The power of specialization

What clients still cannot get from a chatbot is someone who understands their business. Anyone can close a ticket. Reading the business behind the request takes knowing its industry well, and a generalist juggling ten industries at once knows none of them well enough. A specialist does.

That is why picking one industry and going deep is one of the smartest moves an MSP can make right now. The generalist is caught between other MSPs and the client's own AI. The specialist is not.

Danny Tobon has been running that experiment since well before AI made the case for it. His firm, Tsquared Technologies, is fully remote and serves accounting and CPA firms and nobody else. He committed to that single vertical years ago, sat through all the doubt that comes with narrowing a young company's market, and can now show you exactly what the bet bought him. The one thing he would change about the whole journey is that he did not narrow sooner.

A startup that turned away paying customers

Narrowing your market is a strange thing to do when you are small and hungry for revenue. Most young MSPs say yes to almost anything that pays, because scarcity makes fit feel like a luxury they cannot afford. Danny ran the opposite play on purpose. Getting Tsquared down to accounting firms and nobody else meant deliberately shrinking the list of people allowed to hire him.

"We turned away prospects who wanted to pay us, which is a strange thing to do as a startup," he says. Over time the firm also let go of the non-accounting clients it already had, wherever the economics no longer lined up. To an outsider that reads like a business working against its own interest. To Danny it was the smartest bet he ever placed, and the story of why sits in how he got into technology in the first place.

The vertical found him before he chose it

Danny's family came to the United States in 2001, when he was around ten. A couple of years later they bought their first computer, a Gateway from Best Buy, and he broke it almost every week, deleting drivers and poking at things he did not understand until something stopped working. That, he says, is genuinely where it started, because he learned he was not afraid to break something and then work out what he had done. Around the same time a cousin in IT helped him build his own machine, and the thing he loved doing came into focus. He became the person who fixed things for family and friends, back when everyone had viruses.

Then a family friend opened an accounting practice out of her home, preparing taxes. This was the end of high school and the start of college. She was running Drake on a single computer and needed it centralized so more than one person could work, and Danny set it up. It was his first real look at how a firm actually operates. She later introduced him to a trucking company that became his biggest side client, a 24/7 operation where somebody had to be reachable if dispatch went down. He never sat down as a kid and decided he would run a business, but the mindset showed up early. He made jobs out of nowhere, knew he did not want to work for someone else his whole life, and set up the business in 2014 while still holding a full-time job.

The detail that matters is the one that repeats. His first real client, years before any of this became a company, was an accounting firm. The vertical he would eventually stake everything on had already walked through the door once, quietly, before he knew to notice.

The night that forced the decision

Danny started his career at a large MSP in Maryland, straight out of Stevenson University, a firm of more than a hundred people that was vertical in the automotive industry. He began at the bottom, on the support desk. For years he was inside the MSP world without really understanding what an MSP was as a business, because he only ever saw one slice of a large one. A move to a smaller shop, roughly two million dollars in revenue, put him in a network administrator seat and taught him how the business actually runs. Tsquared existed the whole time on the side, two clients and about fifteen hundred dollars a month.

In November 2018 the smaller firm let him go. His daughter had been born that July. They were not happy that he had clients on the side, and there was a non-compete. It happened fast, boxes packed, done. Telling his wife with a newborn at home was not a good night, and he asked her for three months to figure it out.

So he went all in. He found Nate at Tech Pro Marketing on LinkedIn, could not afford a retainer, and asked instead to be taught the process one time for a one-time fee, promising to come back once there was real revenue. Nate showed him the ropes and Danny ran that motion for about a month. The first inquiry that came in, in February 2019, was a client accounting services and bookkeeping firm. He was not marketing to accounting at all. His very first part-time client years earlier had been an accounting firm, and now his first inquiry as a full-time owner was one too.

