeks until you need to start the next project, so that seems like enough time.

You can imagine what happens here, where previously you have the core of the project scaffolded in 2 weeks, this time you only accomplish

1. Researching Laravel boilerplates, but decide to start from scratch to learn the most
2. Learning Docker, as this ships with the default Laravel project and you haven’t used it before
3. Trying different authentication packages,  as the ecosystem is huge and there are many options
4. Going down a rabbit hole on Laracasts, learning about Pest PHP testing framework
5. Getting stuck with PHP syntax, as you haven’t used it for a few years

You get the picture.

You’re not too stressed at the start because you have 6 weeks, but as time rolls on, you're not where you want to be. Perhaps you bit off more than you can chew. Sure, you learn some new tech, but it’s rushed and stressful - and importantly, even though it seemed like a long time, you ended up going 2 weeks over the 6-week deadline, which ate into the start date of your next project (making that one unnecessarily stressful also).

Learning is an essential part of life, especially in technology, but it’s important to choose your battles, and understanding our brains - and that Parkinson’s law exists in all of us - can help get your priorities in order.

For example, you could stick with NextJS, and pump this project out in 4 weeks in your usual high quality in the ecosystem you’re comfortable with. Then you have 2 weeks up your sleeve before your next project - a perfect time to start learning Laravel without any pressure.

How to avoid Parkinson's Law

If you know humans have a bug, where we “fill the time available” - then we can try and squash this bug. As developers, debugging is what we’re good at right?

Often it’s easier to debug a piece of code than it is to debug your own behaviours and habits. But to get work done efficiently, and not fill up the time available, these hints might help.

1. Always have a crystal clear scope

If you don’t have a plan, you’ll never know if you’re on track.

Think deeply before coding, and run all edge cases and question past your client in order to agree on what should be done. I capture super detailed user stories and acceptance criteria in Userdoc (yes that’s a shameless plug) and then get my client to approve the requirements even before starting development.

This helps mitigate the “oh I just realised this…” and “We also need to handle that” which can often occur, and instead makes you more focused on exactly what was agreed.

2. Estimate tasks, and try and stay within the range

Developers traditionally hate estimates, but they serve one good purpose - reminding you how long you thought the task would take.

If the postcard-sending lady (in the origin story of Parkinson’s’ law) had estimated the task would take 30 minutes and planned her day around that eventuating - then she wouldn’t have taken 1.5 hours writing the message… or at least she would have had to mentally acknowledge that she’s taking considerably longer than planned.

It’s ok if things take longer - as long as you are actually ok with that, and you are mindful about the time spent.

Timeboxing techniques have worked for centuries to keep you mindful, and methods like Pomodoro have served to get the most out of the time you have and make you more aware of what you are doing right now - compared to what you should be doing.

3. Do what’s most important first

If the project you’re working on is an AI house blueprint generation platform - where you upload photos of houses and it uses AI to predict what the blueprints and floorplans look like - then just straight into generating the AI blueprints.

Although starting with the server setup, the DB structure, then the user authentication and admin website etc might seem more of a natural starting progression - it’s not what's most important… it’s fine to do things out of order and put the important parts first.

To avoid filling the time available, you want to start with the most important, most risky components, focusing on getting them up and working as required, but saving any bells and whistles to the end (assuming there’s time that is).

4. Don’t get stuck on trivial edge cases

It’s hard to know what’s trivial in a project, as it varies from agency to agency (and client to client), but a good example might be trying to get a website to be pixel-perfect on a super old version of Chrome on Android. Sure, it’s nice to have - but if it’s literally going to affect maybe 0.5% of the users of the website, that should only be something you tackle if the project is delivered perfectly, there are no other improvements, and you’re under time and budget.

My mentality is to log the small bugs and improvements in your project management tool, but first, get the feature done. I love the saying “make it work make it right make it fast”.

You can then triage the issues and decide when and if they need to be dealt with. You don’t want to spend an unplanned day on something that actually adds little value to the project, so always asking yourself “is this trivial?” or “is this a super edge case?” is a good way to detect that it’s occurring.

