The 5 core principles of successful scale-ups

If every awkward customer decision still lands in your Slack, you have not scaled. You have built a bigger business around the same decision bottleneck.

“Can I just check this before I reply?” “Do we still honour last year’s discount?” “Sarah is unhappy. Do you want me to call her, or are you handling it?”

This is not delegation, it is founder dependency. The company has added people, customers and revenue, but it has not built enough judgement into the team. So the same decisions keep coming back to you, through more channels, at higher speed, with more people waiting.

That is the difference between growth and scale. You can grow by adding people, tools and meetings. You scale when revenue climbs without dragging cost, complexity and founder involvement up at the same speed.

Scaling Up, the operating system popularised by Verne Harnish, is built around four decisions: People, Strategy, Execution and Cash. Those pillars matter. But they are useless if daily behaviour contradicts them. A one-page plan will not save a company that hides bad news. A cash plan will not save a company that hires people who need constant supervision. A meeting rhythm will not save a company where every awkward decision still waits for your nod.

The five principles below are the daily tests. They show whether the business can keep scaling without turning you into quality control, escalation point and the person everyone hands their anxiety to.

What changes when you scale up a business?

When you are small, being everywhere feels useful. You hear the customer calls. You spot the weak hire. You jump into delivery. You rescue the project before the customer knows it is wobbling.

That is useful at 10 people. It becomes a liability at 100.

Somewhere between 50 and 100 people, the job changes. Complexity is no longer an occasional annoyance. It becomes the work. The customer is further away from you. Your best people are stretched across more decisions. Managers have picked up people leadership without any training. Culture is no longer what you say in the all-hands. It is what happens in Teams, Slack, sales calls, delivery meetings and performance conversations when you are not there.

This is where you hit the Founder Ceiling. The company cannot grow past what you are willing to let go of. Not because you are stupid. Usually the opposite. You are good at too many things, so the business keeps using you as the answer.

But scale is not created by making you faster. It is created by building a business that can make better decisions without waiting for you.

What are the 5 principles of successful scale-ups?

The five principles are simple. That is not the same as easy.

1. Put the customer at the centre of every operating decision.
2. Hire for the business you are becoming, not the business you used to be.
3. Make bad news travel fast.
4. Keep teams small enough to own outcomes.
5. Turn managers into coaches.

They depend on each other. Customer obsession fails if you hire people who do not care. Small teams fail if managers cannot coach. Honest communication fails if bad news is punished. Hiring fails if you tolerate people who hit the number and damage the culture.

You do not need more values on a wall. You need operating principles that change what people do on a Tuesday afternoon when the customer is annoyed, the margin is thin and you are not in the room.

1. Put the customer at the centre of every operating decision

Customer obsession is not a poster. It is what people do when the easy option is worse for the customer.

Who gets listened to in meetings? What gets measured? Which customer complaints reach the people who can change the system? How quickly can a frontline person fix a billing error, a broken promise or a missed handover without asking their manager for cover?

At Rackspace, Fanatical Support^TM worked because the customer was not a department. The customer was the point of the company. That distinction matters. A customer-service team cannot rescue a business whose operating model makes it hard to do the right thing.

The difference showed up in the way we treated SLAs. Competitors often used them to define the worst acceptable experience. If the SLA said a ticket would be answered in 15 minutes, the customer got a response at 14 minutes and 59 seconds. We had SLAs, but the contract was not the ambition. If we breached one, we paid out. We did not wait for the customer to notice an outage and complain. We told them, credited them and fixed the cause.

That is not soft customer service. It is commercial. In 2001, Rackspace UK was selling servers from around £99 a month. Five years later, the UK business had £26m turnover and some customers were spending around £50k a month. That did not happen because we had a better poster on the wall. It happened because customers trusted that the promise meant something.

This is why Net Promoter Score has served me well. Fred Reichheld, who created NPS, did not design it as a vanity score. He introduced the thinking in the Harvard Business Review article “The One Number You Need to Grow” and developed it in The Ultimate Question. If the person running your NPS programme has never read either, do not be surprised when it fails.

The same applies when companies change the question, change the calculation and then complain the system does not work. That is like deciding to fix a car without a repair manual or tools. You are not running NPS. You are running a survey that happens to use the same initials.

The point is not the score. The point is seeking criticism. If a customer is not where you need them to be, you fix it. If you only have 10 major customers, do not hide behind poor survey response rates. Ring them. Ask the question. Listen hard. If a strategic customer will not respond, that is a signal, not missing data.

