August 2026

In 1975, a 17-year-old named Peter Cancro bought his first sub shop in Point Pleasant, New Jersey. On Wednesday, that shop’s parent company went public at a $7.3 billion valuation. Jersey Mike’s and its backers raised $1 billion in an IPO priced at $23 a share, more than 10 times oversubscribed, Bloomberg reports.

The chain now has more than 3,300 US and Canadian locations, plus an expansion deal with Cancro to open roughly 300 more across the UK and Ireland. Revenue grew 11% to $724 million last year, and Jersey Mike’s US sales growth has outpaced rivals like Jimmy John’s, while Subway has actually declined, as diners lean toward sandwiches over pricier bowl concepts.

Blackstone, which bought Jersey Mike’s last year for about $8 billion including debt, will retain 68% of voting power after the IPO. Cancro stepped back from CEO to chairman in April, handing the role to Charlie Morrison, who previously ran Wingstop and Salad and Go. Shares began trading Thursday on the NYSE under the ticker JMKE.

In 1975, a 17-year-old named Peter Cancro bought his first sub shop in Point Pleasant, New Jersey. On Wednesday, that shop’s parent company went public at a $7.3 billion valuation. Jersey Mike’s and its backers raised $1 billion in an IPO priced at $23 a share, more than 10 times oversubscribed, Bloomberg reports.

The chain now has more than 3,300 US and Canadian locations, plus an expansion deal with Cancro to open roughly 300 more across the UK and Ireland. Revenue grew 11% to $724 million last year, and Jersey Mike’s US sales growth has outpaced rivals like Jimmy John’s, while Subway has actually declined, as diners lean toward sandwiches over pricier bowl concepts.

Blackstone, which bought Jersey Mike’s last year for about $8 billion including debt, will retain 68% of voting power after the IPO. Cancro stepped back from CEO to chairman in April, handing the role to Charlie Morrison, who previously ran Wingstop and Salad and Go. Shares began trading Thursday on the NYSE under the ticker JMKE.

Most leaders assume performance problems start with talent. When a project misses the mark or engagement drops, the instinct is to question whether the right people are in the right seats. After years of leading schools, I have learned a different lesson: most performance problems are communication problems wearing a talent costume.

The people on your team may be highly capable. The real question is whether they have the support required to perform at their best. I learned this long before I became a business owner. I learned it standing in a classroom.

People don’t know what’s expected

As a teacher, I quickly discovered that capability and performance are two different things. I had students who were eager to learn but struggled to complete assignments or meet expectations. It would have been easy to label them as underperformers. Instead, I learned to ask a better question: have I made success clear enough for them to achieve it?

The same lesson applies to adults. One of the biggest mistakes I made early in leadership was assuming that because I explained something, people understood it. I would paint beautiful pictures of the future I was building and walk everyone through the reasoning behind a decision, believing that all that explanation would create understanding. People were still leaving meetings with unanswered questions — chief among them: how does this affect me?

When expectations are vague, people fill in the blanks themselves. That creates inevitable confusion and frustration. Whenever I introduce a new initiative or process now, I run it through five questions:

• What needs to happen?
• Who owns it?
• When is it due?
• Why does it matter?
• How will it affect each person involved?

Clarity removes guesswork. When people know exactly what success looks like, performance improves dramatically.

People don’t feel safe speaking up

Many leaders say they want feedback but fail to create an environment where feedback feels safe. Psychological safety has become a popular leadership topic, but its impact remains deeply practical: if people fear being punished for speaking honestly, valuable information never reaches leadership. That silence becomes expensive.

One of the strongest teams I ever led included employees who challenged my thinking regularly. Those conversations were sometimes uncomfortable, but they pointed out blind spots and highlighted issues I couldn’t see from my position. Those employees became some of my most trusted team members — because honesty is a form of commitment. People who care enough to tell you the truth are helping you improve. People who stay silent while problems grow are usually protecting themselves.