Note: I’m not encouraging delivering buggy code by any means, I’m encouraging being self-aware of what’s the most important - and making intentional use of the time you have.

Sometimes it’s fun to spend a couple of days tinkering with new technology, or getting sidetracked going down a stack overflow rabbit hole for a few hours  - and often that’s how we learn and grow.

But if you’re trying to increase productivity, and trying to hit your project deadlines (whether they are self-enforced or enforced by your clients) - then integrating the techniques above might help you out.

Writing an article on the internet about burnout feels a bit like writing about “how to wake up early”, or “eat better and feel great!” - it feels clickbaity and it’s all been said. Am I really that hard-pressed for content?

But the truth is, burnout’s a bitch…

It hits us differently, at different times, and for different reasons.

And the other truth is, agency life (or running any business) can lend itself towards burning out.

It can feel like an endless marathon, and even if you “make it”, it can feel like best case scenario - you’ll just get more work, more projects, and you’ll just have to work harder.

I don't have all the answers, but I have a couple of examples so far in my life.

My first taste of burnout

I have two distinct memories of work-related burnout… so far at least.

The first time was during a startup I was trying to get off the ground before I started my agency. We’d built the MVP over a year (way too long!), we took part in a startup accelerator and were trying to raise money (way too early!), and not too surprisingly, were struggling to find customers and product market fit.

One morning I woke up and felt like I’d hit a wall. I had zero enthusiasm for my startup.

The little things seemed impossible. In fact, I felt like I never wanted to speak to any of my team members again, never wanted to talk about my startup again, and I had lost all faith in my ability to do anything.

It felt like I almost instantly fell out of love with the things I loved.

But the truth is, it accumulated over many months. Slowly, and sneakily, my brain and body wearied as I spent thousands of hours trying to get this thing off the ground.

I thought I had an ok work/life balance because I didn’t sit at my computer all day - but really - all I thought about was my startup, all I talked about was my startup, and all I cared about was my startup.

When burnout hit, it hit hard, and I didn’t have the skills back then to tackle it, or even realise what it was. So I just quit.

It took many months to regain my confidence and sense of self. It was only then I started to realise my situation crept up slowly, and most of it was of my own doing.

My second dance with burnout

Fast forward… I’d worked for a few other startups, then gone into freelancing, and finally started my own agency.

4 years into my agency, we were doing awesomely. We had a team of 13, a fantastic office, and great clients - but something felt… off.

I recognised the symptoms: being unmotivated with things I’d previously loved, constantly tired and stressed, having (and causing) conflict with my business partners, and where I was normally a confident person - feeling completely out of my depth and overwhelmed.

The difference was this time the business wasn’t a struggling startup, it was an agency going well. Really well, lots of money, lots of clients, externally this was exactly what I wanted.

Without having a clear time when it started, you can start to normalise burnout, and question yourself that “maybe I always felt this way?” - but the second time around I knew it wasn’t the case. That’s not the real me, and I need to do something about this.

Avoiding burnout

Now, this is the bit that really feels like a self-help article, but I can’t talk about burnout without talking about how to avoid it. I’m certainly no pro - and as with everything on Dev to Agency, this is just my experience, and yours may totally vary.

First thing, everyone's different. We all have different triggers, and that means we all have different ways of mitigating burnout. But first things first, you need to do some searching, and figure out how you work.

1. Know your triggers

It might be crazy long work weeks, missing out on important events, or not keeping up your general mental hygiene. Triggers for me are starting to ignore physical activity, and drinking too much alcohol as a temporary way to escape from my stress. I know that if I let my health or fitness slide too long it starts to compound my stress levels. I function quite well in a high-stress environment but if I keep that up for many months and it starts to take its toll.

2. Put yourself first

Take time just for yourself, and treat your physical and mental health like it’s the most important investment. Working super long weeks is sometimes necessary, but don’t use it as an excuse to eat like shit, or avoid things that make you feel good about yourself.

