In a world where companies are under pressure to do more with less, or work smarter with what they already have, the real winners in this space will need to do something that feels counterintuitive: Help us have fewer, better meetings.

The silent killer of work

We’ve normalised a world where meetings are the default setting for getting anything done, and a lot of them are garbage.

This is not just my view, it’s backed by data (sources at the end):

• Only 37% of meetings have an agenda
• 64% of recurring and 60% of one-off meetings happen without one
• Executives consider 67% of meetings a failure
• Individual contributors often spend half their week in meetings they didn’t ask for and don’t understand

Quotes from personal sources:

“The amount of hours I spend in useless meetings is insane.”
—Product Leader, Enterprise

“I slack the organiser just to understand what it’s about.”
—IC PM, SMB

“We waste a lot of time scene-setting because the purpose wasn’t clear from the start.”
—Marketing Operator, Scale-up

We’re 25 years into the 21st century and still wasting huge amounts of time.

Everyone’s building the same AI. That’s a problem.

There are at least two dozen products that transcribe your meetings, summarise what happened, and tag follow-ups. They wrap it in a clean UI, store it somewhere, and call it productivity.

The result? They just make it easier to have more of them 🤦‍♂️.

But none of them fix the actual problem:
The meeting itself is broken.

Before AI can improve our meetings, we need to talk about how it thinks, and why context is everything.

Better inputs = Better outputs

Language models don’t just need data. They need context.

A study on how context affects LLM factual predictions showed that adding relevant information significantly improves model accuracy. AI isn’t magic. It just does better work when it knows what it’s supposed to focus on.

If your AI knows the meeting is a sales call, it can spotlight objections, buyer intent, and next steps. If it’s a design review, it’ll highlight feedback and action points.

But if there’s no title, no agenda, and no stated goal? AI is just guessing.

The missing link: Pre-meeting context

Everyone’s excited about agentic AI, tools that can take your meeting, distill the actions, and just…do the work. But here’s the catch: if the meeting was a mess, the AI output will be too.

Poorly defined goals, vague decisions, and bloated invites don’t magically translate into high-quality execution. They translate into AI agents doing mediocre things with misplaced confidence.

If we want agentic AI to actually be useful, we need to fix the upstream inputs. That means better meetings, or better yet, fewer of them.

Imagine your meeting assistant nudging you with:

• “What’s the goal?”
• “Can this be async?”
• “Why are these eight people here?”

Not just a note-taker, more like a gatekeeper. A pre-check before anyone hits ‘Join.’

The result? Meetings that are sharper. Context that’s crisp, and downstream agents that can actually do their job, because they’re not starting from chaos.

Want smart AI? Start with smarter meetings.

Final thoughts

AI tools are about to flood your calendar with summaries, tags, and follow-ups. But if the meeting itself was aimless? All you’re doing is polishing a turd.

The real frontier isn’t just post-processing. It’s upstream context.

If we want AI to take admin off our plates so we can focus on what matters, we need to feed it clarity before anyone clicks ‘Join.’

The smartest question your AI can ask might just be the one we avoid the most:

“Why are we meeting at all?”

The problem wasn’t organization. The problem was recall. And it hit me: I’m literally using AI assistants daily, so why am I still manually hunting through my notes?

The note-taking paradox

We take notes to remember things. Then we accumulate so many notes that we… forget what’s in them. Ironic, isn’t it?

I’m rather fond of Markdown (it’s clean, portable, and future-proof), which is why Bear has been my go-to for years. But even with tags and search, finding the right note often felt like archaeology, digging through layers of keywords hoping to unearth what I needed.

Meanwhile, I’d been using AI assistants (most notably Claude) more frequently for various tasks, watching them make connections between ideas and synthesize information in ways that felt surprisingly… human.

💡 The lightbulb moment: What if my AI assistants could search through my notes directly?