Leaders often assume employees hold back because they lack confidence. In many cases, employees hold back because previous experiences taught them that honesty comes with consequences. If you want better communication, examine your reaction to difficult feedback. Do people leave conversations feeling heard, or feeling punished? Your response teaches them whether speaking up is worth the risk.

Leaders assume understanding instead of confirming it

I learned this the hard way when a leadership team questioned an employee rewards program I had implemented. I thought the reasoning was obvious. It turned out I had explained what we were doing but never why. Once I walked them through the financial benefits for both employees and the business, the confusion disappeared. What I thought was resistance was simply an incomplete understanding.

Communication is only successful when the other person understands the message the way you intended it. Instead of asking, “Any questions?” try asking, “How do you see this working?” Their answer will quickly reveal whether alignment actually exists.

How to know if your team feels safe

Many leaders believe they have an open-door policy. The better question is whether people actually walk through the door.

One of the most effective tools I’ve used is anonymous feedback surveys. But the key isn’t the survey itself — it’s what happens after. Too many organizations collect feedback and never act on it. Employees notice, and trust declines fast.

Recently, I conducted a survey with leadership teams across my organization. The feedback revealed a consistent need for additional classroom support. Based on that input, we restructured resources and made significant staffing changes. The survey mattered because the feedback led to action. Trust grows when people see evidence that their voice influences decisions.

If your team consistently shares ideas along with constructive criticism, that is usually a sign of safety. If everyone agrees with everything you say, you may have a communication problem hiding beneath the surface.

A weekly clarity check-in

Improving communication doesn’t require complicated systems. Start with a simple weekly check-in built around three questions:

• What is your top priority this week?
• What obstacles are slowing progress?
• What support do you need from me?

These conversations create alignment, uncover challenges early and reinforce accountability. More importantly, they give employees space to speak honestly before frustration turns into disengagement.

The team you need may already be there

The strongest teams are built through clarity, trust and consistent communication. When people understand what is expected of them, feel comfortable raising concerns and have confidence that their leaders are listening, performance tends to follow.

It’s easy to look at a struggling team and search for individual weaknesses. But our job as leaders is to examine the environment we have created around them. That’s often where the real opportunity is. The way we communicate and clarify expectations shapes how people show up every day.

Most leaders assume performance problems start with talent. When a project misses the mark or engagement drops, the instinct is to question whether the right people are in the right seats. After years of leading schools, I have learned a different lesson: most performance problems are communication problems wearing a talent costume.

The people on your team may be highly capable. The real question is whether they have the support required to perform at their best. I learned this long before I became a business owner. I learned it standing in a classroom.

People don’t know what’s expected

As a teacher, I quickly discovered that capability and performance are two different things. I had students who were eager to learn but struggled to complete assignments or meet expectations. It would have been easy to label them as underperformers. Instead, I learned to ask a better question: have I made success clear enough for them to achieve it?

According to a recent report, 94% of 250 surveyed enterprise C-level executives plan to ramp up spending on AI visibility efforts in 2026. However, while almost all executives agree that generative engine optimization had a positive impact on their business in the previous year, a HubSpot study showed that 32.5% of marketers have no clue how to monitor AI citations — let alone measure their impact.

Unlike traditional search optimization, tracking a brand’s AI visibility isn’t as easy as opening Search Console. For many businesses, it’s not even as easy as signing up for an Ahrefs subscription — although there are already similarly designed products available. The truth is that the most effective AI visibility tracking requires a layered approach. Here’s the system I’ve been running since the start of 2026.

1. Get clear on what you’re actually tracking

Before you touch a single tool, decide what success looks like. In my experience, most founders lump together several very different signals and then wonder why their reporting tells them nothing useful.

The first is citations. A citation is when an AI engine links to your website or clearly uses your page as a source inside its answer. It is the closest thing AI visibility has to a traditional SEO signal, which is why so many teams start there.

