SOPs - The KPE Way

SOPs - The KPE Way

 - Lessons Learned, episode 61

In Episode 60 of this series I mentioned that we had “our way” of doing things and we impressed this on all of our technicians. My company was called KPEnterprises, so we called it The KPE Way. It is inspired in part by HP’s famous The HP Way, which made its way into a book by David Packard, but is literally a set of guidelines on how HP operates and intends to show up in the world. (For more on The HP Way and how it showed up in my business from the start, see Episode 4 in this series: https://blog.smallbizthoughts.com/2024/11/lessons-learned-hp-way.html.)

I have long believed that one of the luckiest things that ever happened to me in my business is that someone recommended to me the book The E-Myth Revisited by Michael Gerber. I have since recommended that book thousands of times in the last thirty years. Gerber's philosophy is to have standardized, documented, enforced processes and procedures for everything.

Lucky for me, that fit perfectly with my ENTJ Myers-Briggs personality. (Google it)

I believe that there’s a right way and a wrong way to do things, and I believe we have an obligation to do things the right way (See Episode 10). Combining the concept of “Our Way” of doing things and a commitment to standard operating procedures, we came up with The KPE Way.

In some sense, the KPE way is a set of all the rules defined in this series.

• This is how we label drives in a storage array.

• This is how we say hello when we enter the client’s office.

• This is how we enter notes into tickets.

• This is how we migrate data.

• And so forth.

But beyond rules and regulations, The KPE Way includes how we treat each other, as well as clients and suppliers. It includes how we discuss things, how we document decisions, and how we build our common culture. It is literally everything we do. Why? Because nothing in your business exists in isolation. Even when we say, for example, that knowledge is siloed in different departments, that really means things are messed up somewhere in the system because knowledge isn’t share appropriately.

How is The KPE Way (and Your Company Way) implemented, evangelized to employees, and how does it become embedding in company culture? It starts with a good old “Mission-Vision-Values” statement. Why does our company exist and who do we serve? And, finally, it lists a few values that show up in our company code of ethics. “Few” means about five and no more than ten.

Those things get posted on the wall, on bulletin boards, and in the employee handbook. They also get posted on our web site. We talk about them all the time. And most importantly, everyone in the company works to make sure our actions and behavior are consistent with our stated values. 

Note that this is a lot more than a client-facing PR document. The Code of Ethics is intended to be a reflection of how our company chooses to operate. It is both internal and client-facing. For example, it says we will be competent in any services we sell. But it also says we only work with people we like, and that we value work/life balance. So if a client is abusive to one of our technicians, they get one warning and then we fire them. Yes, really. Clients can count on the fact that we’re competent and professional, and employees can count on support from the company when a client is unprofessional.

Everything’s Connected

Normally, when we talk about SOPs, we’re talking about the nuts and bolts of how the company is run and how service is delivered. And when we talk about culture, we’re talking about human interactions and the shared habits of the group. In other words, there’s the formal documented side and the social human side.

In reality, these are totally interconnected. Your company has procedures for hiring (the formal side) and those procedures should include mechanisms to insure that a new employee is competent and would be a good fit for the team (the human side). When we don’t follow the process, we end up with an employee who’s a bad fit. We have procedures for invoicing and collections (the formal side) and those procedures should include guidelines for treating clients professionally and with respect (the human side).

Once again, every single thing in your company is connected directly or indirectly with every other thing in the company. Therefore, it’s the totality of everything you stand for, everything you do, and all the people you interact with that make up The Your Company Way.

Once you adopt this perspective, you’ll find that writing procedures becomes a little easier, as does implementing them. And the more people share this vision of your company and hold each other accountable, the closer your actual culture is to the ideals in your mission-vision-values documents.

How to start? Absolutely anywhere. Adopt a systems view of your company and the holistic perspective will show up in everything you do, if you let it. And don’t forget to document what you can – cuz that’s The KPE Way!

Feedback always welcome.