I find it hard to justify “time off”, but I now know it’s key. I take time off for my mental health, and you should too. Maybe that’s sleeping in and starting work a couple of hours late one day, going to watch a movie, or cancelling plans just to read a book and chill.

I didn’t do these things initially. I liked to work, computers were my hobby, so it was easy to stay in front of a screen and tinker.

But just like eating healthy - these things don’t always feel good right away. There’s no “I FEEL AMAZING!” moment, it’s just that it can stop you from feeling less amazing.

3. Emphasise family and friends

Talking to people who aren't in my work bubble is key for me to get out of my own head, and realise there’s a big wide world out there (and everyone has their own problems!).

There were times when I felt too busy to attend a dinner party or to see a band with mates - but these days I make it one of my highest priorities. Taking those few hours won’t kill your business, in fact, I’m confident that (for me), it cultivates more energy to put towards my business.

Recovering from burnout

With my first taste of burnout, recovery took a while, close to a year. I felt a lot of guilt because I didn’t understand what happened. I lost trust in myself, how can my feelings for something just change so fast? Will this keep happening to me? Why couldn’t I stop it?

But my second time, I saw it coming before it was too late, and I took some time off with my wife in southeast Asia.  I was still very much in the pre-burnout zone, but that trip gave me time to break the loop of my endless day-to-day thinking and gave me space to think slightly more long-term, and strategically.

Within the next 6 months, we choose to let around 50% of our clients go, and sadly a few staff members. We needed to operate at a sustainable long-term pace, I didn’t want anyone working more than 38 hours per week - including myself. We put in place strong processes and procedures with our clients to reduce everyone's stress, and we because a lot better at setting realistic expectations.

I learnt to delegate. I hired my first senior-level solutions architect, and along with my wife as project manager, they took over one of our largest clients so I could focus on the remaining part of the business.

I went to the gym religiously right in the middle of the day, even though I rarely wanted to. It was really hard for me mentally to leave work at that time, but afterwards, it made me feel good, feel proud of myself - and for me, feeling like you are accomplishing things is a huge part of recovering.

I’ll be honest - at that time I was very close to just shutting down my agency. Just like with my original startup, I wanted to completely walk away, and never speak to a client or staff member ever again. But I’m really glad it didn’t happen - because the next 4 years at my agency were even better than the first, and a lot more sustainable and happy.

Burnout happens, especially in states of prolonged stress or uncertainty. And running your own business is a recipe for this, as it’s easy to put yourself under a lot of pressure.

This is a big reason why many businesses fail. Not just because they run out of money, but because the owners run out of steam.

So get in early. Prioritize your health, and carve out time to decompress - however that looks for you. Your business will be better off with long-term sustainability, than constant intensity.

Unfortunately, not all clients are easy to work with. Some may be overly demanding, always negative always blaming you - or even disrespectful or abusive.

Often times these are also the clients that are always asking for discounts or penny pinchers - and the ones you have to chase up their invoice payment 5 times before seeing any money.

Working with toxic clients can take a toll on your mental health and overall well-being. It’s important to remember that your time and energy are valuable, your work is valuable - and it’s not worth sacrificing your well-being for a client… especially one that doesn’t appreciate you.

The value of a positive work environment 💕

When I first started my agency, all I cared about was finding work and making enough money for my business to survive. But after we’d built up a regular client base, that fear of “not enough work” started to dissipate - and I could take a breath, and reassess. Had I actually built a company I loved working in?

I mean, I loved most of it. I loved coding, working on cool projects, and hanging out with my team. But, there were a couple of clients who were always negative, they always demanded more, and even though they didn’t spend much money with us, took up a huge amount of time, and expected to get priority treatment over everyone else.

Relationships like these can bring the energy down, and as a small tight-knit team - our energy and culture were integral.