The second is mentions. A mention is when your brand name appears inside the response, whether or not the AI links back to you. Mentions matter because they show your brand is part of the model’s vocabulary on a topic. But mentions can also flatter you. A brand can be mentioned as a passing example and still lose the commercial intent of the query.

That is why I treat recommendations as a third and separate metric. This is the question that matters most: When someone asks for the best option, does the AI actually suggest your product, company or service, or does it just acknowledge that you exist? As I wrote in my previous Entrepreneur piece on how AI recommends local businesses, being visible and being recommended are not the same thing.

If you only track citations, you can end up celebrating movement that never turns into revenue. Track citations, mentions and recommendations separately, or your reporting will blur the thing you actually care about.

2. Build a prompt library that sounds like a real customer

Nothing in AI visibility works without a serious prompt library filled with the questions a real buyer would ask to discover a brand like yours.

I always start manually. Before I ask any AI tool for help, I write the first 10 to 20 prompts myself. That matters because you already know the language your customers use, the objections they have and the competitors they compare you against. Start with the obvious commercial prompts, then expand into comparison queries, pain-point queries, and local variations.

Good prompt libraries also need specifics. Add city names where geography matters. Add competitor names where comparison matters. Add budget, company size, use case or industry where those filters would realistically shape the answer. OpenAI’s own data shows how conversational ChatGPT usage has become, which means generic one-line prompts often miss how people actually search.

Once you have that manual base, use Claude or ChatGPT to generate variants and cluster them by intent.

It’s better to have 50 good prompts than 300 bloated ones. Too few prompts and you miss the long tail. Too many, and you start tracking noise instead of buying intent.

3. Use platforms for scale, but understand their limits

A growing number of tools now cover AI visibility directly, including Peec, Semrush, Ahrefs and DataForSEO. What makes them useful is not just that they collect data. It is that they make the data operational.

A good platform can track multiple engines at once, automate daily checks, visualize trend changes, generate reports for your team and often let you set a location. Some also suggest new prompts to monitor, identify competitors you had not considered and surface content gaps that may be hurting your visibility. Once you spend the time setting them up properly, the maintenance burden is relatively low.

But there is a big catch. A lot of this tracking still depends on search-enabled environments, model snapshots or vendor-specific ways of querying the models. 

That matters because the answer a user gets from a live AI session can look very different depending on whether web search is active, what context is available and how the system decides to compose the response. In other words, platform data can be directionally useful without being a perfect reflection of what every real user sees.

This is where teams get overconfident. They subscribe to a dashboard, see a neat visibility chart and assume they now understand the market. They do not. They understand one layer of it.

That does not make the tools useless. It just means you should treat them as monitoring infrastructure, not ground truth. For a useful overview of how these products fit together, this guide on measuring AI visibility in 2026 is a solid reference point.

4. Keep a manual tracking layer for the prompts that matter most

The most labor-intensive part of AI visibility tracking is also the most revealing. Once a month, I like to take the most commercially important prompts from my library and run them manually across ChatGPT, Claude and Gemini in fresh chats.

The point of doing this is control. You can test the exact prompt phrasing, add the location directly into the query when geography matters and compare outputs side by side. You also get the full richness of the response instead of a summarized score inside a platform dashboard.

From there, I save the responses and use a high-reasoning model to analyze them. I want a clean breakdown of how often my brand was cited, how often it was mentioned, whether it was actively recommended, how prominently competitors appeared and what patterns keep repeating across answers. You can also use this layer to ask for hypotheses about why certain competitors keep outperforming you on specific prompts.

This approach takes more effort, but it gives you something automated tools often flatten: context. You see not just whether your brand showed up, but how it showed up and what narrative surrounded it.

In practice, the best setup is usually hybrid. Use a platform subscription to monitor broader patterns, and use manual checks on the prompts that actually matter to your pipeline.