At Rackspace, we told every director to speak to at least one customer every week. That meant every senior meeting carried live customer evidence, not anecdotes filtered through four layers of management. When we made decisions, we were not guessing what customers cared about. We had just spoken to them.

Macquarie Technology Group is worth studying here. Joseph Michelli wrote Customer Magic about the way Macquarie built a customer-obsessed operating system. The lesson is not “send more surveys”. It is: know whether the customers who drive the bulk of your revenue actually love you, and put enough ownership behind the programme that feedback turns into action.

If you are serious about scaling, customer feedback cannot live in a survey dashboard somebody remembers once a quarter. It has to show up in your weekly rhythm, product decisions, hiring standards and team scorecards. Revenue tells you what already happened. Customer behaviour tells you what is about to happen.

The test is simple. Can someone close to the customer make a sensible promise, solve a problem and protect the relationship without waiting for three layers of approval?

If not, you do not have customer obsession. You have a nice sentiment trapped behind permission.

2. Hire for the business you are becoming

The team that got you here may not be the team that gets you there.

That sentence is uncomfortable. Good. Loyalty matters. Institutional knowledge matters. But a scale-up cannot pretend the next stage needs the same capability as the last one. This is especially true of the leadership team.

You are usually running the biggest company you have ever run. The early leadership team is often there because they were strong individual contributors, then became managers, and now have to manage through managers. That is not the same job with a bigger title. As you approach 100 people, a real management layer starts to form. The systems that worked at 50 start to creak. They need changing again as you move towards 300.

Some people make that jump. Some do not. The danger is pretending loyalty can replace capability. You need people who understand the pace of the next stage, know which things break before they break, and can see which problems need fixing now versus which ones the business can live with for another quarter.

The research backs that up. The ScaleUp Institute’s 2025 Annual Review names talent and leadership as one of the main constraints on scale-up growth. The Department for Education’s SME Skills Horizon 2025 makes the same point from the SME side: skills and recruitment are not admin. They decide whether growth turns into capability or chaos.

In one Scaling Up masterclass, I used the Topgrading contrast to make the point. Brad Smart’s process at GE aimed for 85% confidence that a hire would be an A-player. In the UK data I referenced, only 25% of hires were A-players after 12 months. Another 25% became obvious C-player mistakes. The expensive middle was the remaining 50%: B-players who were not bad enough to fire, but not strong enough for the next stage.

At Rackspace, that meant looking outside the obvious talent pool. People from other hosting companies often arrived with the same service-cost mindset we were trying to kill. Some of our best hires came from hospitality: waitresses, bar staff and door people who already understood pace, pressure and customers. They did not need to be persuaded that service mattered. They already felt it.

That is where scale-ups get stuck. Not with spectacularly bad people. With decent people in roles that now require more range, speed, judgement or leadership than they can provide. They say yes in the meeting, wait for you afterwards, then ask, “What do you want me to do?”

The answer is not to go hunting for a Google CV. It is bigger, not big. If you are moving from 50 to 100 people, someone who has helped build from 100 to 300 can be gold. They know what team structure, cadence, decision rights and management habits need to be in place before the next stage arrives.

The hiring questions get harder:

• Has this person already operated at the next level of complexity?
• Have they managed managers, not just managed individual contributors?
• Do they know what breaks between 100 and 300 people?
• Do they raise the standard of the people around them?
• Can they make decisions without pulling you back in?
• Do they care about customers when nobody is watching?
• Are they a culture fit as well as a performance fit?

A high performer who damages trust is not an A-player. They are a problem with a good CV.

3. Make bad news travel fast

Scale-ups do not die because bad news exists. They get into trouble because bad news is late, softened or hidden. One of our values at Rackspace was simple: bad news first, no surprises.

Honest communication buys you time. It lets you fix problems while they are still small. It stops gossip replacing truth. It stops customers learning about issues before your senior people do. It lets managers coach performance before frustration hardens into resentment.

This is especially difficult in founder-led businesses because people learn what you reward. If you react badly to mistakes, people bring you good news and hide the rest. If you only praise heroic rescue work, people learn to tolerate chaos until it becomes dramatic enough to be valued.

At Peer 1, this became visible through an award called Cock-up of the Month. It was inspired by Henry Stewart‘s The Happy Manifesto, which I used as part of the cultural handbook for new starters alongside the formal HR material. One of Stewart’s 10 principles is to celebrate mistakes. The point was not to laugh at incompetence. It was to make it normal to say, “We got this wrong”, learn from it and stop the same mistake hiding in three other places.