Building the bridge

For the uninitiated, MCP (Model Control Protocol) is Anthropic’s standard for connecting AI assistants to external tools and systems. It’s what allows AI models to interact with the real world rather than being trapped in a conversational bubble. Read more here

I decided to build a custom MCP server that would connect my Bear Notes database directly to Claude. No middlemen, no cloud dependencies, no sharing my personal notes with third parties.

What’s so great about it?

• 🔍✨ Transforms notes into semantic vectors that capture meaning, not just keywords
• 🤖📚 Claude can now retrieve relevant context from notes I haven’t explicitly mentioned
• 🔒🎉 Everything runs locally, no data leaves my machine, no API faff, no cloud dependencies

How it actually works 🤓

Under the hood, the system uses the Xenova implementation of transformers.js with the all-MiniLM-L6-v2 model:

• It converts each note into a 384-dimensional vector that captures its semantic essence
• These vectors form a searchable index that understands meaning, not just text matching
• When I ask Claude something, it can query this index to find relevant context from my notes

Why this matters

This isn’t just about retrieving notes more efficiently. It’s about a fundamental shift in how we interact with our personal knowledge:

• From active to passive recall: Instead of having to remember what I know, I can ask what I know
• From siloed to connected: My notes are no longer isolated islands of information but an interconnected knowledge base
• From keyword to meaning: Finding information based on concepts and ideas, not just exact words

And perhaps most importantly, it happens entirely on my machine. In an age of cloud-everything, there’s something deeply satisfying about a powerful system that doesn’t need to phone home.

The unexpected benefits

Building this system yielded some unexpected advantages beyond just finding notes:

• Rediscovering forgotten ideas: I’ve bumped into old notes that sparked new thinking
• Connecting disparate concepts: Claude often finds connections between notes I didn’t realize were related
• Removing mental overload: I no longer worry about forgetting where I stored information

Try it yourself

If you’re a Bear Notes user with a penchant for tinkering, the entire code is available on GitHub if you want to dig into the implementation details or contribute to improving it.

The setup is detailed in the README, but the TL;DR is:

git clone https://github.com/ruanodendaal/bear-mcp-server
cd bear-mcp-server
npm install
npm run index  # This creates vector embeddings of your notes
# Configure your MCP settings and you're good to go

The bigger picture

Beyond the practical utility, this project represents something I believe in deeply: technology should augment our thinking without compromising our privacy or agency.

By connecting my notes to AI in a way that keeps everything local, I’ve created a system that feels like an extension of my own memory rather than a separate service I’m querying.

Would love to hear your thoughts or answer questions if you decide to build something similar!

Now user manuals are nothing new, but I’ve found them an invaluable tool to better understand your colleagues.

Let’s all be a little more transparent. Below is my manual.

What I’m like

• I’m a morning person so I do my best-focused work in the morning.
• I plan my day and leave specific time to check emails/slack etc, so if I don’t respond right away I’m not ignoring you on purpose, much 😉
• Related to ☝️ sending me emails titled “Urgent” are completely useless. Email is not a good delivery mechanism for real urgent actions. Doing this tells me that it really isn’t urgent. If my house was on fire I wouldn’t email the fire brigade.
• I can get easily distracted, so work to keep distractions at a minimum.
• I hate being invited to meetings with just a meeting name and no agenda. What is the objective? Meetings with no clear agenda or objectives are a waste of time.

How I work

• I’m a visual person so like to draw or map things out.
• I apply ruthless prioritisation to everything! There are only so many hours in the day if it doesn’t make the top of the list I forget & ignore it. I truly believe if it’s important enough it will come back around.
• When I’m given a large task/piece of work I like to discuss the broad scope and the underlying ‘chunks’ of work. I get lost in big tasks without breaking them down.
• Related to ☝️ I’m great at starting and ending projects/work but can get lost or lose focus in the middle if the goals aren’t clear.
• I get excited by new ideas, but need to ask enough questions to understand the core goal(s) - don’t mistake this for being negative, I just need to understand how things work and how it contributes to the wider company goals.
• I like to do things quickly, ship it, get feedback and then iterate. People have different views on what is ‘perfect’ but we don’t know until we implement it. So ship it, as long as we’re measuring the impact.
• If I’ve got headphones on I’m trying to stay focused on a task. I am happy to be disturbed if important, however, please be considerate.