Note: If you’re looking for a good Code of Ethics you might adopt or adapt for your company, please see The National Society of IT Service Providers’ model COE at https://nsitsp.org/code-of-ethics/.

(*) There has been a minor update, but you can find the original HP Way document at https://www.hp.com/hpinfo/abouthp/histnfacts/publications/measure/pdf/1977_07.pdf. Some have argued that HP moved away from these guidelines after Carly Fiorina took over.

-- -- -- 

Episode 61

This Episode is part of the ongoing Lessons Learned series. For all the information, and an index of Lessons Learned episodes, go to the Lessons Learned Page. 

Leave comments and questions below. And join me next week, right here.

Subscribe to the blog so you don't miss a thing.

-- -- -- 

Karl W. Palachuk is an executive coach and author of several books, including Managed Services in a Month and Relax Focus Succeed. He has built, bought, and sold several businesses, including two successful managed service businesses in Sacramento, CA. He advocates a holistic view of business, viewing the company as a system. You can find him at karlpalachuk.com or on LinkedIn. No artificial intelligence apps were used in the writing of this post.

:-)

ACES Conference 2026: Built to Scale

If you work in the Apple-focused MSP and consultant space, there’s a specific challenge that doesn’t get talked about enough: growing past the “good technician” phase into an intentionally designed business.

That’s what ACES Conference 2026 is built around.

Theme: Built to Scale

When: May 2026

Audience: Apple-focused MSPs and consultants serving 5–50 employee clients

ACES is not a technical event. This isn’t about macOS tips, deployment tricks, or MDM features. It’s about running and growing a sustainable services business in the Apple ecosystem.

The agenda focuses on the levers that actually move the business:

• Pricing and packaging

• Operational scale

• Sales and positioning

• Security as a differentiator

• Tool stack decisions (RMM, PSA, compliance, etc.)

• Leadership and team growth

In other words, it’s about building a company that works — not just one that keeps you busy.

New in 2026: An Employee Track

One addition this year stands out.

ACES is introducing a dedicated employee track. Attendees can bring a key team member for a separate day of programming focused on:

• Operations and process maturity

• Client communication

• Recognizing revenue opportunities

• Handling ticket overload without burning out

That’s a meaningful shift. Too many events assume the owner learns everything and then “downloads” it to the team. This approach invests directly in the people who are actually running delivery.

Intentionally Small, Intentionally Practical

The room is capped at roughly 125 consultants.   At that size, you can have real operator conversations.  Just working business owners sharing what’s actually working in their environments.

If you’re in the Apple-focused MSP space and looking to scale intentionally instead of accidentally, this is aligned with that goal.

There’s a 10% discount available using code FriendOfDave:

https://www.eventbrite.com/e/1665307221729/?discount=FriendOfDave

More details on the event are at:

https://acesconf.com

If you attend, I’d be interested in what you take away — particularly around pricing discipline and operational scale. Those are the conversations that tend to matter long after the conference ends.

Building Technicians Rather than Finding Technicians

Building Technicians Rather than Finding Technicians

  - Lessons Learned, episode 60

 

Here’s an age-old question: Is it better to “build or buy” a technician? The reality is that almost all business owners do some of both. In talking with several dozen people in the Small Biz Thoughts Technology Community, I find the most common answers to “How did you find your best technicians?” to be:

• 
  
  Technical recruiters
• Job boards (e.g., Monster)
• College internships and placement offices
• Friends/family/acquaintances

Less common but also mentioned are:

• Referred by a current employee
• Someone who worked at one of our clients
• Someone who just walked in the door looking for a job

I put that list in order from Most to Least likely to have the most professional experience and certifications. I shouldn’t have to say it, but we all probably know that the correlation between experience/certification and quality employee is present but weak. The relationship between cost and experience/certification is stronger.

We started out with the friends and family option because it just sort of happened. When it came time to hire an administrative assistant, I put ads out and had very good luck collecting resumes. So, after that, we primarily hired by placing ads on job boards (and primarily Craigslist). On a couple of occasions we used a job recruiter. We got some good candidates but a lot of people who just weren’t a good fit. I suspect the recruiters needed to send those folks out on interviews in order to justify their existence.