5. Measure business impact, not just AI visibility

Visibility is interesting. Impact is what pays for the work.

The most obvious place to start is Google Analytics. Track identifiable AI referral traffic where possible and monitor how those visitors behave compared with other channels. That still will not show you the full picture, because some people will discover your brand through an AI answer and come back later through a branded search, direct visit or referral.

That is why I also like simple operational fixes. Add “AI assistant” as an answer option to your “How did you hear about us?” field. If your business uses sales calls, train the team to ask whether the lead first heard about you through ChatGPT, Claude, Gemini or another AI tool. It sounds basic, but this kind of qualitative data becomes surprisingly valuable once patterns start repeating.

Watch for indirect signals too. When your recommendation rate improves on important prompts, do branded search, demo requests and direct traffic rise soon after? If your visibility numbers look better but none of those downstream indicators move, something in the chain is broken.

According to a recent report, 94% of 250 surveyed enterprise C-level executives plan to ramp up spending on AI visibility efforts in 2026. However, while almost all executives agree that generative engine optimization had a positive impact on their business in the previous year, a HubSpot study showed that 32.5% of marketers have no clue how to monitor AI citations — let alone measure their impact.

Unlike traditional search optimization, tracking a brand’s AI visibility isn’t as easy as opening Search Console. For many businesses, it’s not even as easy as signing up for an Ahrefs subscription — although there are already similarly designed products available. The truth is that the most effective AI visibility tracking requires a layered approach. Here’s the system I’ve been running since the start of 2026.

1. Get clear on what you’re actually tracking

Before you touch a single tool, decide what success looks like. In my experience, most founders lump together several very different signals and then wonder why their reporting tells them nothing useful.

Key takeaways

• When buyers ask AI platforms hyper-specific, bottom-of-funnel questions, the models often hedge, hallucinate, or contradict themselves — because the authoritative source material simply isn’t there.
• The founders who publish clear, structured answers into those gaps first will become the default response AI gives for years to come.

Identifying the gap between what people want to know and the information available to them is the game — and good SEO practitioners have been playing it for years.

Generative AI is just the next evolution of that process. Traditional search shows you what people are asking. Generative AI exposes something more valuable: the high-intent questions that search engines still answer poorly. That’s the opportunity.

Stop optimizing where everyone else is

Founders naturally focus on what’s visible when building strategy. Competitive analysis provides benchmarks. Market research examines established demand. Industry commentary reveals dominant trends. Everyone looks at the same signals — which is exactly why so many strategies end up looking identical.

In traditional SEO, this shows up as fighting for contested keywords and racing to outbuild backlink profiles just to land a spot under AI Overviews and sponsored ads. Marketing budgets drain quickly, and returns rarely follow.

With Generative Engine Optimization (GEO), search has shifted from keywords to intent. Platforms like ChatGPT, Claude, Gemini and Perplexity are now where people compare products and services. Users aren’t searching in fragments anymore; they’re writing full sentences and expecting full answers.

If you want your brand recommended first in that environment, understand what these systems actually reward: the most structured, authoritative and complete knowledge on a topic. Keywords no longer determine rank. What matters is how thoroughly and clearly your company is represented to the world.

How negative space reveals opportunity

When a buyer asks a bottom-of-funnel question and the retrieval pool is thin or non-existent, the model comes unstuck. AI is trained to answer confidently — so when the information it needs is missing, incomplete or poorly structured, it hedges, invents or contradicts itself.

That’s negative space: the unanswered questions, the under-explained problems, the contradictory responses, the outright hallucinations. It exposes what’s absent from the knowledge base these models draw from.

If your brand isn’t present in that space when prospects are asking the questions that precede a decision, you’re losing deals you never knew existed.

These gaps won’t stay empty forever. The companies that publish authoritative content into them first will shape how future AI responses are generated. Timing is the competitive advantage.

