Executive summary
Performance brands are entering a technology cycle that financial services has already lived through. The language is different. The vendors are different. The market context is different. The underlying pattern is familiar: growth creates complexity, complexity creates pressure, pressure creates technology spend, and technology spend without decision maturity creates more complexity.
This paper argues that many scaling performance brands are approaching what Corbelle calls the Performance Ceiling: the point at which organisational complexity begins growing faster than organisational capability. The ceiling is rarely visible at first. Revenue may still be growing. The brand may still be strong. Customers may still love the product. The warning signs appear behind the scenes: slower decisions, fragmented data, inconsistent reporting, rising system costs, increasing dependency on vendors, and growing uncertainty about AI.
The comparison with financial services is not used to claim that a 75-person endurance brand is the same as a global bank. It is used because financial services has already paid for a painful education in technology decision-making. Banks and fintechs learned that technology does not create strategy, data does not automatically create clarity, and AI does not compensate for weak operating discipline. Those lessons are now directly relevant to founders and COOs of performance brands.
This matters most for brands between roughly 25 and 500 people. Startups can still rely on founder intuition and lightweight systems. Large global brands can afford senior technology teams, data leaders, enterprise platforms and specialist integrators. The greatest risk sits in the middle: the scaling performance brand that is big enough for technology decisions to matter, but not yet mature enough to make those decisions consistently well.
Performance brands are not short of technology. They are short of technology decision maturity. Financial services already learned what happens when that gap is ignored.
Key findings
| Finding | Implication for founders and COOs |
|---|---|
| The maturity gap is real, but it is not only a spend gap. | Financial services spends materially more on technology, but the more important difference is operating discipline: governance, data ownership, decision frameworks and senior accountability. |
| AI is accelerating the ceiling. | AI increases the number of possible initiatives faster than most teams can evaluate them. This creates the illusion of opportunity while increasing decision fatigue. |
| Scaling brands are the highest-risk segment. | The most exposed firms are not Gymshark, WHOOP or Strava. They are the sub-500 employee brands entering their first serious phase of technology investment. |
| Vendor-led strategy is a major risk. | When internal decision capability is weak, vendor roadmaps become de facto strategy. This is one of the clearest parallels with financial services transformation programmes. |
| The next advantage is decision quality. | As technology becomes more accessible, advantage shifts from access to judgement: knowing what to do, what not to do, what to sequence and what to measure. |
Why this paper exists
Performance brands are commercially led organisations whose product, service or experience helps people perform better physically. The category includes endurance events, sports nutrition, recovery technology, coaching platforms, performance apparel, fitness communities, clubs and sport-related operating businesses.
These organisations are often founder-led, brand-rich and community-driven. They understand their customer deeply. They move quickly. They build trust through product quality, lived experience and cultural relevance. For a long time, that was enough.
The sector is being pulled into a more complex operating environment. Customers expect personalisation. Boards and investors ask about AI. Marketing teams rely on increasingly fragmented data. Events businesses manage participant journeys across multiple systems. Nutrition brands operate across DTC, wholesale, subscriptions and retail. Coaching platforms blend human expertise with algorithmic recommendations.
The result is a new kind of pressure. Founders and COOs are being asked to make technology decisions they were not trained for, in a market moving faster than traditional playbooks can cover, with vendors offering confident answers to questions the leadership team may not yet have defined.
The founder and COO problem
In a 50-person business, the founder can often hold the operating model in their head. They know the customer, the team, the systems, the bottlenecks and the priorities. Technology choices are tactical. If a tool fails, the blast radius is limited.
In a 250-person business, the operating model has changed. There are more teams, more systems, more handoffs, more territories, more channels and more stakeholders. The same decision habits that made the company fast now begin to make it fragile. The COO becomes the person expected to convert ambiguity into execution.
The Performance Ceiling is the point at which organisational complexity begins growing faster than organisational capability. It appears first as operational friction, then as decision drag, and eventually as a constraint on growth.
The hypothesis and the verdict
The original hypothesis tested in this paper is direct: performance brands are about to repeat the expensive technology mistakes financial services made over the last decade.