As a software development agency, it’s important to create a positive work environment for your team. This includes not only the physical space in which you work, but also the energy and culture of the team. Working with difficult clients can create a negative atmosphere, which can affect the productivity and morale of your team.

What’s their problem? 🤔

Just because a client is showing toxic traits this doesn’t mean they are a toxic client, it could mean that your relationship is damaged, but that can be fixed.

This could be due to many reasons, the way the relationship started, or they may have a problem that they have not communicated, or they may think that's how they have to deal with businesses to get what they want.

Whatever it is, there's really only one way to find out: Communication.

Arg... hard conversations suck, but they are hard because they are important. They’re hard because they are often things that should have been discussed a lot earlier.

Have a respectful chat with your client, not via email, but in a meeting (virtual or in person). Certainly don’t call this a “TOXIC CLIENT CHECKUP MEETING” - just a catchup meeting.

Assuming there is no specific issue, start it off as a general check-in, with questions such as.

• How do they feel things have been going?
• Do they feel you are meeting their expectations?
• Have they been happy with the work you are doing?
• Have they had any specific problems?
• What do they think you could improve upon?

These will help tease out some information you may not currently know.

If you know there’s a specific issue you haven’t talked about, or an issue that comes up in the general discussion, confront it head-on. You need to have this issue resolved, and removed.

Maintain composure, be polite, be emphatic - but be clear.

If it’s a problem with the quality of your work, clearly explain the issues. But also mention that software development isn’t like lego - it is significantly more complex. If you feel there were some quality issues that were below standard, own it, and discuss how you will endeavour to stop these issues from taking place in the future.

If it’s an issue with money - although no one likes talking about money - it has to be done. If they always have a super tight budget, explain to them that they need to allow more padding, because issues will always arise. For example, if they have $10,000 and cannot spend a cent more, then I would recommend we only scope $8000 worth of work, and have a $2k buffer up our sleeve.

Whatever the problem is, try to calmly, and collaboratively solve it right there on the spot. This cannot happen all the time, but most clients when confronted with a direct conversation will be open to a solution (or at least agree to try a solution).

Don’t get me wrong, the toxic relationship isn’t all your fault, it might not be your fault at all. However you are in the client services business, and it’s your job to try and build and repair relationships. At least try.

What’s your problem? 🧐

Hearing them out is great, but you also need to clearly set or reset your boundaries.

Be clear about what’s not working for you, and how you will need it to work in the future. This may be.

• User acceptance testing is completed within 2 weeks of handing it over
• That all communication is done in a respectful manner
• Invoiced are paid within the agreed timeframe
• Uncontrollable scope creep is having a negative impact on the projects

Whatever it is. Everyone has different problems - but you need to communicate these unambiguously, and make sure your client understands.

This may sound like an intense conversation, but it doesn’t need to be as intense as it sounds. Remember, show empathy, be future-focused (don’t bring up past issues, focus on future solutions), and always be respectful.

When it’s time to say goodbye (detox time) 👋

But after all that sometimes, despite your best efforts, you may find that a client is simply not a good fit for your business. This happens.

When possible, let them go with grace and poise, and ensure your firing is at the same high standard as your onboarding.

1. Finish all work to completion (if possible), to the same high quality you always have.
2. Give them advanced notice. You don’t need to be harsh and list the reasons, just tell them you don’t feel this will work out long-term, or your agency is going in a different direction.
3. Provide them with a list of alternative agencies or freelancers that could help. Depending on the reasons, you may not want to recommend your friends - however, a quick google search will help you hear. You are in a better position to judge tech companies than they are, so help them out.
4. Provide a handover of all source code, documentation, passwords (securely transferred) etc, plus any other information the next agency or freelancer may need - it’s much easier to do this now than in 6 months when a random agency contacts you asking about an API key or something.
5. Be sure to officially sign over any code and intellectual property to them, again, this isn’t something you want to have to deal with in the future.
6. Try and end on a high note. Don’t burn bridges unless the bridge is already completely on fire, as you never know what will happen in the future.

