20 years combined tech and asset management. Learning and sharing about entrepreneurship and business building. https://xitter.gq/t.co/hB3DddxTQA. NFA
Joined June 2024
- Tweets5.4K
- Following831
- Followers1.1K
- Likes2.1K
While there's no shortage of business podcasts out there, you don't often get a detailed breakdown of how a home service business can get to $60K/month after just two months.
A floor coatings entrepreneur was on Koerner Office (@mhp_guy ) a few months ago to describe how it happened:
> Tile installer made lateral move to concrete floor coatings
> Didn't wait to become an expert - taught himself by studying online + experimentation
> Simple advertising message
> Offered reasonably detailed breakdown of average ticket size, square footage, ASPs, implied volume of jobs required, etc to reach the $60K/month
Worth a listen.
playbooks.beehiiv.com/p/60k-…
🤖 Made with AI
Based on the commentary from the discussion, a potential pro forma margin improvement from tech and AI implementation in the median US machine shop yields some pretty incredible possibilities:
🤖 Made with AI
Manufacturing is one of the sectors that has the most to gain - $4.0Tn - from implementing AI.
It's lower margin (than other sectors) + multiple systems and processes spanning both humans and machines in the loop = lots of 'surface area' for possible improvement.
Terrific podcast on this from Daniel First of Axion in an episode of Inspired last November.
Key takeaways:
> Manufacturing is a $30T+ annual industry; 10% of revenues are lost to waste and quality problems
> First defines "quality" as more than avoiding defects - it's when supply and demand meet. Hence quality problems occur when insufficient customer feedback loops and data create mismatches
> For manufacturers, implementing AI is as much as enabling field service reps to gather and report on warranties, reliability, and customer feedback to create a self reinforcing loop of input that can drive product, inventory and resource allocation decisions
Great episode and worth a listen, especially for entrepreneurs focusing on manufacturing.
playbooks.beehiiv.com/p/manu…
For the full episode:
youtu.be/bhQQpYe4HsQ?si=p5Nd…
Yes the facts are very supportive.
And much of this growth began many years ago (pre AI).
Entrepreneurship will become a much more common path .
Kelso is one of the best 'career pivot to entrepreneurship' stories you'll ever hear.
Steve's shared his thoughts several times over the last 1-2 years, and I learn something new each time.
The key takeaways for me was how he achieved the sweet spot of the S curve of growth, around 2022, related to how he structured acquisitions starting with his 2nd deal.
This type of creativity is perhaps a necessary (but not sufficient) condition to achieve a ~10x'ing of revenue growth in five years (~$17m to $1.5Bn) .
More here:
playbooks.beehiiv.com/p/the-…
Each of the podcasts where Steve Carroll shares the story of Kelso (Owned and Operated, Buyers and Builders, Acquiring Minds)
youtu.be/Ru9nyK8_nAU?si=n6AY…
youtu.be/WLlvYnPmApw?si=3C5I…
youtu.be/WLlvYnPmApw?si=3C5I…
@WilsonCompanies @PrivatEquityGuy @StevenHartC
Private equity-backed Kelso Industries is aggressively rolling up MEP companies
~40 deals and ~$1.2B+ revenue in 5 years
~$130M - $150M EBITDA (est.) for an MEP platform of Kelso’s size and scale
MEP platforms trade for 7.5x - 9.5x+ EBITDA in this market and I would put Kelso at the upper end of that range
Enterprise Value of Kelso (estimated)
Low: ~$1.15 billion
Mid: ~$1.25 billion
High: ~$1.35 billion
Incredible journey for a 5 year old platform
M&A updates in commercial & industrial HVAC, plumbing, & electrical
👇🏼
SchryverCo.com
SMB M&A base rates base rates can be helpful for aspiring entrepreneurs.