I remember dealing with the CTO of an overseas market research business after a serious hardware failure. Their data had been trashed and the backup was corrupt. One of our newer technical people had spotted a loose cable, plugged it back in and made the situation worse. I rang the CTO straight away, told him what had happened and took responsibility. He told me one of his customers had fined him £10,000. I offered to pay it. He would not let me, but we upgraded the infrastructure and changed the process so it could not happen again.

That is why “bad news first” matters. It is not about enjoying failure. It is about protecting trust while there is still time to do something useful. At IT Lab, we had to knock the old MSP habit out of people who blamed delivery delays on BT Openreach. The customer does not care whose fault it is. They care whether you take responsibility and get it fixed.

Bad-news-first cultures are not negative. They are calmer. The truth arrives early enough to be useful.

You can see it in the rhythm of the business. Do team meetings include customer complaints, missed commitments and broken processes? Are people allowed to say “I got this wrong” without being punished? Do managers practise direct feedback, or do they wait until performance review season and unload six months of irritation in one meeting?

If your business cannot tell itself the truth, it cannot scale. The same problems keep turning up in different places, customers feel the inconsistency and your best people get tired of pretending everything is fine.

4. Keep teams small enough to own outcomes

Big companies love departments. Scale-ups need customer ownership. I like teams of no more than 12, and preferably five to nine. Big enough to hold the work. Small enough to know each other.

The mistake is copying the big-company org chart too early. Finance builds finance expertise. Sales builds sales expertise. Support builds support expertise. Engineering builds engineering expertise. The functions get tidier, but the customer has to travel through more handoffs to get anything done.

The idea clicked for me when I read The Service Profit Chain by James L. Heskett, W. Earl Sasser Jr. and Leonard A. Schlesinger. They showed how customer value gets destroyed at the boundaries between departments. Every handoff is a place where service drops.

You still need functional expertise. Engineers need engineering standards. Account managers need sales leadership. Customer experience people need a craft home. But day to day, the work should sit with small cross-functional teams wrapped around a defined chunk of customers, revenue or effort.

At Rackspace, Peer 1 and IT Lab, I used versions of this model. At Rackspace, teams owned chunks of revenue. At IT Lab, the top customers had the best account manager, customer experience support and senior engineering wrapped around them. They had a clear route into the team, and that team had a daily huddle to talk about what happened yesterday, what was happening today, what else mattered and what needed to happen next.

This is not a free-for-all. People can still belong to their function, but the team owns a customer set and a scoreboard. Revenue, margin, customer satisfaction and delivery quality sit in the same conversation. At IT Lab, that operating shift helped move NPS from -7 to +55 while the business became more profitable.

The test is not whether the team is busy. The test is whether the team owns a measurable outcome that matters to the customer and the business. “Support response time” is activity. “Customer renewed because the issue was fixed before escalation” is an outcome. Better for the staff. Better for the customers. Better for the bottom line.

Your external service will only ever be as good as your internal service. If the inside of the company is full of silos, queues and blame, the customer will feel it eventually.

If nobody owns the outcome, you will.

5. Turn managers into coaches

When a business scales, you cannot be the only person developing people. But “turn managers into coaches” is too easy to say unless you give them a rhythm that actually happens.

This is where the business quietly breaks. You promote the best salesperson, engineer, consultant or operator into management. That person knows the work. They do not necessarily know how to coach performance, give feedback, run useful one-to-ones, handle conflict or build successors.

So they either avoid the people work, or they manage by answering every question themselves. Both routes create dependency. The manager becomes a smaller version of the same bottleneck.

The practical answer is the 10-minute weekly check-in. When I spoke to Jim Harter, Gallup’s Chief Scientist for Workplace Management and Wellbeing and co-author of It’s the Manager, this was the management habit that mattered. Not a top-down telling-off. Not the manager reading out a score. A short, regular coaching conversation around the employee’s own scorecard.

This is why I do not treat engagement as fluffy HR. Gallup’s 2026 State of the Global Workplace keeps employee engagement, wellbeing and job climate in the same conversation, using 2025 data. Its Q12 meta-analysis links engagement with outcomes including customer loyalty, profitability, productivity, turnover and absenteeism. In other words, the weekly check-in is not an HR nicety. It is part of the operating system.