Choosing accounting, and why the reasoning holds

Even then he had not committed. That came when he started working with an MSP coach, Joe Rojas at Stargrow Manage, who pushed him toward picking a market. They agreed to try it, and once they did, everything clicked. Clients want somebody who already knows their industry and does not need the basics of their own work explained back to them.

Why accounting specifically comes down to two things, and the first is a resemblance most people miss. An accounting firm's work looks a lot like an MSP's. Bookkeeping is their version of proactive monitoring, staying on top of transactions the way an MSP stays on top of patches and inventory and anything new entering the environment, keeping things clean so nothing compounds. When Danny understood that, the fit made sense.

The second is need. Every accounting firm he looked at had gaps and things done the wrong way. These firms inherit enormous responsibility for their clients' most sensitive information, especially the ones doing tax and compliance work, and in a way it is not fair to them. They are trying to serve their clients, and the government hands them a security burden they never signed up for. They need someone to guide them through a world they cannot be expected to know. The parallel runs further than that. A CPA with the knowledge can start a firm from the house tomorrow, the same way an engineer can start an MSP, and then set up their own systems with no idea what a scalable foundation looks like. It is exactly what happens if a technologist sets up their own books with no finance background. Eventually someone has to come clean it up. Danny loves being the person who does that on the technology side.

When other owners ask him how to specialize, his answer starts with the clients they already have. Find the one you are genuinely excited to work with, where the interest runs past the technology into what the business actually does. Most MSP owners already like technology, so the thing to look for is the second passion underneath it. Then ask whether you can see yourself in that industry for years, and whether that client's culture is the culture you want more of. The mistake he warns against is chasing money into a vertical whose rhythm you will come to resent. If an industry runs around the clock and you cannot picture building or staffing for that, no revenue number is worth the burnout that follows.

The cost was the point

Nobody really tried to talk him out of it. Joe was encouraging the whole way. The resistance was internal, and the biggest enemy was Danny himself. Even after deciding, he kept questioning it. The firm was not where he wanted it to be on revenue, and here he was deliberately limiting who could hire him. That doubt still surfaces occasionally.

What he gave up was real. Turning away prospects who wanted to pay, and later letting go of non-accounting clients where the economics were not aligned, is a hard thing to do when you are still building. The surprise was that it made those decisions easier rather than harder. As a generalist, especially early, you say yes to almost everything out of scarcity even when the fit is bad, because you need the revenue. The niche gave him a filter, and the criteria became obvious.

That honesty runs through how he talks about the strategy now. The cost of concentration never fully goes away. All the eggs sit in one basket, and the loudest question in accounting today is whether AI replaces the work, since categorization that used to require an accountant is now largely automated. If that industry takes a hit, Tsquared feels it, and Danny says plainly that still scares him. The gift, though, is the opposite of what he expected. Narrowing did the marketing for him.

When it stopped being a marketing angle

The first thing that changed after committing was Danny himself. He still runs the sales and marketing side of the business, so he had to learn to speak the language of accounting firms and be where they gather. Then the clients started doing the work for him. Accounting firm owners are a family. They know each other, attend the same conferences, and sit in the same communities, and referrals began moving through the clients he already had. "One accounting firm tells another that this company actually knows our industry," he says. "No generalist ever earns that sentence." It also compresses the sales cycle, because by the time a firm reaches him the expertise is assumed and the conversation starts at what are you trying to accomplish rather than proving he is good at technology.

It stopped being a marketing angle the moment operations started looking different from a generalist's. Tsquared works with remote-first firms that scale headcount hard around tax season, and capacity for busy season is one of the biggest issues in the industry, which means onboarding and offboarding happen constantly. So the firm built provisioning into a real product and leaned in further than a technology company normally would. It does not act as HR, but it goes well past where most providers stop. A generalist provisions the account, ships the laptop, and moves on. Tsquared verifies device check-in and runs a technology onboarding call with the new employee, and that extra step is what makes the firm feel like an extended part of the team instead of a help desk. It could only build that because every client shares the same shape.

