How we built Lake Effect in two weeks

Lake Effect started as an AI event.

That was probably mistake number one.

Not because AI was the wrong topic.

AI is a huge part of why Lake Effect exists.

Agenda.

Sessions.

Learning objectives.

Speakers.

Rotations.

Tickets.

Sponsors.

All the normal event stuff.

What I actually wanted to build was a room.

A room filled with smart people who were willing to be honest with each other.

Not just the wins.

Especially not just the wins.

I wanted us to talk about three things:

That became the philosophy behind Lake Effect.

And in a slightly ridiculous twist, those same three questions describe the process of building Lake Effect itself.

So before the first ferry even leaves Toronto for the island, here is my first Lake Effect contribution.

The messy story of how we built it.

Mostly from my phone.

With AI sitting beside me.

While changing our minds repeatedly.

And somehow filling a room faster than I expected.

This may have been the most important decision.

I could have spent three months developing the perfect concept.

Instead, we put something into the world.

The original Lake Effect was much more structured.

It was aimed primarily at accounting firm owners.

The theme was essentially:

There were versions with tables of five.

Versions with conversation stations.

Versions with four structured rotations.

At one point the event was built around 75 potential attendees.

The original ticket concept was also much more premium.

$397 early bird.

$497 regular price.

Dinner.

Open bar.

Ferry.

A private evening at the Royal Canadian Yacht Club on Toronto Island.

On paper, it made sense.

But then something more useful happened.

People reacted.

And we listened.

That is where Lake Effect really started.

The first major breakthrough came when I realized Lake Effect should not really be an AI conference.

People can watch an AI webinar anywhere.

They can watch one while eating cereal.

I did not need to bring people to an island to give them another presentation.

What people were missing was connection.

So we changed the event.

Instead of building a conference about AI, we started building an evening for entrepreneurs and trusted advisors.

Accounting firm owners.

Bookkeepers.

Entrepreneurs.

Technology leaders.

Students.

People building things.

People trying things.

People willing to talk about the messy middle.

That shift mattered.

The island was no longer simply the location.

It became part of the experience.

Get on a boat.

Leave the city for a few hours.

Sit around a table.

Eat a great meal.

Have a drink.

Talk to people you might never normally meet.

Then come back across the water with a few new ideas and hopefully a few new friends.

That felt much more human.

The original pricing was $397 early bird and $497 regular.

Then I asked myself a different question.

What if this simply cost roughly what someone might spend on a really good night out?

We moved the public ticket price to $297.

Then we experimented.

Was this a perfectly designed pricing strategy?

No.

Did it help us discover where the market was?

Absolutely.

And something else started happening.

People began buying seats for other people.

That changed everything.

I originally explored a fairly normal sponsor strategy.

Technology vendors.

Accounting apps.

Partners.

Packages.

Support opportunities.

I did not want booths around the room.

I did not want someone buying five minutes of microphone time.

I did not want people sitting through commercials between courses.

The better model became:

ApprovalMax sponsored seats for CPA students.

Paul Jackson from Method decided to join us.

Other people offered to buy or donate seats.

That felt much closer to the spirit of what we were trying to create.

Instead of:

Look at our logo.

It became:

Meet this person.

One of my favourite decisions was creating seats for future CPAs and CPBs.

At first, Lake Effect was mostly about established owners and leaders.

But if we are talking about where accounting and technology are going, why would we leave out the people who will actually inherit the profession?

We started opening sponsored seats to students.

Suddenly the room was not just people who already had the answers.

It included people asking completely different questions.

That matters.

Jim Harris is coming.

Jeff Cates is coming.

Chris Fudge is coming.

Experienced firm owners are coming.

Bookkeepers are coming.

Technology leaders are coming.

And students are sitting at those same tables.

There were many versions of the room.

One long table.

Small tables.

Nine conversation stations.

Rotations.

Eventually I kept hearing the same concern.

At a long table, you really talk to the person on your left and the person on your right.

Maybe the person across from you.

That is not what I wanted.

So we went with five round tables of eight.

Forty people.

And we started designing movement into dinner.

Appetizer with one group.

Entrée with another.

Dessert with another.

Not speed dating.

Not networking bingo.

Just enough movement so that by the end of the evening you have actually met the room.

We then started intentionally balancing those tables.

Industry people.

Firm owners.

Students.