I would say, overall, I had the most luck with paid internships promoted through local tech schools and community colleges, followed by Craigslist. And that brings us to the build vs. buy question.

Note: A large part of the following discussion is true for me because we were rigorous about having “our” processes and procedures. We had our way of doing things, so we were committed to training new employees on our processes no matter where they came from or how qualified they were.

As we settled into recruiting from tech schools and Craigslist, we also settled into training technicians to create great techs out of good techs. This meant that our hiring process focused primarily on finding people who had a good attitude, a good culture fit, and a focus on good service. You’ve probably heard it a thousand times, but you can train people on the technology, but it’s harder to train them to have a customer-focused attitude toward service.

Lessons Learned: First, if I were to start over again today, I would only rely on internships and Craigslist to find people who are motivated and have technical aptitude. We would “build our own” from the start. Only if there was an urgent need for a specific talent or high-level guru would I spend money on recruiters.

Second, I would develop a more formalized training process for our team. We really figured out and documented our hiring process. It is detailed and works great with our service-focused approach. Where we could improve is evaluating which skills are needed for which technicians and providing them with the training to learn what they need to learn.

 

How do you feel about buy vs. build?

Feedback always welcome.

-- -- --

 

Episode 60

This Episode is part of the ongoing Lessons Learned series. For all the information, and an index of Lessons Learned episodes, go to the Lessons Learned Page. 

Leave comments and questions below. And join me next week, right here.

Subscribe to the blog so you don't miss a thing.

 

:-)

 

We Need a Real Data Center

We Need a Real Data Center

 - Lessons Learned, episode 59

 

In 2006 I bought a friend’s company. Well, essentially, I acquired his company in exchange for him coming on my staff at a very nice salary for a couple years. His company included a few regular consulting clients and a huge number of web sites set up in a rack full of crappy old “servers” and “network devices” that ran mostly on Unix.

I acquired all that so I could get his one big client. They became our largest client at the time. We also got a couple of smaller clients that I was happy to keep. The plan for all the micro clients (1-3 users) and all the hosted web sites was to pass them off to other consultants. First, we needed to move those web sites to the cloud so I didn’t have to stay awake at night worrying that the Frankenstein rack of junk hardware would fail.

At the time, I had a mostly-full equipment rack that housed three servers for us, switches, and a couple of servers that housed products we were test-driving for vendors. It was nice and clean and organized and beautiful. We used blue network cables on this rack and it was known as the blue rack.

Then we had the other rack full of junk, housing hundreds of web sites. We used green network cables on that one so, as you guessed by now, it was called the green rack.

All of this was in a room we’d built up to be our “server” room. It was about ten feet by four feet, had a dedicated through-the-wall air conditioner, lots of 30-amp electrical outlets, and a bunch of UPS backups.

This was a decent little server room, but hardly a data center in any sense of the word. In 2006 we were committed to moving all servers and services to the cloud, both for our clients and ourselves. So this little server room would do nicely until we could move things to a real data center. We’d already started planning that move. And then . ..

The owner of the building next to ours assumed that, since his parking lot had been there for decades, it would be okay to dig a hole in it for something. The back hoe went down about two feet before it cut through a major electrical cable and blew up a transformer, forcing a power outage for blocks around that lasted more than a week! The transformer was one of those 4-foot cubes you see from time to time. I guess you don’t replace those things quickly.

Anyway, we were instantly on battery backup, but that wasn’t going to last days. I ran down to Home Depot and bought a generator, then built a cable beefy enough and long enough to go from the parking lot to our server room. I didn’t know whether the UPSs were rated for generator usage, but it didn’t really matter. I plugged them in and ordered new UPSs to be delivered as soon as possible.

And then we signed the deal to set up our racks in the real data center. Our plan hadn’t changed, really. We were just forced to speed up the execution. Unfortunately, that experience cost a good deal of money. Aside from the UPSs, the generator, and the cable, we had to pay someone to sleep in the office because we couldn’t close the doors with the big cable running through the place.