The focusing did not stop at accounting either. Once inside an industry, Danny found verticals within it, and narrowed again to cloud-native, tech-forward, remote-first firms, owners already sharp on technology who need an expert layer rather than hand-holding, often with no headquarters at all and staff spread nationwide or offshore. Specialization, in his telling, is not one decision. It keeps focusing.

That depth is also what let him leave the commodity fight. A generalist is stuck arguing that they cost less or do the same thing or have more features. Once Tsquared became the accounting firm experts, prospects stopped comparing it to the MSP down the street. Danny is candid about the internal side of this. Clients from before he went vertical are grandfathered into old pricing, and he looks at those numbers now and wonders what he was thinking. He keeps them there anyway, not out of fear they will leave but because the relationships are good and they send referrals, and it feels like the right thing to do. When an accounting firm comes to him today, the price is meaningfully higher, because he knows what the work is worth. The gap between those old numbers and today's is the value of specialization, measured in dollars.

From provider to advisor

As soon as Tsquared went vertical, it was clear the firm should not operate like a break-fix shop or a generic IT department that shows up for computer setups. Danny held that view well before AI was the topic of every conversation. A firm was never coming to him just to get one issue fixed. They wanted guidance through their technology journey, because technology is now one of the biggest things in their business. The technical answer was never the scarce part. Google existed long before ChatGPT and Claude. What a client cannot get anywhere else is someone asking why they are really asking for a thing, going one level deeper than the request.

One example carries the point. A firm's real problem was employees leaving to start their own practices and taking information with them. They had the roadmap, and it walked out the door, which is insider threat in everything but name. A generalist hears that and locks down file access, and the ticket is closed. But it was never only a technology problem. It lived across process and procedures, hiring and vetting, training, and finally enforcement, and enforcement is the only piece that is technology. Giving the owner a real roadmap for closing those gaps takes a CTO and CSO perspective, not a ticket. That is the difference between fixing what a client asks for and diagnosing what they actually need, and it is the part that never gets replaced, because it is the relationship.

Going vertical sharpened everything behind that advisory role. It told Danny exactly what talent profile to hire for, so the team fits the industry and the client culture, and it let him standardize the tool stack instead of maintaining a sprawl. Documentation is where it compounds most. Because the firm knows what an accounting firm's foundational technology looks like and what sits on top of it, documentation, training, onboarding, and computer setup all run off templates. A generalist taking on a law firm next would reset all of that to zero. For Tsquared the playbooks stack up. Write it once, refine it forever, and bring a new engineer into one world instead of ten.

One hiring lesson he learned runs against the usual advice for a young firm. Early on he brought in junior, very green people, and he would build differently now. That approach fits a mature MSP that already has layers and a real org chart, not a small shop still finding its feet. For a small firm he would hire experienced, advisory-minded people first and add the junior layer underneath them later. The reason goes past the quality of the work. Bringing in people who are smarter than you in their own domain, real A players, is what finally lets a founder step away. The founder who knows the most technically cannot delegate his way out while everyone below him still needs him, and a business stuck in that shape never learns to run on its own.

The AI reckoning

If specialization is the bet, AI is the stress test, and mostly it has proven the advisory case rather than undercut it. What AI changed first, for Danny, was client expectations. Every client now has an expert at their fingertips and lives in a world that expects an immediate answer, which puts real pressure on the provider. If Tsquared is not fast enough, the client finds the answer somewhere else. That means the old model has to change. Both MSPs Danny worked at were help-desk shops, essentially call centers where people phoned in to get things fixed, and that is how he built at first because it was what he knew. He thinks it should be chat-based now. Ask in the chat, get an immediate answer, avoid opening a ticket at all, and escalate to a higher-level engineer only when the problem needs real judgment. AI becomes the first line, and humans handle what actually requires a person.