It’s ok to let go of clients that are bringing you down. And while it’s important to always be professional and respectful, it’s also important to protect your own well-being and the well-being of your team.

Setting boundaries, being honest, and prioritising your values are key to maintaining a healthy and successful business.

Remember, not all clients are for life, and it’s ok to let go of toxic clients (or outgrow clients) just the same as it’s ok for clients to let you go.

Without boundaries, you don’t have a clear cutoff point of what you’re willing to do, and what you’re not willing to do.

Whilst this may be ok at the beginning of your business, in the long term knowing exactly where your boundaries are, and communicating these to your clients and staff - will allow you to be more proactive and less reactive.

Let’s run through an example without any boundaries set, then the same example of clear boundaries set.

Without boundaries 🤷‍♀️

• It’s 7:15 am and your phone rings just as you get out of the shower, you see it’s a client you worked with mid-last year, so you pick it up.
• Turns out their website’s stopped working for some reason, they are pretty angry as a lot of their business is online.
• You let them know you’ll look into it right away, quickly get dressed, and head over to your computer (a little annoyed, and a little sleepy)
• The client rings you back, just as annoyed - and asks if you have fixed it. You mentioned you have literally just sat down to start looking into it, and you will let them know once it’s fixed.
• It takes you an hour or so to figure out the problem, which turns out to be one of the APIs they use on their homepage has changed overnight, and it’s stopped the site from loading.
• The API provider didn’t announce this change which is frustrating. You find the updated API documentation, and in just over an hour you have fixed the problem.
• You ring your client back and let them know you have fixed the problem. They are still annoyed, but less annoyed.
• You’re happy that the problem’s over as you have a lot of other work to do, finally eat a quick breakfast - and head into your office.
• At the end of the month, you invoice the client for the 2 hours of work.
• They are furious, and ring you instantly, saying how it’s completely unprofessional to deliver a website that breaks and loses them money - and then charge them to fix it. They refuse to pay the bill.

This is an extreme example - but it’s also typical of how many client services business function.

There are no clear boundaries, expectations were not correctly set - and it affects everyone.

Let’s look at this scenario again, if clear boundaries had been set.

With boundaries 🙅‍♀️

• It’s 7:15 am and your phone rings just as you get out of the shower, you see it’s a client you worked with mid-last year, but you let it go to voicemail. You have no ongoing agreement with this client, and besides - your day doesn’t start until 9 am.
• You have your shower, do a quick meditation, and have some coffee and breakfast.
• You get into the office just before 9 am, check emails, and then check your voicemail from the client. You understand the issue and confirm that their website is actually down.
• Instead of calling the client back, you shoot them a quick email letting them know that you understand the problem, but are very busy at the moment, and as they are not on any support or retainer agreement unfortunately you won’t have time to look into the issue for 2 days. You also remind them (like you have before) of the support plans you offer and attach a PDF outlining the different response and rectification times they get for each plan.
• They reply to the email slightly annoyed - but understanding that as they chose not to go on a support plan previously, the response and rectification times are “as soon as possible”.
• Two days later you look into the problem, see it’s a change in the third-party API that’s in use, and take a quick look at the documentation. You then get back to the client and let them know the problem is not to do with the code you delivered, but that a third party you rely on has changed their API. The fix will be around 2 hours -  and you check if they are happy to pay for this - if so, you can schedule the fix for tomorrow morning, so should be fixed by 10 am.
• They agree, the fix gets done, and they pay their invoice.

With the above example, boundaries and expectations had already been set with the client.

You had previously let the client know about the support plans you offer, and the different response and rectification times they get for the monthly payment.

They previously understood this, and specifically choose not to go on a support plan - which they then understood meant they get the lowest priority if they have any issues.

This meant that you didn’t need to answer the call out of hours - in fact, you didn’t have to reply to them until you had the time. This is a boundary that you understood, and they understood - so they could not be surprised when you didn’t pick up the phone.

The client also understood that after the project is signed off and live - all work is charged for and that there is no free support. This is a boundary you set, and they understood.