Lessons learned. Well, this is a tough one. Aside from learning that the building owner next door is a moron, the rest was just poor timing. We had a plan to move everything. We just didn’t get to do it on our schedule.

We did not have a week-long power outage in our emergency planning, but our experience dealing with disasters made this one pretty low stress under the circumstances.

Looking back, my final assessment is that we should have been done two weeks earlier.

But for context, most of the rest of the industry didn’t gets servers out of the office for at least five or ten years after us!

 

Feedback always welcome.

-- -- --

 

Episode 59

This Episode is part of the ongoing Lessons Learned series. For all the information, and an index of Lessons Learned episodes, go to the Lessons Learned Page. https://blog.smallbizthoughts.com/p/lessons-learned-blog-series.html

Leave comments and questions below. And join me next week, right here.

Subscribe to the blog so you don't miss a thing.

 

:-)

Business of Tech Plus Content Now in the Small Biz Thoughts Community

For those who haven't noticed yet, Business of Tech Plus content has started appearing in the Small Biz Thoughts community resource library.

How It Works

Kara set up an automation that pulls Plus content directly into a new section at the bottom of the resource library. Once it's posted on the Plus side, it shows up here automatically—no extra steps needed.

What's Included

Here's a rundown of what you'll find:

Full-Length Interviews

Instead of the shorter public clips, you get access to the complete conversations. The recent interview with Mike Riggs and Wes Spencer, for instance, runs about 22 minutes in full versus the six-minute public version.

Monthly Briefs

A summary at the start of each month covering the key news from the previous month and why it matters.

Vendor Strategy Briefs

In-depth looks at various vendors—helpful if you're doing due diligence on potential technology partners. There are currently 12 briefs available covering: Rewst, Liongard, SuperOps, Auvik, Huntress, Pax8, N-able, ConnectWise, Kaseya, Acronis, Slide, and NinjaOne.

Each brief goes through a thorough process: vendors are given the opportunity to fact-check the content, and the briefs are also independently fact-checked to ensure accuracy. New ones come out regularly.  

Leadership Content

Additional pieces covering topics like the evolving role of channel chiefs and other industry trends.

A Few More Coming

This is the first batch of content to make its way over. There are still a few more pieces that will be added soon.

If you have any questions about finding this content, just head to the resource library and scroll down to the Plus section.

“Agentic AI”: Hype vs. Reality for MSPs

 Let’s talk honestly about “agentic AI.” No doubt you’ve seen vendors pitching it as the next big thing for managed services and IT consulting. As always, the concept has a solid technical backbone—but the marketing narrative is another animal entirely.

That means it’s on us to decode the buzzwords, sort out what matters to our businesses, and steer clear of distractions.

Before We Go Further: Two Meanings of “Agentic AI”

Here’s a key distinction that’s getting missed:

- Agentic AI (marketing version): The promise of AI systems acting independently, making decisions, and running environments on autopilot—sometimes with less oversight than reality can support.

- Agent-based automation (actual tech): Software agents that operate semi-independently within carefully crafted guardrails, reliably handling tasks like patching, identity changes, and security responses. The architecture is proven; it just doesn’t run free of ownership and review.

My concern isn’t with agent architectures—they’re part of every MSP’s toolkit. It’s the vendor story of frictionless autonomy that leads us astray, downplaying how risk, responsibility, and oversight actually work in real service environments.

Why Accountability Rules—Now and Always

Here’s the non-negotiable: In managed services, someone’s always legally and contractually responsible for the result, whether a person or an automated agent made the change.

If a system triggers downtime, data loss, or a compliance blunder, the client won’t accept "the agent did it" as the answer. They'll pull up the contract, the SLA, and maybe even their insurance—because what matters isn’t philosophy, it’s economics.

This core accountability isn’t optional. It’s the backbone of our business model.

Automating With Agents—The Real MSP Example

Let’s anchor this with something MSPs live every day: patch management.