Bookkeepers.

CPA4IT team members.

Different perspectives.

Different personalities.

Even making sure people with history between them did not end up sitting together all night.

The seating chart stopped being logistics.

It became part of the experience.

This experiment would have been almost impossible for me to execute this quickly without AI.

But not because AI came up with one magical prompt.

There was no:

prompt.

AI helped because we used it hundreds of small times.

Most importantly, AI let me iterate ridiculously quickly.

I could have an idea while walking somewhere.

Open my phone.

Talk it through.

Challenge it.

Rewrite it.

Post it.

See what happened.

Come back.

Change direction.

Repeat.

AI did not build the community.

People did that.

AI reduced the distance between having an idea and testing the idea.

That distinction matters.

Another thing that worked was letting people watch the event come together.

We did not disappear for six weeks and emerge with a polished campaign.

We shared the messy middle.

Four early supporters.

Then eight.

Then twenty something.

Then over thirty.

Then the room changed.

Then capacity changed.

Then more people joined.

On September 15 we were at 36 of 40 seats filled.

Then forty were spoken for.

Then we had another problem.

We had roughly 45 people for a room designed for 40.

I was suddenly asking the club whether we could squeeze another round table into the room.

That is objectively a much better problem than trying to fill thirty empty chairs.

But I think people enjoyed watching it happen.

They were not just buying a ticket.

They were watching something get built.

Some of them were helping build it.

At the beginning we talked about limited seats because that is what event marketing tells you to do.

Later, limited seats became an actual operational problem.

That changes the energy.

We stopped needing to push.

I even stopped wanting aggressive calls to action.

No more constantly asking people to comment “curious.”

No manufactured panic.

The conversation shifted toward:

We are full.

We are trying to make room.

Here is the waitlist.

That feels much better to me.

Some of the most interesting reactions came from people who said:

I wish this was closer.

That got me thinking.

Maybe Lake Effect is not one event.

Maybe it is a format.

Vancouver already has a possible name:

Rain Shadow.

Calgary could become Chinook.

Other cities could have names connected to their own geography and identity.

I joked that if enough people wanted one in Saskatchewan in winter, apparently I would be going to Saskatchewan in winter.

That may still happen.

But that idea only exists because we launched something before knowing exactly where it would go.

Everything was brilliant.

Every strategy worked.

It was all part of the plan.

We made lots of mistakes.

The first Lake Effect had too much curriculum.

AI adoption.

Workflow documentation.

Automation frameworks.

Structured sessions.

Rotations.

Exercises.

Thirty day plans.

Those are useful things.

They may still appear in future Lake Effect experiences.

But I was trying to justify the event with content.

The venue was already telling me what the event wanted to be.

An island.

A ferry.

A beautiful club.

Round tables.

Dinner.

Conversation.

The content did not need to carry the whole experience.

People did.

$397 and $497 were defensible.

This was a premium venue.

Food.

Drinks.

Ferry.

Small audience.

A curated experience.

But it made the decision feel bigger than I wanted it to feel.

Once we moved toward $297 and introduced community pricing, the psychology changed.

The question became less:

Can I justify buying this professional development event?

And more:

Would I spend this on an interesting night with interesting people?

That was a much better question.

It worked.

But if you asked me exactly which variable caused ticket demand to accelerate, I could not prove it.

That is one of the dangers of building the airplane in the sky.

Sometimes you land it.

At different points Lake Effect was built around:

75 people.

45 people.

40 people.

Then potentially 48.

That is too much movement.

When I finally locked five round tables of eight, the event became easier to understand.

Then we filled them.

Then I accidentally kept going.

Suddenly we had about 45 people attached to a 40 person room.

A nice problem.

Still a problem.

Capacity should have been locked much earlier.

RCYC is not just another downtown hotel.

That is part of the magic.

It is also part of the operational challenge.

Cherry Street.

City Station.

RCYC launch.

Not the Jack Layton Ferry Terminal.

Different departure times.

A possible Foot of Yonge option depending on weather.

Guests asking whether they needed to book the ferry.

None of these mistakes destroyed the event.

But every confusing logistical email spends a little bit of trust.

Next time I want transportation to be boringly clear.

Tables of five.

One long table.

Nine stations.

Five round tables of eight.

Rotations by course.

Hosts.

Anchors.

Movement.