So when it came time to pay for the time spent resolving the issue, there were no questions.

What boundaries to set 🚫

You don’t want to be an over-the-top dictator who demands everything has to be done your way - that’s not the way to keep clients long-term.

The boundaries you want to set are the ones that keep you and your team happy, efficient and make for a healthy sustainable work environment.

For me this meant…

1. The working hours for me and my team keep (Mon-Friday, 9-5, not including public holidays or the week between Christmas and New year)
2. Unless they are on a high-priority support plan which can include weekends
3. The response times for getting back to client requests
4. During project delivery - 1 day
5. After project sign-off and the project is live - 2 to 3 days, unless on a support plan in which case it depends on their level of support
6. Resolution times for dealing with issues reported
7. During project delivery - the next sprint
8. After project sign-off and the project is live - As soon as possible, but may be up to 2 weeks depending on current projects, unless on a support plan in which case it depends on their level of support
9. Costs for dealing with issues
10. During project delivery - no cost
11. After project sign-off and the project is live - all work is charged at the standard hourly rate
12. Issue communication methods
13. Email is preferred
14. Phone if urgent and on a support plan
15. Scoping new work (after the project is live)
16. As soon as possible, but may be up to 4 weeks - unless on an ongoing retainer (in which hours can be used for scoping)

Your business may require other boundaries, such as having graphic design files delivered X weeks before development, or never contacting you directly, but instead via your account manager. This is all up to you, and the boundaries you need for your business to thrive.

How to set boundaries ✅

You don’t do this in a “boundary-setting session” - you do it in an ongoing, consistent manner throughout the life of your relationship.

It starts with setting expectations up front, letting them know how you work, along with what you are willing to do, and what you’re not willing to do. A project workshop is a good time to introduce this, but you should also capture this in a signed agreement with your client, as talking is great - but things need to be written down.

As the project nears delivery you should remind them how you handle ongoing support and maintenance - and that if they want priority service they will need to pay monthly for this.

And lastly, you need to reinforce your boundaries every time they get stretched.

Whilst it’s ok to break your own rules every once and a while (e.g. doing a small change on the weekend for a very important client) - you need to show your standards and be consistent - so boundary setting is often an ongoing task.

Without setting boundaries, you will always be reactive, acting based on how you feel on the day as opposed to a strategic, consistent method you’ve laid out - and clearly communicating to your clients and staff.

Build a business you want to work in

Build a business with boundaries that protect you and your team.

When it comes to software development, one of the most important goals is to create a product that is bug-free… or at least as bug-free as possible.

We all know software bugs can cause a range of issues, from minor annoyances to major disruptions and even security vulnerabilities.

Bugs mean more development work and more time re-testing, and if the system is very buggy - it can impact the relationships with your clients.

But not every bug is the same. To create software that is as bug-free as possible, it's important to first understand where bugs are most likely to occur and how to prevent them.

There are three main places where bugs can happen in software:

1. Bugs in requirements
2. Bugs in understanding the requirements
3. Bugs in code

Each type of bug arises differently and requires a different method of mitigation.

Let’s jump into each of these in more detail.

Bugs in your requirements 🐞

These are issues that arise when the requirements for a software project are either incorrect or incomplete.

An example is your client asking for an automated text message to be sent to them when a user upgrades their account, but this feature was either missed or forgotten during the scoping and confirmation phases. So it never became part of the requirements.

The main reasons requirements bugs arise:

1. Poor communication between the client and yourself
2. Ineffective documentation/requirements-gathering practises
3. Ineffective client confirmation/requirements sign-off

To avoid requirements bugs, ensure you have great communication from the start.

Spend a lot of time deeply discussing the problem and the possible solutions (a project workshop is a perfect place for this), and ensure you are documenting all the outcomes.

I would take a lot of notes, but also do audio recordings of all workshops. The audio is integral as if you are both facilitating the workshop and trying to remember all the discussions and outcomes - this is a monumental task!