You set up a patching agent to scan, download, and deploy updates across client endpoints. The vendor’s glossy pitch might suggest it “handles updates for you.” But in practice, you:

- Set strict approval workflows (no patch goes live without your review)

- Track all actions in an audit log

- Roll back if something conflicts or fails

- Take ownership for outcomes in client-facing reporting

Automation speeds up the routine, but it doesn’t shift responsibility. You, not the agent, are named in the contract—and you’re who the client will call if something breaks.

That’s the pattern again and again, whether with identity changes, automated ticket responses, or security rules. Agents are powerful. They’re not autonomous in any risk-free sense.

Where Vendor Narratives Create Confusion

The hype version of “agentic AI” imagines fleets of self-directing systems making high-stakes choices without human check-ins. But in real managed services:

- Workflows require business context and review

- Systems need clear, documented guardrails

- Compliance is audit-driven

- Risk management is built on accountability

What we get from technical advances (LLMs, orchestration, agent architectures) is better task decomposition, smarter interfaces, and robust automation—but we don’t get out of the loop.

In short: The marketing label is ahead of the architectural reality.

On “Citizen Development”—And Why Governance Still Matters

You’ll also hear the claim that natural language AI will make everyone a kind of software developer (the so-called “citizen developer” revolution). In MSP terms, this sounds like clients being able to “write” their own automation inside their business systems with just a few prompts.

That’s interesting—but here’s what’s missing: It’s not just skill levels that matter, it’s governance and risk. Whether your technician, admin, or client is using an agent-powered tool, someone has to:

- Establish rules and boundaries for changes

- Validate outputs before they go live

- Track accountability for actions taken

AI can simplify building processes—but it can’t remove the need for structured thinking, domain knowledge, and clarity about who’s responsible when things go wrong.

If You’re an MSP: What Matters Most Right Now

Where should you focus?

1. Leverage agent-style automation for repeatable, auditable tasks. Patch approvals, overnight reporting, security hygiene—these are real wins.

2. Build governance into every deployment. Keep approval flows, audit trails, and risk frameworks in place from the start.

3. Don’t sell autonomy—sell reliability and accountability. Your clients are trusting you, not your bots, to deliver business outcomes.

4. Educate clients about their process, not just their tech. Help them understand what makes sense to automate.

5. Stay in the loop. Oversight, review, and judgment are why you exist.

Automation is the engine, but you’re always the driver.

Summing Up: Agentic AI Is Useful—But Frictionless Autonomy Is a Myth

Here’s the upshot:

Agent architectures and automation are here to stay, and they’re great tools for any MSP. But the hype around “autonomous AI agents” doesn’t align with the reality of economic liability, SLAs, or how clients expect risk to be managed.

The opportunities are real. The responsibility is, too.

When something goes wrong, you won’t point to the agent—you’ll answer for the fix. That’s how the managed services model works, and it’s not going anywhere.

We Can't Afford Every Client

 We Can't Afford Every Client - Lessons Learned, episode 58

 Once I started turning almost all work over to technicians, I quickly saw an “established” bad practice that needed to change. Here’s how I noticed it and  how I addressed it.

When I started my business, I made the same error that almost all new consultants make: I did not place enough value on my time. I don’t mean the amount of money I charged so much as how I chose to use my precious hours. And perhaps back then I would not have called them precious. Now I treat them like rare gems.

I made a profit on all jobs (with rare exceptions), but some jobs should not have been done. Specifically, there were jobs that seemed profitable because my time didn’t “cost” me anything. Even then, I knew that I probably wouldn’t be able to do these jobs if I had to pay a technician.

A great example is (unfortunately for me) one of my favorite clients. They ran a feed store a few  towns away, up the foothills toward Tahoe. At first they were undeniably profitable. They had a big problem that was driving them crazy. That got my foot in the door. While there, my network assessment exposed a server that was choking and a backup that hadn’t worked in more than two years. On top of that, their point of sale systems were in the feed barn. In other words, a barn with cash registers. This is just about the dustiest environment you could imagine.