Why targeting negative space matters

Targeting negative space creates informational authority where none currently exists. It’s an evolution beyond competing for visibility.

Most companies optimize for visible demand — and as traditional SEO has shown, that market is saturated, costly and difficult to rise above. Negative space, by contrast, is unmapped territory. Most companies don’t yet realize it exists. Where definitions are inconsistent and structured knowledge is incomplete, there’s room to become the answer.

The work is to find the questions where AI responses lack depth, structure or clarity, and provide the authoritative explanation. Instead of creating content for top-of-funnel curiosity, focus on the questions prospects ask when they’re preparing to make a decision. If your answer is the one they encounter — clear, credible, complete — they’ll remember it.

Strategy shifts from volume to informational leverage. AI systems don’t rank pages; they use them to construct an answer. A single, well-structured, explanatory article beats a stack of SEO-optimized copy. When your content is the first authoritative source — entity-structured, statistic-rich, citation-heavy — the AI is more likely to draw from it.

From visibility to authority

Negative space is the gap between what your prospects are trying to find out and what AI can confidently answer. Where traditional SEO optimizes for attention, AI SEO builds better interpretations for AI systems. That’s the next real competitive advantage.

Be the first credible explanation in an empty space, and you don’t just show up in the answer — you define how the topic is understood, and you’re front and center at the moment prospects are ready to act.

Key takeaways

• When buyers ask AI platforms hyper-specific, bottom-of-funnel questions, the models often hedge, hallucinate, or contradict themselves — because the authoritative source material simply isn’t there.
• The founders who publish clear, structured answers into those gaps first will become the default response AI gives for years to come.

Identifying the gap between what people want to know and the information available to them is the game — and good SEO practitioners have been playing it for years.

Generative AI is just the next evolution of that process. Traditional search shows you what people are asking. Generative AI exposes something more valuable: the high-intent questions that search engines still answer poorly. That’s the opportunity.

Stop optimizing where everyone else is

Founders naturally focus on what’s visible when building strategy. Competitive analysis provides benchmarks. Market research examines established demand. Industry commentary reveals dominant trends. Everyone looks at the same signals — which is exactly why so many strategies end up looking identical.

You’re doing the work. The phones are ringing and your reviews look good. But you’ve noticed that your leads are not the same as they were last year. Referrals that used to flow in steadily have slowed as well. You’re not losing deals. You are losing the opportunity to compete for them like you used to.

The number that should concern you: 46%. I studied 400 companies across five industries. That’s the percentage of business owners who are completely invisible to AI. Not underperforming. Not dropping to page two. Gone. When a prospect asks ChatGPT, Perplexity or any AI assistant for a recommendation, nearly half of all business owners I studied are invisible in the answer.

And here’s the part that should surprise you: Your company may show up, but if the owner behind it is not visible, AI treats the business differently. The algorithm appreciates the company info but craves information about the owner or key person in the business — who is the human behind the brand and are they credible?

The gap that nobody is talking about

Companies where both the owner and the business were visible to AI saw an average year-over-year traffic increase of 2.5%. Companies where both were invisible? Down over 28%. That’s a 30% difference — same industries, same time frame. The difference is whether the person behind the company had any presence in places AI considers trustworthy.

On the surface, a 2.5% increase seems minimal. However, Gartner research predicts that organic traffic to websites will drop by 25% this year. Over the next three years, that kind of gap translates into millions of dollars in revenue the invisible owners never enter their sales funnel — not because the work was bad, not because the market or economy shifted, but because humans and machines couldn’t see who was behind the company. This is not a marketing problem or a traditional SEO problem.

I know what you’re thinking: “So I need to post more on social media.” Nope. This is an entity recognition problem. AI doesn’t really care about your Instagram reels or TikTok followers. It cares whether you are credible on the platforms AI approves as relevant to your expertise and authority. The venues change often, but the basics include published articles, podcast appearances, speaking engagements, industry directories and local press. The common thread: places where you are findable and verifiable.