My leadership style

• I’m a servant leader. My job is to make my team successful, I lead through others.
• I focus on outcomes 🎯, not outputs.
• If you ask my opinion I’ll happily give it but will also give a 1-5 rating. 1. Have no strong opinions to 5. I’m highly opinionated about this.
• I trust by default.
• Strong opinions, weakly held –> Probabilistic opinions, objectively held, decisively executed, diligently evaluated. I have opinions about a lot of things but I’m always up for a constructive debate and am happy to change my opinions provided with new data. Credit Shreyas Doshi

How to get the most out of me

• Articulate the customer value in the task at hand, who is it for and why do they need it?
• Give me some context and the specific objective(s). I almost always want to know why we’re doing it, who is it for and how is it benefiting our customers.
• Be competitor aware but customer focused. Don’t show we what a competitor is doing and say we need to do this. Show me the customer data that informs our own signals.
• I have an “open-door” policy, but please check my calendar before scheduling a meeting and respect my focus time.
• Short meetings or calls can save a lot of back-and-forths.

Communication channels

Consider the channel you use to message me, the below is a guide to how I treat each with my personal SLAs:

• WhatsApp: Urgent P0 level needs [ Response time: ~minutes ]
• Slack: Important short-form communication [ Response time: ~hours ]
• Email: Non urgent long-form communication [ Response time: ~weeks ]

How I communicate

• I prefer direct communication.
• I don’t shy away from difficult conversations.
• I ask lots of questions and like to understand the root cause or the why.
• Lots of context-switching kills me and its impossible to get into flow. Carving out times for deep-work is important to me.
• If I’m particularly deep in work I mute all notifications so I’m unlikely to respond to Slack or emails until I’m done.

• I try as best as possible to prefix messages to support asynchronous communication:

  Prefix	Description
  [ACTION]	There is an important action required in the message
  [MUST-READ]	A must read important message
  [HELP]	A request for help or question
  [FYI]	Sharing information for visibility (non-urgent)

How I give feedback

• I give feedback (privately) as soon as possible while the context is still fresh in our minds.
• By default, I look for how things can improve. This can sometimes come across as overly critical, so let me know how it feels for you. However, I’ll always suggest we ship something and learn i.e. progress over perfection.

How I handle feedback

• I prefer critique over compliments as I like to know how to do it better. I’m very self-critical, however, the occasional praise for a job well done is appreciated.
• I will ask for feedback often as I value continuous improvement and personal growth.

My son, let’s call him ‘L’, has recently turned three and he (amongst other things) is a motivating factor in my life. When he was born my wife and I created a list of the person we wanted him to become:

• Happy
• Faithful
• Positive
• Honest
• Ambitious
• Independent
• Kind
• Empathetic
• Curious
• Confident

But this week he taught me something.

In the last few months L has become more and more independent (nailing one item in the above list 💪). He now doesn’t want any help when taking himself to the toilet. Except there is a problem, for weeks now he has been pulling reams of toilet paper off the roll and it was getting rather annoying.

It goes like this. L

Papa I’m going to the toilet and DON’T help me

Me

OK, no problem, let me know when you’ve finished

A few minutes later, he calls and I walk in to see toilet paper all over the floor. I tell him off because I think he’s playing with it (and to be honest, I just think he’s generally being a 3 year old). Those of you with children will understand.

Then one day, I ask him

L, why do you keep pulling all the toilet paper off?

He responds

Papa I can’t break it

The penny finally drops. He keeps pulling the paper because he doesn’t know how to tear it off! Now I feel like the world’s worst parent, how could I be so naive?

Now you’re probably thinking to yourself, why is this guy going on about his son and what has this got to do with product development?!…

Well, I see this same scenario during product development play out all the time. We make assumptions and try our best to understand customers’ behaviour, we pour over the data, we send surveys and we conduct interviews but we don’t hear what they’re saying.