After your requirements gathering sessions, go back over your notes right away and flesh them out whilst the conversations are still clear in your head. If you need to, revisit the audio recordings, and then do another round of fleshing out the notes.

I would then create user stories, user personas, and user journeys from all this information - and would use a tool such as Userdoc to centralise it all.

Ask questions right away, confirm everything you need, and then add the answers back into your requirements. The longer you leave this process, the greater chance you miss small details.

Gathering the requirements is one part, but confirming the requirements is equally important.

Send your client a copy of the requirements (e.g. invite them to Userdoc - or send them a digital version), as often they might need some time to look through it all and prepare any questions

Then organise a session with your client where you take them through every requirement one by one, verbally discuss it, and get them to confirm if it seems correct and if anything is missing.

This may take a couple of hours or more, but in my experience, it’s an integral step in avoiding requirement bugs - and setting up the project with the best chance of success.

Bugs in understanding 🐛

These are issues that arise when the requirements for a software project are correct but are misunderstood during the implementation process.

An example is a requirement around “Sending notifications on account upgrade” originally being asked to be a text message by your client, but was not communicated correctly to the developers, so they assumed it was an email notification, so implemented it as such.

If you’re the one implementing the code then this is less likely to occur - as you’ve already gone through the process of avoiding requirements bugs, and you will know the system very well.

But if others are building or helping build the projects, this increases the chance of bugs around misunderstanding the requirements.

The main reasons misunderstanding bugs arise:

1. Poor communication between you and your team
2. Unclear or ambiguous task descriptions
3. Lack of project management processes

To avoid misunderstanding bugs within your team, you need to solid handover process to ensure the project context ends up in your team's head.

Developers and designers are not expected to know every single requirement by heart, as they may not work on them all. But it’s still important to take them through the agreed requirements, so they can understand the larger goals, and be aware of what the end result will be.

If you skip this step and just assign people tasks without educating them on the broader system - then it’s more likely they have misunderstandings.

When breaking down the implementable project tasks, I recommend you link them back to the source of truth in the project requirements, instead of just re-writing them. These requirements are what the client agreed to, and what they should use for testing, so they should also be what the developers use for implementation. If you store your requirements in a manner where each one has a distinct URL (such as Userdoc, a page in Confluence, or Notion), then your assigned tasks can link directly to the specific page.

But no matter how hard you try, misunderstandings still occur - and that’s ok.

The best thing you can do is to catch them early, and a great way (if you're a programmer) is to always do peer code reviews. Before a task can be deemed “complete”, ensure that the code matches the requirements, as besides educating the coders and giving them detailed requirements, the next line of defence in ensuring they have implemented it as planned is the code itself.

Bugs in your code 🪲

These are issues that arise when the requirements are correct and understood correctly, but there is an issue with the implementation of the code.

An example is a requirement around “Sending notifications on account upgrade”. The developer correctly implements the feature to send a text message to the client when a user upgrades their account, however, they make the mistake of also sending a text message on all account status changes, including when the account is downgraded (which was not what the client wanted).

The main reasons code bugs arise:

1. Basic coding/logic errors
2. Lack of attention to detail
3. Lack of personal testing

To prevent code bugs, it's important to have a solid coding process in place, including good personal testing practices, automated testing, peer testing, and peer code reviews.

Basic coding and logic errors can be caught by code linters, but also doubled checked in your code pipeline - so enforcing a team-wide standard of tooling can catch a lot of these low-hanging bugs.

I have a high standard when it comes to development bugs. I’m certainly not “bug-free”, but I’ve spent a long time working on my own testing and debugging practices, and I teach my team to apply the same thoroughness.

Make high-quality code something your whole team strives for, and importantly, holds each other accountable for.

Preventing bugs in software requires a combination of thorough requirements understanding, clear communication among team members, and a solid coding process.

By focusing on these areas, software development teams can work towards creating a product that is as bug-free as possible.