Think of it this way: Would you hire a contractor with no reviews, no references and no one who could vouch for them? AI works the same way. It loves third-party signals of trust.

The owner is the multiplier

Here’s what surprised me in the research: The company’s brand visibility matters, but the owner’s visibility is the multiplier. A business with a strong brand and an invisible owner underperforms a business with a strong brand and a visible owner or subject-matter expert — every time. The owner’s visibility doesn’t just add to the company’s credibility; it amplifies it.

This makes sense if you think about it from the customer’s point of view. People want to know who they’re trusting with their money, their home, their family’s health. A faceless logo doesn’t answer that question. A person does. AI is simply reflecting what humans have always deferred to: I trust people more than brands.

What invisible owners get wrong

Most invisible owners are busy running their business. They’re not hiding on purpose. They assume their work speaks for itself. They’ve convinced themselves that marketing can handle visibility for the company.

But the world has shifted. Your prospects aren’t searching on Google alone anymore — they’re asking AI to help them decide. And AI builds its recommendations from a variety of sources and trust signals most business owners aren’t knowledgeable about.

The good news is that you don’t need to become a content machine. You don’t need to go viral. You need to exist in the places that matter to AI. Get your name in a local business journal with your point of view related to your industry. Show up on someone else’s podcast. Speak at your Rotary club or chamber event. Film it with your phone and post it to LinkedIn and YouTube. Share what you know in a way that helps people. Forget the sales pitch. Provide genuine value. When you teach, you build authority. When you build authority, AI pays attention.

The question you need to ask yourself

If someone asks an AI assistant to recommend a business like yours in the area, does your name show up? Not your company’s name — your name. Check it now.

The data is clear. When the owner is invisible, the company pays the price. When the owner is visible, everything compounds: trust, referrals, traffic, revenue. You built the business. You hired the team. You took the risk. The one thing you need to focus on now? Let the people — and the machines — know who’s behind it all.

That’s the invisible owner problem. And 46% of you are living it right now. The good news is that visibility isn’t reserved for people with PR teams and huge marketing budgets. It’s available for any owner willing to step out from behind their logo and be known for what they do.

You’re doing the work. The phones are ringing and your reviews look good. But you’ve noticed that your leads are not the same as they were last year. Referrals that used to flow in steadily have slowed as well. You’re not losing deals. You are losing the opportunity to compete for them like you used to.

The number that should concern you: 46%. I studied 400 companies across five industries. That’s the percentage of business owners who are completely invisible to AI. Not underperforming. Not dropping to page two. Gone. When a prospect asks ChatGPT, Perplexity or any AI assistant for a recommendation, nearly half of all business owners I studied are invisible in the answer.

And here’s the part that should surprise you: Your company may show up, but if the owner behind it is not visible, AI treats the business differently. The algorithm appreciates the company info but craves information about the owner or key person in the business — who is the human behind the brand and are they credible?

Opinions expressed by Entrepreneur contributors are their own.

Running a small accounting firm in 2025 means answering a question nobody has fully solved: how do you build a business clients trust deeply when half your team is working from a spare bedroom two states away?

I have spent four years wrestling with it. The honest answer is that I got it wrong before I got it right. What I learned did not come from a management book or a productivity framework. It came from losing a client I should never have lost, in a way that had nothing to do with the quality of our work.

That experience changed how I run everything.

How I lost a client of seven years to a birthday

Not a missed deadline. Not a calculation error. A birthday. One of my remote staff members mentioned in passing that a long-term client had a milestone coming up. I nodded, said “great,” and did absolutely nothing with that information. Three weeks later, that client called, annoyed about something unrelated, and signed with another firm before the month was out. When I dug into it, the birthday was just the final straw. We had quietly become a transaction to them. They sent documents, we sent returns and somewhere along the way we stopped being people they actually knew.