It’s easy to enter into any kind of customer research with preconceived ideas, personal views or even possible solutions. Just like when crossing a road Stop, look, listen (another child reference, do you see what I did there?)

Next time you’re doing any customer research, take a step back, leave all preconceptions behind and listen. You’ll be amazed at what you’ll hear!

I look forward to another one of his lessons next week.

Remote working can force great new ways of working and in my experience when done well, keeps meetings focused.

Don’t get me wrong, doing remote working well is hard, it takes a shift in ways of working, I think for the better, especially for companies not set up for it and in my opinion many still focus on presenteeism.

Trust your team

It goes without saying you should do this irrespective of being remote or not, however, some people will say things like

What are they up to? They’re not always available

In my experience, the people who ask these questions are typically not measuring success by outcomes but by output. If you haven’t already set some KPIs or OKRs or whatever works for your business but set some NOW! This is a whole other post topic I could talk about.

Schedule your time

It’s easy for people to feel pressured to show they are present. Hopefully, you have a trusting manager and team, so keep focused on your objectives and not how many hours you’re online.

It’s too easy for time to disappear too since there won’t be the typical human interruptions or social queues for lunch or grabbing a coffee etc. Schedule break time, practice the Pomodoro technique or some other time management techniques so you don’t get burnt out.

Turn your video on

Seeing a real person at the other end of a call makes a difference.

Sounds stupid but turn your mic off when not speaking

I can’t tell you how many times people forget this and all you hear is them chomping away at a chocolate bar. It’s impossible to hear each other and you get insanely jealous that you don’t have one 😂

Learn to turn your notifications off when in “meetings”

In most meetings, I try not to have my laptop open and I take a notepad. However, now you’ll be on your laptop 100% of the time. Learn to turn off your notifications so that you stay in the room and aren’t distracted by constant email or slack notifications.

Over communicate

I can’t say this enough but over communicate everything. Write up meeting notes, send clarification messages and make time for personal updates.

Make time for pleasantries

Being remote can remove some of the human element out of communication. Take time to ask how people are and listen before you jump straight into work mode.

Learn with your team

Every team is different, we all work slightly differently but make time to discuss remote working techniques that you’d like to try. Some simple things I’ve done:

• Socialise virtually - create a meeting whose sole purpose is to catch up over coffee.
• Make sure there is one facilitator, who should ensure each person in the meeting has a say.
• Create a Google Doc for each meeting and have everyone collaborate on it in real-time.
• Make space for things to be parked; items that can be discussed later

Get out the house

Make time to get out of the house where safe to do so. Go for a walk and get some fresh air (whilst following social distancing or your local advice).

Working at home for extended periods can be hard if you’re not used to it, but create a routine and get some fresh air.

There are so many things you can try as an individual and as a team; test, learn and adapt….what other choice do we have?

Whether you evaluate your priorities as Urgent vs Important or Impact vs Effort or use a scoring model, there is typically one element missing: courage.

So what do we do?

The world we live in is full of distractions. Every device we own is vying for our attention, we get invited to meetings we may not need to attend and we receive emails that we most likely don’t need to read.

We try to do it all. We pretend we’re prioritising. We spend our time trying to please others while holding on to our belief that we’re being effective. We create to-do lists that keep rolling over to the next day and when we cross off a bunch of items on that list we think we’ve been successful.

The reality is, you haven’t.

The most common answer I’ve heard for having not done something, or succeeded is time. You had time, but at some point, consciously or unconsciously, you decided to prioritise something else. We say things to ourselves like:

I won’t start that now as it’s too hard, but I can do these two things in 30 mins instead.

Doing the right thing

Time management is not just about getting more done, it’s about getting the right things done and ignoring everything else.

It takes courage to put your ideas out there, courage to say no, courage to ignore everything else in the world you deem less important and courage to stay focused until it’s done.

