Hey Everyone!

Hope you all had a great Memorial Day weekend! It was a rainy one here in Charlotte, but honestly I'll take it. We've been in a pretty serious drought the last few months, so a soggy long weekend actually felt great! 

Now we are officially in summer mode and it's easily one of the best times of year in Charlotte.

💼 Fun Fact: According to data from 50 B2B SaaS keywords tracked in Q1 2026, websites holding top three search rankings saw click-through rates drop 18% to 34% once AI-generated answers appeared above them, even though their rankings never changed.

📣 Shout-out: To Emily, founder of LS Adaptive, who created a brand making thoughtfully designed everyday clothing for anyone navigating disability, recovery, aging, or mobility challenges. What she created is pretty cool.

— August Shah

Last issue I shared the Hiring Scorecard, which works great once you've already talked to a candidate and need a way to actually compare them “apples to apples”.

That is only half of the equation.

So here is the other half of the equation: Hiring Toolkit. 

Inside you'll find:

Take it, make it yours, and if you build a better version, send it back so I can steal yours hahah (just kidding).

I'll be honest, this is something I didn't pay nearly enough attention to in the early days of PMG. Overhead was basically that thing in the back of my mind that the bookkeeper handled and thats no way to run a business.

When I talk about overhead I am not talking about payroll, rent, or your big software contracts. I am talking about the indirect expenses. Unbillable hours, rework not priced into the project, the meeting that had four people that didn't need to be there, and unused software stacking up across the team.

None of this has that “wow this needs to change ASAP”.  It tends to just show up over time and one day you notice your margin is worse than it was a year ago, even though revenue is up.

The easiest place to start is what you can see. Pull your last three months of card and bank activity and go line by line, paying extra attention to the small recurring charges. The first time we did this at PMG I found a tool I genuinely could not remember signing up for. The harder part is the stuff that doesn't live in a statement.

Then comes the part that gets missed. Cutting. Go back through what you found and ask one question on each line of if its viable to keep. If the answer is no, or even a maybe, cancel or iterate. Put the review on the calendar quarterly and have someone thats not you run it. Not fancy or glamorous, but every time we have this review, we have come out with a better picture of the business and more money in the bank.

Answer:

The signals were probably there early on, you were just looking in the wrong direction. These are the two I have learned to watch for.

Step 1: Same risk, no movement.

Compare your last few status reports side by side. If the same risk has been sitting there each time with nothing really progressing, that is your tell. It can mean the person closest to it is either hoping it resolves itself (it won't) or could be avoiding a hard conversation they do not want to have. Meanwhile the rest of the project keeps moving, and by the time the risk is dealt with, it has worked itself into a couple other workstreams. What could have been a quick conversation is now a problem to untangle.

Step 2: Don't shoot the messenger.

This is a leadership problem, and It affects every project at the organization. If your team thinks leadership is going to react negatively, you stop getting flagged risks. The way to know whether you have this dynamic is to look at how you and your project leaders react when somebody brings bad news. If you tense up, get short, or jump straight into "why did this happen" before you have heard the whole story, that is the dynamic your team is reacting to.

The point of this is to stop relying on what's directly in front and focus on the other factors. By the time a project is on fire the options are limited. And in my experience, the earliest signals are almost never on the spreadsheet, they are in how people are acting.

Have a question you want answered?

Submit a Question

Last issue I talked about Viktor and how useful I found it, which is still very much true. For smaller use cases the Slack-native experience is honestly fantastic and great at handling tasks while you are on the move.

The problem is once we started leaning on it across the org and running multiple agents, the costs got out of hand pretty quickly. By the end we were spending close to $700 a month on Viktor and still hitting walls anytime we wanted to build something a little deeper. So over the past month we moved over to the Anthropic ecosystem. For comparison, we are now spending around $300 a month with more output for less spend, including the more complex workflows.

What I found is the ecosystem is actually built to scale in a way that works for a small business. You can research and prototype something in Claude Chat, hand it off to Claude Code to build into a real thing, and then connect that to whatever else you already run. We have routines, internal tools, and automations going across Airtable, ClickUp, Slack, our website, and a few other places.

Pro tip: Pick the most painful, repeatable thing you do every week and just put an agent on that. The mistake people make is trying to AI their entire business in week one, and what they end up with is a pile of nothingness and a bunch of AI slop. Be intentional.

That's a wrap! I'm going to spend the rest of the week trying to convince an agent to do my inbox triage for me (without using CAPS). Wish me luck.