That was the moment I understood what remote work actually costs an accounting firm, and it has nothing to do with productivity.

Every article you read about managing remote teams talks about communication tools, meeting cadence and tracking output. All of that matters. But none of it addresses the specific thing accounting firms sell: trust. Not software. Not efficiency. The feeling a client gets when they believe their accountant genuinely knows them and is watching out for them. That feeling is extraordinarily hard to manufacture through a screen, and most firms are not even trying.

What the office was actually doing for us

Before we went hybrid, the office was running a relationship-maintenance operation that none of us noticed or appreciated. A client would call the front desk and mention something offhand. The receptionist would relay it to the file owner. Someone would follow up. It wasn’t a system. It was proximity and human instinct doing the work automatically.

Remotely, that entire invisible operation collapsed overnight. Nobody was overhearing anything. Nobody was walking past anyone’s desk. The phone calls still came in, but they landed in a vacuum. Information stopped traveling sideways across the team the way it used to, and we didn’t notice until clients started feeling like strangers.

The fix was not what I expected

My instinct was to add more check-ins. More internal meetings. More structured communication. What I actually needed to do was much simpler and much harder. I needed every person on my team to take personal ownership of the relationships inside their files, not just the work inside them.

That sounds obvious. In practice, it means the accountant handling a business return knows that the owner’s daughter just joined the company, knows the lease renewal is coming up in spring and picks up the phone once in a while for no reason other than to check in. It means treating client files less like tasks and more like ongoing relationships that require actual attention.

We now build what I call a relationship note directly into every active file. Not tax notes. Not billing notes. Personal context — what is going on in this client’s life right now that we should know about. It takes two minutes to update, and it has changed the quality of our client conversations more than any software we have ever purchased.

The staff problem nobody wants to admit

Remote work exposed something uncomfortable about our industry. A certain type of accountant, perfectly competent technically, has no interest in the human side of the work. In an office, that person still participates in the culture whether they want to or not. They overhear conversations, they get pulled into hallway discussions, they absorb the relationship norms of the firm just by being present.

Remotely, that same person retreats completely into the technical work and becomes, from the client’s perspective, someone who processes their documents and occasionally sends a PDF. Clients sense this quickly even if they never say it directly. It shows up in the tone of emails, in the length of calls, in how often they reach out with questions — and how often they quietly start wondering if there is someone better out there.

I am not saying those staff members are bad at their jobs. I am saying that remote work removes the guardrails that used to compensate for the gaps, and as a managing partner you have to consciously rebuild them.

What I actually changed

I stopped measuring my team only by what they completed and started paying attention to how their clients talked about them. Not formal surveys. Just listening. When a client calls the front desk, what is the tone? When someone refers us to a friend, what do they say about us specifically?

I also made phone calls a non-negotiable part of the job. Not video calls, not Slack messages — actual phone calls with clients at regular intervals that have nothing to do with a deadline. Some of my staff resisted this. A few still do. But the clients who get those calls are the ones who have stayed with us through fee increases, staff changes and every other reason someone might think about switching firms.

The last thing I changed was the hardest. I stopped pretending that remote work is neutral. It isn’t. It has real costs, and they fall unevenly on client-facing professional services firms in ways a tech company or a marketing agency will never feel. Acknowledging that inside the firm made it possible to actually address it, instead of just adding another tool to the stack and hoping for the best.

That client we lost over a birthday is with a competitor now. I think about it more than I should. But it is the reason we run the firm the way we do today — and I would rather have learned it then than be learning it now.

Running a small accounting firm in 2025 means answering a question nobody has fully solved: how do you build a business clients trust deeply when half your team is working from a spare bedroom two states away?

I have spent four years wrestling with it. The honest answer is that I got it wrong before I got it right. What I learned did not come from a management book or a productivity framework. It came from losing a client I should never have lost, in a way that had nothing to do with the quality of our work.

That experience changed how I run everything.