But, having said all that - it's important to remember that no software is ever completely bug-free, and it's inevitable that some issues will arise even with the best prevention measures in place.  That's why it's important to have a tight process for identifying and fixing bugs as they arise so that any issues can be addressed quickly and efficiently - without wasting too much time, and costing too much money.

When starting out, it’s hard to know exactly how much to charge.  And that's totally fine. You learn more from every project and get better and better.

But as your business grows, you will often want to raise your prices. Knowing how to raise your prices is important, as you want to make sure your clients (or at least the ones you want to keep!), don’t feel like you are taking them for granted.

Price increases and client education 📖

As your client services business grows, your costs will likely change, for example

• Realigning salaries to market rate (you may pay yourself too little at the start)
• Investing more in sales and marketing
• Hiring more staff
• Economic inflation

And specifically as a client services business in the technology industry

• Developers cost more as they move from junior to mid-level, to senior
• Devs in certain technologies demand higher wages as popularity grows (e.g. React)
• Staying up to date with technology can involve more professional development time

So it should never come as a shock to your clients that you will eventually raise your prices. But if you never mention your prices will change, then your clients won’t think about it.

Make this clear right from the start.

Let them know you update your prices yearly, and make this a normal part of the conversation.

Talking about money can be awkward - but it’s business, and money cannot be ignored.

If you are doing a fixed-price project, the pricing will obviously not change.

However, they need to understand that just because you charge $X per hour on this project - you won’t charge that forever.

In your proposals, list your pricing as your “current pricing” which reinforces the expectation of future cost changes. Plus it also means a client cannot come back in 6 or 12 months and say “let’s start”, and expect the same pricing that was quoted back then.

For any ongoing contract you work out with your client, ensure you mention your hourly rate is subject to change over time. Don’t make the mistake of being locked into a specific price for many years (unless this is intentional of course).

Letting clients know about a price increase 💬

Say your price is going up $15 per hour. For some clients, this might be annoying, but not a showstopper. However, for other clients, this may be a huge problem. Often you don’t know how a client will react until you tell them - so give them an advanced warning

I would let my clients know 3 months in advance that my agency's prices were going up, and I would only let active clients know. I wouldn’t bother telling clients I haven’t worked with for 12 months, they can find out when (and if) they need more work.

Communicate your price change in an email, and include a small description of why you are increasing your prices.

Don’t just mention “To keep up with staff salary costs”. Describe the benefit to them, for example, “to keep offering competitive staff salaries, and ensure the high-quality standards you are used to”.

Don’t expect you can simply raise your prices, earn more money, and everything turns out perfect - you should be prepared that you might lose some clients.

Raise prices, but keep your clients 💪

If you think there are specific clients that may not respond well to your higher pricing, there are a few things you can do to make the transition easier for them.

1. Offer a gradual price increase
2. Offer a reduced rate to go on a monthly retainer
3. Grandfather clients on your previous price

Offering gradual price increases allow you to increase pricing for long-term or loyal clients gradually over a set period of time (e.g. 6 months). For example, if you are raising your prices by $15 dollars per hour, you could offer them a gradual raise of $5 every 2 months, over 6 months.

This works well with regular ongoing clients who are money conscious, and also encourages them to front-load their work with you, as the more that can get done sooner, the cheaper the rate.

It’s also a great time to get your clients onto an ongoing retainer.

Having clients on a retainer is gold - I am always happy to give them a small discount if they lock in for 12 months or more.

Retainers provide better visibility of your future schedule,  guaranteed monthly income, and create more of an ongoing partner relationship over the transaction project-based work.

It’s also totally fine to “grandfather” client pricing, meaning you keep older clients on older prices, and only raise your prices for new clients.

When I first started my agency, I hated the idea of clients at different prices.

It just seemed messy, and I feared it would add a lot of admin overhead.

Whilst everyone dreams of a productised service business that sells a cookie-cutter solution and basically runs itself, humans are different. Your relationships with these humans are also different, so do what feels right for you.