Peter Lehrman of Axial was on Minds Capital in January 2026. Key takeaways:
> Each year 20k-40k attempted SMB M&A deals - only 25% (!!) close
> Strongest interest for B2B and industrial services
> Most deals complete at just below 7.0x EBITDA - we can debate if that's high or low (seems high to me)
> Indy sponsors have become the most active buyer cohort by closed deal activity, growing 29% CAGR over recent years
I added a second chart shared earlier today by @synopsi on base rates of the categories that see the largest companies
Charles Wang and Mark Leonard both pursued enterprise software to build massive companies and personal fortunes.
But their approaches were fairly different. One built a centralized process for acquisition integration, the other was highly decentralized.
Why could they both win? I’d argue it’s due to the quality of the underlying asset - software’s stickiness, high margins, scalability and fragmentation drove most of the results, regardless of M&A style.
We spend a lot of lionizing Constellation and Mark Leonard and this is for sure deserved.
But Charles Wang built as enviable a record with a more centralized approach to software roll-ups.
Both monetized the same core insight - software is sticky, high margin, and fragmented.
playbooks.beehiiv.com/p/the-…
Many investors assert US SMBs offer more opportunity and inefficiency due to the sheer number of private companies - 36m - relative to the number of public US companies (~4,000).
But the scarcity in privates is in earnings power. With the median private company garnering a median EBITDA of ~$111k vs $20m in median earnings for a public company, the abundance in companies in private markets is offset by the scarcity of earnings power.
And ultimately investors are not buying companies, but earnings power.
Doesn't mean the private markets aren't inefficient or target rich. But there's a reason they're called 'small' businesses and expectations should be consistent.
playbooks.beehiiv.com/p/36-m…
🤖 Made with AI
Investor to SMB owner is possible.
Prior finance experience might even be uniquely helpful.
Helpful experience from an April podcast I've summarized:
playbooks.beehiiv.com/p/priv…
🤖 Made with AI
We spend a lot of lionizing Constellation and Mark Leonard and this is for sure deserved.
But Charles Wang built as enviable a record with a more centralized approach to software roll-ups.
Both monetized the same core insight - software is sticky, high margin, and fragmented.
playbooks.beehiiv.com/p/the-…
The original software roll-up OG is likely someone you've never heard of.
"Roll-ups" are all the rage, but no one brings up Charles Wang and what he built with Computer Associates starting in the 1970s.
It's arguably the best software outcome financially, perhaps better than even Mark Leonard's record at Constellation. Charles Wang pioneered software M&A before Constellation, Vista, and Thoma got their starts.
Highlights:
> Started in software sales
> Key observations: enterprise customers want reliability, software margins are high, software companies have good products, but not good sales teams
> Founded Computer Associates late 1970s, IPO'd early 1980s raising $3.2m
> Acquired companies to get the product; consolidated back office functions, eliminated executive overlap, and better aligned sales
> 200+ acquisitions later, became top 3 US software company by revenue and built a $25bn market cap company
>$3.2m of IPO proceeds to $25bn represents 7,800x
Wang built on centralized consolidation vs the more decentralized Constellation model, but both achieved incredible financial outcomes by taking a few simple concepts and implemented relentlessly.
I suspect the lessons of Computer Associates can inform today's entrepreneurs attempting roll-ups or 'build-ups', as Henry Kravis calls them.
Full breakdown:
playbooks.beehiiv.com/p/the-…
One big takeaway is how much the return profile of search or ETA jumps simply by closing a deal.
From 2.1x MOIC for all funded search to 3.7x MOIC as the expected value of all closed searches.
It’s obvious but an important reminder: two things are hard - finding a business and then operating it to create value.
But if you just move from searching to finding, you land in the pool of outcomes that’s empirically nearly 2x better than simply looking.
Chart below from CGPT:
Yes and perhaps happening at a time when the trend in ongoing token cost reduction is flattening.
Can interpret as 'new cases will be needed to support pricing as costs continue to move down' or 'new cases will be created because costs keep coming down'.
Flattening of the cost curve can help companies and investors discover where 'structural margins' lie at each layer of the stack.
(yes, the chart was built by CGPT but citations included; costs per token flattening in $s and cents, although % reductions are still steep from ~now through '29)