All of which is to say, we started doing regularly monthly maintenance. At first, we a lot of fix-up and clean-up work. Then it settled down to about 30 minutes per month to check server, network, and basic stuff, and 30 minutes to clean up the POS machines. [Literally. We opened them up and vacuumed them out because there was so much dust.] Total onsite = one hour.

Again, when I was doing this, I’d head out late morning so I could take care of the client and then eat lunch in this cute little town in the foothills. It took about 45 minutes to drive each way.

Then I started sending technicians. I don’t remember the exact fully burdened rate, but let’s call it $60/hour. Drive time is 1.5 hours and service labor is 1.0 hours. That’s 2.5 x $60 = $150. We didn’t charge for drive time (one of the few things I would change if I went back to the beginning.)

We charged the client for on hour onsite at $150/hr. Zero profit. But if that tech was driving around Sacramento, we might have had 30 minutes drive time in those 2.5 hours and billed two hours of labor, putting us ahead by $150.

We got into this situation because I used to make the mistake of believing that my time cost nothing. There’s the cost out of pocket and the cost of NOT doing another job that pays better.

Serving Out-of-Towners

That client was an outlier (literally) on the mountain side of the valley. In the other direction is the Bay Area. We had clients down in the Delta (where the Sacramento empties into the Delta on the way to the Pacific). We also had clients closer to Oakland. The Delta was closer but on tiny winding roads. The Bay Area clients were about eighty miles, which might take 90 minutes or might take four hours to drive.

Eventually, we drew a line to represent about 30 minutes drive time. We did not charge for travel inside that area. For outlying areas, we set a four hour minimum. That way we were pretty much guaranteed to be profitable.

We gently passed off the feed store to another consultant in our SMB IT user group. The Bay Area folks didn’t bat an eye. At the time, this was a $600 minimum, so I was surprised. But we didn’t hear a peep or complaint out of them. To be honest, I was quite shocked.

From time to time we culled clients because they were too small and didn’t meet our minimums. And once we set the four hour minimum for out of towners, we never had a client who was unprofitable again. Really.

Lesson Learned: The owner’s time is valuable. Especially if the owner’s primary responsibilities are sales and client relationships, that’s the most valuable time in your company. When calculating jobs, assume you’re a well-paid technician and make sure the job still looks profitable. You can’t give away your time.

 Feedback always welcome.

-- -- --

 Episode 58

This Episode is part of the ongoing Lessons Learned series. For all the information, and an index of Lessons Learned episodes, go to the Lessons Learned Page. 

Leave comments and questions below. And join me next week, right here.

Subscribe to the blog so you don't miss a thing.

 :-)

Join Today:

 

The ASCII Group Appoints Ted Roller as Chief Community Officer

 Washington, D.C. – January 14, 2026 – The ASCII Group, North America’s original and vendor-neutral community for Managed Service Providers (MSPs), is pleased to announce the appointment of Ted Roller as Chief Community Officer, a newly created executive role focused on strengthening member engagement and expanding ASCII’s community across the United States and Canada.

In this role, Roller will lead initiatives designed to deepen member connection and engagement, working closely with MSPs to help shape the future of the ASCII community. His focus will be on ensuring ASCII continues to reflect the real-world needs of MSP business owners and remains a trusted resource for peer insight and collaboration.

Roller brings more than 25 years of experience in the IT channel, with a career spanning MSP ownership and senior channel leadership roles. Earlier in his career, he owned and operated an MSP that was recognized locally as one of the Fast 50 fastest-growing companies, providing firsthand experience with the operational and strategic realities MSPs face.

He has since held senior channel leadership positions at LogMeIn, Intronis, Mailprotector, CloudRadial, Zomentum, and FlexPoint, where he helped build partner programs, strengthen community engagement, and support growth through indirect sales models. In addition, Roller has contributed to industry education as faculty for CompTIA and has been recognized by organizations including MSPmentor, The Channel Company, and SMB Nation.