How do you prioritise?

Be honest with yourself, do you choose to only spend time on the most important things?

Do you have the courage to make the difficult decisions about priorities and stay focused until it is done?

It’s not the prioritisation framework that’s the problem, it’s you!

We all have meetings (some more successful than others); we can’t escape them. However, meetings aren’t always bad. Well structured agenda’s, good timekeeping and clear actions make meetings more successful, on the whole.

But what about the discussion: the core part of the meeting where opinions & facts are debated? These can take longer than is really required, and for the majority of the time this is due to individuals debating something that they are actually less opinionated about.

We also solicit feedback directly, but it’s sometimes difficult to know how invested the other person is in the feedback they’re giving. Let me give you an example: in a design critique we’re asked to critique a specific piece of work and a designer might receive the below sort of feedback.

This carousel feels clunky, can we make it smoother

I prefer this button to shake when it can no longer be clicked

How is the designer expected to know which of these to action? Sometimes people default to the most senior person in the room (which is all sorts of wrong), action everything, or only action the feedback that they personally agree with.

So how do we resolve this?

We ask a simple question: how much do you really care between 1-5? (1 = less, 5 = more)

If we identify that a discussion is about to start, we ask this question first. If both or all individuals are <2 we go with the initial recommendation, or make a decision then and there.

If people are at opposite ends of the scale, the person at 5 gets to drive the conversation.

If both are at 5 we either continue the discussion because getting to a decision is the highest priority, or we schedule a separate discussion as to not disrupt the flow of the current meeting.

In the feedback scenario, imagine the same feedback but with a clear indication of how much that person really cares:

• “This carousel feels clunky, can we make it smoother”

  I’m a 4, this feels weird against all of our other interaction patterns

• “I prefer this button to shake when it can no longer be clicked”

  I’m only a 2 on this so take it with a pinch of salt

What have we found

By doing this, we’ve found we just spend less time on things that people weren’t that passionate about or that don’t require lengthy discussions.

So saving us time, sanity and more importantly means we can focus on the things that really are important.

Give it a try, does it work for you?

Previously, Pusher hadn’t had anyone dedicated to this, either out of necessity or due to a lack of time or resources. As we’re scaling, we’re now in a position to bring visual communication front of mind and align all of our efforts in a much more unified way.

These views had been widely held in our team, but hadn’t been documented; it became obvious when onboarding a new hire that we needed to better communicate our process.

Painting

When you start a new piece of work, how do you go about determining the message and the feeling you want your users to be left with when experiencing your product?

Anything a user experiences on your site will be communicating something; what users see, hear, and feel can drastically change their perception of your brand and your product.

So how do we make sure we’re doing our best to communicate the right thing at the right time?

In a post by Stewart Scott-Curran, he talks about thinking in terms of a sliding scale:

“It’s useful to think about your visual identity on a sliding scale that you can dial up or down depending on what you’re producing.”

For us, that means deciding where we sit on a scale from serious to playful for a given project. We’ve taken this a little further and started articulating our visual identity across three possible points on a scale. Level one is articulated as more serious and less playful; level two as a balance between the two, and level three as more playful.

So starting with level one (our core message), what impression do we want our customers to be left with, and how do we articulate this on our sliding scale?

Start with the customer; what is their primary objective when visiting our site? We articulated it using job stories, starting with the below:

When I visit Pusher.com, I want to feel as though the products are suitable for creating professional, scalable, and reliable applications, so that I’m confident it will provide value to my customers faster than if I were to build it myself.

*yes this is only one JTDB, to keep the post readable 😉

We created a list of words that came to mind when thinking of the above:

• Hosted
• Ambitious
• Delightful
• Reliable
• Premium
• Scalable
• Secure
• Simple
• “Just works”
• Fast

This helped shape our thinking about our constraints/position on a sliding scale for “level one”:

• More geometric than organic
• More contemporary than retro
• More serious than playful
• More realistic than abstract
• More calm than energetic
• More simple than detailed
• More bold than subtle
• More obvious than clever

None of these are absolute; as they exist on a sliding scale they can be explored throughout the work, whilst ensuring that the above statements stay broadly true.

