For the last four months, we’ve created a dedicated space at IMPACT for our team to get together, train each other, and learn.
We call it the Expert Exchange. It happens every Monday at 3:30 Eastern, and it’s 90 minutes. Yes, 90 minutes. I know that seems like a lot of time to bring an entire company together every week, and it is.
But for the business we’re in, and with how fast everything is changing, we need it to stay on the cutting edge of AI and to be as valuable to our clients as we need to be.
We start with show-and-tells. Two or three people get five to ten minutes each to share something they’re doing with a client, or experimenting with on their own, that the entire company should know about.
Then we move into a training, facilitated by someone on our team or a guest.
In October, I was leading a session on building with AI.
I asked everyone to write down an answer to a question: “What’s something that you heard AI can do that you’ve wanted to try but you haven’t tried to build yet?”
Then I gave them an hour to figure out how to build it, with everyone together on Zoom, chatting, supporting each other, and having fun.
For many people on our team, it was the first time they had built anything with AI. Some played it safe, and plenty spent the time working out where to begin.
We still had wins. Michael Accuosti, our controller, had never built an AI agent. By the end of the session he had started on his first one, a financial analysis assistant he was testing with sample data to help identify what he might be missing.
It was a good first step, and we wanted to see how far people could go with a second try.
This past Monday, four weeks later, we did it again. An hour this time. Pick something to work on, build with AI, and show us how far you get.
The entire company went straight to work on real problems, and the leap from October was obvious. Here’s what five people showed:
Brian Casey, one of our coaches, worked on a manufacturer’s quoting problem.
Every job this client does is custom. They’ve been hesitant about self-service tools because they couldn’t see how a pricing calculator would work for their business.
Brian had been talking with them about where sales were breaking down. Their HubSpot reports showed lost deals because they couldn’t develop a solution fast enough, didn’t receive drawings in time, or couldn’t meet the technical requirements.
Their quoting process also created delays. A buyer would get on a call, learn what information they needed to provide, then go gather it. The team would have to follow up to move the quote forward.
Brian built a prototype called a “Quote FastPass.” It collects project details upfront, flags timelines the manufacturer likely can’t meet, and helps the buyer book a walkthrough call. He also showed a view for the sales team to evaluate fit, identify missing information, and prioritize promising projects.
By the end, he had something he could put in front of the client to show how self-service could work for their custom business.
Stephanie Baiocchi, who leads our Endless Customers Live event, worked on helping attendees get approval to come.
Someone interested in attending often has homework to do before they can buy. What will the whole trip cost? What if they bring three people? How do they explain the value to their boss?
Stephanie started a “Plan Your Trip, Convince Your Boss” tool during the session. It lets someone choose their team size, tickets, travel plans, hotel nights, and meal budget to estimate the investment.
Then they select the business problem they want to address by attending, and it drafts a letter to their boss. They can copy it, email it, or save it as a PDF.
She had a working version to demonstrate, with some details still being refined. It brought the budget and the approval process into one place.
Austin Mock, who helps clients improve their video content, worked on making video reviews more useful with AI.
We’ve had success using AI to review written content. Video has been harder. Reviewing a transcript leaves out what’s happening on screen, and those visuals affect whether a video holds attention and builds trust.
Austin explored a Gemini-based review process that considers the visuals and audio together. Then he tested it on an eight-minute video he’d already reviewed himself.
It flagged a text-heavy graphic, suggested ways to make the video more engaging, and recommended additional footage that could strengthen it without a full reshoot.
Much of that matched the feedback he’d given manually. He was clear that it needs more work before he shares it broadly, but he made progress on a task that takes considerable time.
Ashley Jensen, who leads our content efforts, worked on coordinating a growing group of contributors.
We have newsletters, articles, webinars, and podcasts in production, and a lot of people across the company who want to contribute. They need to see what’s planned and where they can jump in.
Ashley built a shared content calendar using information from ClickUp. It shows what’s being published and when, along with open webinar and podcast opportunities.
She demonstrated how a contributor could claim an open article topic or pitch an idea that becomes a task in her queue.
She still needs to test sharing it more widely, but the pieces are coming together to make participation easier and reduce coordination work.
Allison Riggs, one of our coaches, worked on a question she keeps hearing from home services clients.
They want to know whether they’re appearing in Google’s online estimates and what to do if they aren’t. Allison had been explaining how to check, and she wanted to give them a repeatable process they could follow.
She built a tool where clients enter their business name and location, receive suggested searches, and record where they appear. The checks still require the client to run the searches themselves.
Their answers feed a score and a report with suggested priorities, competitor comparisons, and a reminder to retest in two weeks.
She turned her recurring instructions into a guided tool clients can work through.
What I loved about these examples was how closely they connected to the work our people already do. Each person understood a problem and used the time to make progress on it.
And everyone watching could see possibilities for their own work.
Brian’s tool could apply to other manufacturers. Ashley’s approach could help another team coordinate contributors. Austin’s experiment gave the rest of us something to learn about reviewing video.
That knowledge spreads because we’ve made time to share it.
Most companies have trainings, but they’re typically about company policies. How many are creating space for their people to learn, share, and develop professionally on a regular basis, especially when it comes to adopting AI?
A weekly 90-minute meeting may be too much for your business. I get that. Every other week or once a month might be a better starting point.
Here’s how I’d start:
If we want our companies to become more valuable and more AI-forward, we have to give our people time to learn, practice, and help each other improve.