“Ted’s background as both an MSP owner and a channel executive makes him uniquely qualified to lead our community efforts,” said Jerry Koutavas, CEO of The ASCII Group. “His role will be instrumental in strengthening member engagement and expanding ASCII’s reach as we continue to support MSPs across North America.”

“ASCII has always represented something different in this industry,” said Roller. “It is a community built on trust, openness, and shared experience. I’m honored to join the team and look forward to helping shape the next chapter of member engagement and community leadership.”

About The ASCII Group, Inc.

The ASCII Group is the premier community of North American MSPs, MSSPs and Solution Providers. The Group has members located throughout the U.S. and Canada, and membership encompasses everyone from credentialed MSPs serving the SMB community to multi-location solution providers with a national and international reach. Founded in 1984, ASCII provides services to members including leveraged purchasing programs, education and training, marketing assistance, extensive peer interaction and more. ASCII works with a vibrant ecosystem of leading and major technology vendors that complement the ASCII community and support the mission of helping MSPs to grow their businesses. For more information, please visit www.ascii.com.

Pia Unifies Community and Strategic Alliances Under New Chief Community Officer Role to Accelerate Ecosystem-Led Growth

 TAMPA, Fla. — January 13, 2026 — Pia, the leading AI-led help desk automation platform for managed service providers (MSPs), today announced that longtime executive James Allen has been appointed to the newly created role of chief community officer, unifying community strategy and strategic alliances as the company scales its partner ecosystem. He reports directly to David Schwartz, CEO of Pia.

The chief community officer position represents a fundamental shift in how Pia is approaching growth. Allen will orchestrate community and partnerships as interconnected drivers of the company's expansion. He'll build scalable programs such as advisory boards, user groups, and certifications while managing technology partnerships, distributor relationships, and co-selling initiatives with vendors.

“The creation of this role signals our belief that community and partnerships aren't just support functions; they're core growth drivers,” Schwartz said. "James understands that in the MSP market,success comes from building an ecosystem where partners co-innovate, customers connect, and strategic vendors integrate into our value proposition. By unifying these efforts under one leader, we're building a scalable growth engine that contributes directly to pipeline, revenue, and brand equity."

Allen will ensure the “voice of the community” shapes product, marketing, sales, and customer success teams — keeping Pia's innovation roadmap aligned with MSP needs.  

"I'm excited to take on this role and strengthen Pia's focus on community-led growth," Allen said. "Our community is how we stay close to what MSPs need and how we build the right AI-driven solutions to help them serve their customers. Bringing community development and our alliances together will make it easier for our partners to stay connected to their customers and get the support they need to grow."

Allen brings 15 years of executive experience across finance and MSP ecosystems. Since joining Pia in 2022, he's served as executive vice president of strategic partnerships and senior vice president of sales, cultivating relationships with CEOs, CFOs, and senior MSP leaders while building connections with technology partners, distributors, and vendors. His career has focused on helping MSPs enhance efficiency, improve profitability, and deliver exceptional customer experiences.

About Pia

Pia specializes in transforming the help desk experience for managed services providers (MSPs) with its AI-powered automation platform, Pia aiDesk. The platform leverages advanced technologies such as artificial intelligence, machine learning, and natural language processing to automate and streamline the most common help desk tickets, significantly increasing efficiency and reducing costs. To learn more about how MSPs can experience the future of help desk management with Pia aiDesk – where AI meets operational excellence, delivering consistency, scalability, and customer satisfaction across every interaction – visit https://pia.ai/.

The AI Gold Rush Will End Like Airlines: Great for Users, Terrible for Operators

I've been thinking a lot about airlines lately. Not because I'm traveling more, but because every time I use ChatGPT or Claude or any of the AI tools I've woven into my daily workflow, I see the same pattern emerging that turned aviation from a glamorous frontier into a brutally commoditized business.

And if I'm right, we're about to watch the AI industry follow the exact same trajectory.

The Uncomfortable Truth About Infrastructure

Here's what I keep coming back to: the best infrastructure businesses are terrible businesses to be in.