There was a lot of design energy spent getting to something pretty simple.

People want to meet people.

Round tables are good for conversation.

Moving once or twice helps.

That is basically it.

The final concept is much simpler than the journey we took to get there.

Some responded.

Some joined.

Some sponsored.

Some conversations moved slowly.

The lesson was not that vendor outreach was bad.

It was that personal trust moved faster than sponsorship packages.

The people who understood what I was trying to create did not need a twenty page prospectus.

They needed to hear why the room mattered.

There was a point where I caught myself doing the normal event marketer thing.

Countdowns.

Seats remaining.

Comment CURIOUS.

Discount deadline.

Another CTA.

Another graphic.

Another scarcity post.

It worked to a point.

Then it started to feel wrong.

Once the room was filling naturally, we stopped pushing so hard.

That felt much more like Lake Effect.

This is worth saying.

AI makes execution faster.

It does not automatically make decisions better.

If your assumption is wrong, AI can help you produce a gorgeous campaign based on the wrong assumption by lunchtime.

We created versions of the event that we later abandoned.

Copy we replaced.

Visual concepts I hated.

Images that felt too artificial.

Fonts that looked boring.

Ferry graphics that looked childish.

Messaging that felt too salesy.

The advantage was not that AI got everything right.

I would ask:

Who needs to meet each other?

before asking:

What should I teach them?

That one question would have saved me a lot of early planning.

The people are the product.

The food, venue, speakers and activities help create the conditions.

But the room matters most.

Choose the room.

Choose the tables.

Choose the maximum.

Create a small buffer.

Then sell to that number.

Forty seats should not somehow become forty five unless the room is already capable of holding forty five.

Apparently this needs to be written down for me.

The early Lake Effect attendees did more than buy tickets.

They helped shape the event.

Next time I would formalize that.

Your first ten or fifteen attendees could become the advisory group.

Ask them:

I learned a lot from FirstWave, CatchTheWave, FriendsAndFamily, sponsor seats and student seats.

But there were probably too many versions.

Next time I would make the economics much easier to understand.

One normal ticket.

One early supporter price.

Sponsored community seats.

Done.

No booths.

No commercial breaks.

No forced pitches.

If a partner wants to support the community, help us put the right person in the room.

Sponsor a student.

Sponsor an emerging firm owner.

Bring a customer.

Bring someone interesting.

Then sit down and eat with them.

That is Lake Effect sponsorship.

One link.

One map.

One address.

One sentence saying:

This is NOT the Jack Layton Ferry Terminal.

Launch times.

Transit.

Parking.

What happens if you are late.

Weather backup.

Done.

Absolutely.

But I would use it even more deliberately.

Not as the person making the decisions.

As the person constantly asking me:

That is where AI becomes far more powerful than a copywriting machine.

AI should not just help us produce more.

It should help us learn faster.

Three questions.

That is it.

Imagine if accounting firms shared like that.

Technology companies shared like that.

Entrepreneurs shared like that.

Students shared like that.

Instead of everyone pretending everything works perfectly.

We would all learn much faster.

That is also what I want #GPTEPD to become when the daily experiments start October 1.

Not polished case studies.

Experiments.

Some good.

Some bad.

Some embarrassing.

All useful.

I started this project thinking Lake Effect was an event.

I am starting to think it might be something else.

Maybe Lake Effect is a community built around being willing to say:

And then somebody else takes that knowledge and tries something better.

That is how a small idea creates a bigger effect.

Put the right people together.

Create enough trust for them to be honest.

Let ideas collide.

Then send everyone back into their businesses and communities carrying a little bit of what happened in the room.

That is the Lake Effect I want to build.

And this launch has already taught me one more thing.

You do not need to know exactly what something will become before you begin.

We started with an AI event.

Then we changed the audience.

Changed the message.

Changed the price.

Changed the format.

Changed the room.

Changed the tables.

Changed the sponsorship model.

Changed the marketing.

And kept listening.

Less than two weeks later, forty seats were spoken for.

Then somehow we had forty five people trying to fit into forty chairs.

There are worse problems to have.

The event itself has not even happened yet.

So I have no idea whether Lake Effect will live up to everything I hope it will be.

And that is kind of the point.

This is still an experiment.

On September 30, we run it.

Afterward, I will come back here and share the same three things.

What worked.

What didn't.

What I would do differently.