Where AI falls short, and where Danny thinks the industry overestimates it, is context. A client once brought him a technology upgrade plan written entirely by AI, impressive to read but blind to the client's actual environment. That is exactly how a confident client starts implementing and ends up with downtime and a plan that has to be reworked. The confidence outruns the context, and that gap is why the advisory role holds, because somebody has to hold the full picture.

The worry that stays with him is not the one people ask about. If AI resolves all the entry-level issues, how do junior support people learn? They need experience to become good engineers, and the foundational work is exactly what is being automated. Accounting is asking the same question about junior accountants. The rungs people used to climb are being automated away.

The hardest part of the work has little to do with technology. Deploying controls is an afternoon's job. Getting a firm to take governance seriously and genuinely change how its people work is where the real difficulty sits, and it is why Danny treats trust as the center of what he does. His clients are trusting him to guide them through a responsibility they are often confused about carrying, and making sure a firm is doing the right thing is a heavy obligation to hold.

None of that makes Danny the provider who tells firms to slow down. He refuses that posture. Telling a firm not to use AI is neither achievable nor right, because it just guarantees they fall behind while a competitor takes data privacy less seriously and pulls ahead, and the technology is genuinely remarkable for how much time it saves. His clients are getting deep into these tools, many of them into Claude, and he is walking that journey with them in real time without pretending he knows every in and out, because it is new and the unknowns around trust, privacy, and data handling are real. So Tsquared enables AI with eyes open. It makes firms aware of the risks and the gaps, including the fact that policy and regulation have not caught up, and it corrects one important misconception, that the risk ultimately sits with the firm and not with the IT team. Then it helps them train the team, put guardrails around what is worth avoiding, and make the decision themselves. Nobody has this figured out yet, Danny included, and everyone is navigating it as it happens.

The road ahead

Look far enough forward and Danny thinks the question stops being how an MSP works and becomes what an MSP even is. The industry has kept climbing, from break-fix to managed services to managed security, and the next rung is the specialized, advisory-focused firm that walks a client through their industry's journey and makes the firm itself smarter, including on AI, while the routine work keeps automating. Some of today's model does not survive that shift. The break-fix layer goes, and so does the junior support desk role, because more efficient ways of working are already here. People live in an Amazon Prime world now, ordering from a phone and expecting delivery today, and the same expectation applies to answers.

His bet on where the plumbing goes is specific. RMM, PSA, and the knowledge base collapse into the tools people already use. Email is cumbersome, a message that waits for a case to open and a callback and a round of back-and-forth, when a client could just ask in Slack or Teams the way they message a coworker. The system answers straight from the knowledge base if it is a question, or triggers provisioning on the back end if someone is starting Monday, and most of it runs automatically. MSPs should be investing in that experience now, because a provider who is not using AI to its potential becomes the friction in the relationship, and the client finds one who is not.

He wants two things from the next decade. More MSPs going vertical, and compliance treated as a core offering rather than a side project. Tsquared went deep on the second, mapping FTC Safeguards Rule requirements under GLBA, though Danny is honest that compliance is a hard sell to a small business while it still feels optional. It changes when it becomes something a firm must do rather than should, the way car insurance is not a choice, and if enforcement catches up that way the MSP becomes the trusted guide who walks clients through it, with a safer cyberspace as the byproduct.

The next frontier, Danny thinks, will run the same course security did. Security was a specialty once, and now most MSPs bundle it in without thinking. Helping firms put AI agents to work will make the same trip, from a thing that sets a provider apart to a thing every provider is expected to do. He intends to be early to that too.

None of it changes the question underneath the work, the one the whole firm is built to ask. A client shows up with a request, and the job is to figure out what they actually need. AI has not touched that question. If anything it has sharpened it, by clearing away the easy answers and leaving only the part that takes someone who knows the business. Danny bet years ago that this was the real work, and narrowed his firm until he could do nothing else. The bet is still paying, and it pays a little more every time a client hands their AI a problem it cannot see the whole of.