Gallup’s Q12 starts with a brutally simple statement: “I know what is expected of me at work.” If someone does not have three to five clear measures, tracked daily if possible and weekly at worst, they do not know whether they are winning. That is demoralising. A scorecard gives the work shape. It tells people whether their effort is going into the right place.

Then the weekly check-in becomes easy. Most of it should be praise for a job well done. That matters because another Q12 item asks whether someone has received recognition or praise for good work in the last seven days. If the person is on track, the manager reinforces the behaviour. If they are off track, the manager coaches. What is blocking you? What will you do next? What support do you need?

Weekly is better than fortnightly. If you do it less often than every six weeks, do not kid yourself that it is a rhythm. People know you do not care. This is continuous performance management. It is how you keep A-players engaged, spot struggling people early and stop annual appraisals becoming twelve months of avoided conversation.

It also tells you which managers can actually manage. Look at the team’s Q12 score. Look at whether weekly check-ins are happening. If engagement is low and the check-ins are not happening, you have a manager problem. Coach them, push them or move them.

Coaching is not softness. It is how you multiply judgement. If your managers cannot run this rhythm, every performance issue, people decision and awkward conversation comes back to you.

How do you know which principle is breaking?

Do this before you build another initiative. Five minutes will tell you where the bottleneck really is.

• Customer obsession: Find the last three customer complaints that mattered. Did the team fix the system or just soothe the customer?
• Hiring quality: Name the people you would rehire tomorrow for the next stage. Name the people you keep explaining away.
• Honest communication: Look at your last senior meeting. What bad news arrived late, vague or not at all?
• Small teams: Pick one key customer outcome. Can you point to one team that owns it end to end?
• Manager coaching: Ask your managers what they are coaching this month. If they only describe tasks, you have a management problem.

This is where scale fails: in the gap between what you say the company values and what the business actually rewards.

If you are trying to choose an operating system, read the definitive guide to what Scaling Up means. If the problem is meeting rhythm, start with daily huddles. If the issue is that your senior people do not yet lead as a unit, look at executive team coaching.

Frequently asked questions

What is the difference between growing and scaling a business?

Growing a business usually means revenue goes up while headcount, costs and complexity rise with it. Sometimes productivity falls and profitability gets worse. Scaling a business means revenue rises while cost and complexity grow more slowly than revenue. Complexity stays manageable, productivity improves and profit per chosen unit goes up over time.

What are the four pillars of Scaling Up?

The four pillars in Verne Harnish’s Scaling Up framework are People, Strategy, Execution and Cash. People asks whether the right people are in the right seats. Strategy says where the company is going. Execution creates the rhythm and accountability to deliver. Cash keeps growth funded and controlled.

Why does your founder-led company stall as it scales?

Your founder-led company stalls when you remain the main decision point. The business adds more customers, people and problems, but the authority to solve those problems does not move into the team. You become the system. That works for a while, then it caps growth.

How big should teams be in a scale-up?

There is no magic number, but a scale-up team should be small enough to own a clear customer or business outcome, make decisions quickly and learn from the result. Once a team becomes a pass-through layer, splits responsibility across too many handoffs or waits for senior permission, it is too large or too poorly designed.

What should a founder stop doing first?

Stop being the default escalation point for decisions the team should be capable of making. Pick one recurring decision, define the rules, give a manager ownership, then review outcomes rather than approving every step. That is how authority starts moving out of your head and into the business.

What should you do next?

---

If this post has annoyed you slightly, good. The issue is probably not effort. It is design. The business is asking you to carry decisions, standards and exceptions that should now belong inside the team.

The goal is not to disappear. The goal is to build a company where your best work is not dragged back into every operational tangle.

That is the point. Scaling is not adding more people around the same bottleneck. It is rebuilding the business so the bottleneck is removed.

Four ways to take this further

1. Book a call. If growth is now making the company slower, heavier or more dependent on you, I can help you decide whether the constraint is people, strategy, execution, cash or your role as founder. No obligation, no pitch. You will know quickly whether this is the right kind of help.
2. Grab the book. F**k Plan B covers these principles in more depth, with the practical founder lessons behind customer obsession, honest communication, hiring, small teams and managers who coach.
3. Watch the £30m scaling video. Start there if you want the founder-level version of these principles, using Rackspace and Peer 1 as the proof base.
4. Subscribe to the newsletter. Get direct, practical thinking on scaling, founder bottlenecks, leadership rhythm and building a company that can run without you in every room.

Your move. Open Slack, Teams or your inbox. Find the decision that should not have come to you this week. That is where the scaling work starts.