Critiquing the painting

By articulating our visual treatment like this, it has helped us have more fruitful discussions. No longer are we debating every piece of work in its broadest sense; we have some constraints that help us align our visual communication in a way that we’ve not done before.

This has lead to faster turnaround times and greater consistency in our visual design, interface design, and animations. Too often we discuss these things, but rarely document them, or share and refine them. Doing so enables much greater collaboration, shared understanding and alignment. Like most companies in a similar position we have more work than people available to do the work, so less time re-debating and more time executing is a good thing in my book.

Full disclosure: we still need to articulate our other levels, and will do so when we have work that doesn’t require “level one” type treatment. But as an example: non-transactional marketing emails and newsletters are probably much closer to levels 2 & 3 (more playful than serious) than a level one.

A picture really can paint a thousand words, and because it can, we should be as rigorous with our “pictures” as we are with our words.

Feedback

How do you think about your visual communication? Have you ever thought about it? If you do, or don’t, I am really interested in hearing about your experiences or processes.

We—the Workshop at Pusher—are a newly formed team of three, made up of individuals who have experience in Frontend Development, Design, Product & Project Management.

Our mission…see previous post whoami.

Setting the scene

My background is in Product Management leading large teams of other Product Managers and Developers. Now in those teams we always had a PM, Scrum Master/Tech Lead and Engineers (some even had a BA). These were always within the Scrum defined magic number 7 +-2.

We created user personas, used Jira and had all the typical Scrum ceremonies. We were successful: we delivered on time, on budget, and within the agreed-upon scope—in short, we were a well-oiled machine.

When the Workshop was formed I set up a similar Scrum process: we scheduled our rituals, set up our tools, and agreed on how the process would work. We were confident that we’d soon have a well-oiled machine of our own. We were wrong.

Why did it fail?

It killed productivity. All of my previous experience was in larger companies with much less ambiguity and more tightly defined roles & responsibilities. Pusher is a startup (well, probably a scale up now) and we don’t have these luxuries. (Yes, we’re hiring—check out our careers page and let me know if you’d like to join our team)

With all of our best efforts we spent too long in planning, we didn’t have enough detail to estimate with any level of accuracy and there wasn’t anyone who can refine requirements. We’re a small team with no PM, no Scrum master and definitely no BA.

So what have we changed?

For smaller changes we’ve shifted our focus to explicitly allocate a fixed amount of time for the work. With our larger initiatives we’re experimenting with Design Sprints.

We made sure we all share the same context of the work. We over-communicate about everything—from company strategy, to OKRs, to our goal-oriented roadmap, and all the way down to individual blocks of work.

We’ve changed so much as a company in the last 12 months that our brand identity, communication style and product offerings have all changed. We have only recently defined these and experimented enough to feel confident in our choices.

So far so good, we’re shipping value to users and the business daily.

What the future holds

Now this process won’t work for us forever, however, for where we are as a team today this is producing the best results. I know in 6 months or so as we grow, as new people join the team the process will change again.

With every individual that joins it brings a new perspective, a new experience and a new opinion.

The key lesson here is that even if we have relevant experience it doesn’t mean we can just apply the exact same process to a different environment. We use our experiences as tools to shape the future, every person is different and what works for one set of individuals rarely works for another.

One size definitely does not fit all.

To support, elevate, and connect the ideas of others; to create and act as the custodians of the Pusher identity, and; to create powerful systems that inform and scale our impact.

On a daily basis, I’m having really thought provoking conversations (I would say that though, wouldn’t I), and wanted to start sharing my thoughts on various subjects to trigger discussion, learn new things and allow my assumptions to be challenged.

I’ve never written a blog or shared my thoughts like this externally before, but I wanted to push myself to try something new. If nothing else, at least I’ll have a great documented list of ramblings to look back on 😂 .