Airlines are the perfect example. They require massive capital expenditures. They operate on razor-thin margins. They're completely commoditized—most travelers pick based on price and schedule, not brand loyalty. Running an airline, frankly, sucks.

But the unlock airlines provide? That's massive. Global business travel. Tourism economies. Supply chains. Family connections. The entire modern world depends on cheap, reliable air travel existing.

The same pattern played out with shipping containers. Before containerization, global shipping was boutique, expensive, and labor-intensive. Then we standardized it all. Now it's a low-margin, commoditized business that most people never think about.

But that commoditization unlocked something extraordinary: modern global trade. Your ability to order something manufactured in Shenzhen and have it on your doorstep in Ohio a week later depends entirely on shipping being cheap and boring.

I think AI is heading down this exact path.

Why AI Will Become Boring Infrastructure

Look at what's already happening:

The models are converging fast. Six months ago, there were meaningful differences between GPT-4, Claude, and Gemini. Now? They're all basically the same for most use cases. The gap is closing weekly.

Open source is catching up. We're one or two breakthroughs away from having open-source models that match the commercial offerings for 90% of use cases. When that happens, the pricing pressure becomes brutal.

The capital requirements are insane. These companies are burning billions on compute, data centers, and talent. The circular spending between big tech companies—where Microsoft invests in OpenAI, which then spends that money on Azure compute—can't last forever.

It's starting to feel like the dot-com bubble. I lived through that one. I've got enough gray hairs to recognize the pattern. Massive investment. Sky-high valuations. Everyone convinced they're going to be the winner. Then reality hits.

And here's the kicker: AI spending is currently propping up the US economy. When this bubble deflates—and I think it will, probably not as catastrophically as dot-com but it will deflate—it's going to be painful.

The Good News (For Everyone Except AI Companies)

But here's what makes this bearable: the infrastructure will still be there.

After the dot-com crash, we didn't lose the internet. We lost a bunch of overvalued companies, sure. But the fiber optic cables stayed in the ground. The protocols kept working. The infrastructure remained, and ten years later we built entirely new businesses on top of it.

The same thing will happen with AI. The models will become commoditized utilities. Running an AI company will be a low-margin infrastructure play. But the unlock for the rest of us will be extraordinary.

I'm already seeing it in my own work. I use AI extensively—Notion AI for research and organization, ChatGPT for drafting and ideation, various other tools for production workflow. It's genuinely the most interesting technology I've played with in years.

My goal is that people see it in the velocity of my output, not in the writing itself. It's just me, but more. And that multiplier effect is real.

What This Means for MSPs and IT Services

Here's where it gets relevant for the channel:

In the short term, there's going to be turbulence. When AI valuations correct, it will ripple through the economy. Small businesses will feel it. IT services spending will take a hit. The timeline is unclear—things move faster now than in 2001, but "faster" doesn't mean "four months."

In the medium term, AI becomes a capability, not a product. You won't sell AI services. You'll use AI to deliver everything else more efficiently. Just like you don't sell "internet" today—you assume it exists and build services on top of it.

In the long term, the intelligence unlock is transformative. The ability to automate cognitive tasks that used to require human judgment? That's massive. But it also means the commodity technical work that many MSPs still rely on becomes even less defensible.

This is why I keep hammering on the same theme: MSPs need to move up the value chain now. When AI makes password resets and basic troubleshooting trivially easy, you need to already be positioned as a strategic advisor, not a help desk.

The technology is enabling that transition. But there's a window, and it's closing.

The Airline Lesson

So yes, I'm bullish on AI. I use it constantly. I think it will unlock enormous value across the economy.

But I don't want to run an AI company any more than I want to run an airline.

The best infrastructure is the kind you never think about. It just works, it's cheap, and it enables you to do more interesting things.

AI is going to become that. It's going to be great.

Just don't be surprised when the companies building it discover they've created a commodity business with airline economics.

What do you think? Are we heading for an AI winter, or is this time different? I'm curious where you see this playing out in your business.