Verdict
The hypothesis is supported in direction and pattern, but the language should be sharpened. The evidence does not prove a precise ten-year lag. It does support a maturity gap in technology decision-making, data governance, AI operating models and executive confidence.
What is proven
| Evidence | Source | Why it matters |
|---|---|---|
| Financial services has materially higher technology intensity. | McKinsey, 2024 | Banks' IT spending reached 10.6% of revenues and 20.0% of expenses in 2022. Global banking technology spend reached approximately $650bn in 2023. |
| Sports and fitness sectors remain digitally immature. | Sport England, Digital Futures 2024 | The UK sport and physical activity sector scored 51% for digital maturity — categorised as Digital Experimenter. |
| Sports organisations are still early on GenAI strategy. | PwC Global Sports Survey 2024 | 59% of sports organisations surveyed did not yet have a GenAI strategy. |
| The main barrier is organisational, not technical. | Data & AI Leadership Exchange / Wavestone, 2025 | 92% of respondents believed the primary barrier to establishing data and AI-driven cultures was people and organisation change, not technology. |
| Spend alone does not create advantage. | Bain & Company, 2023 | Banks leading in technology deliver 5 percentage points higher total shareholder return through deft use, not simply higher spend. |
"I also think that the participation events and performance brands are about a decade behind where financial services firms were in terms of the investment and the exposure. It's an exciting time ahead especially if you can catch the front of the wave."
Co-Founder, Ultra X · LinkedIn comment, 2026
What is not proven
There is no single public benchmark that directly compares performance brands against financial services across AI maturity, data maturity, governance maturity, technology spend and decision quality. Some performance brands are highly mature — WHOOP, Garmin, Strava, Oura and other technology-native businesses are better understood as technology companies operating in the performance ecosystem. The Corbelle opportunity is strongest among commercially led scaling brands that are not technology-native.
What financial services learned the expensive way
Financial services is not presented here as a perfect model. The lesson is not that financial services got everything right. The lesson is that it has already experienced the consequences of getting technology decisions wrong at scale.
During the last fifteen years, banking and fintech underwent several overlapping waves of technology investment. The sector discovered that technology access is not enough. Without governance, architecture, operating discipline and clear ownership, technology investment can make an organisation slower, not faster.
A real illustration: when data abundance meets decision immaturity
In 2018, TSB migrated more than five million customer accounts to a new banking platform designed to support future growth.
The technology worked in testing. The migration plan was approved. The programme went live.
The lesson was not that the technology failed.
Technology was not the root cause. It was the amplifier.
The platform exposed organisational complexity that the business was not fully prepared to manage.
When technology capability grows faster than decision-making capability, performance suffers. That's the Performance Ceiling.
This pattern — platform first, strategy second — is not a failure of intelligence or ambition. It is a failure of sequencing. The technology forces every unresolved question into the open at once. Who owns the customer? Which metric matters? What will be decommissioned? A platform cannot answer those questions. It can only make the cost of not answering them unavoidable.
Lesson 1: technology does not create strategy
Many financial services programmes began with a platform decision. A bank bought a CRM. A wealth manager bought a data warehouse. An insurer bought an automation tool. In practice, the technology often exposed unresolved strategic questions rather than solving them. Performance brands are now at risk of making the same sequencing error: buying AI, CDPs, automation tools or analytics platforms before defining the business problem clearly enough to defend in a boardroom.
Lesson 2: more data does not create clarity
Financial services became data-rich long before it became insight-rich. Performance brands face the same pattern. A nutrition brand may have Shopify, Klaviyo, subscription records, retail sell-through and fulfilment data. An events business may have registration data, timing data, CRM, waiver data and volunteer data. The problem is not absence of data. It is the lack of an agreed operating view.
Lesson 3: governance beats heroic effort
In immature environments, good people compensate for weak systems. A data analyst reconciles the spreadsheet. A COO chases the number. A founder remembers why a campaign was launched. This works until scale makes heroics unsustainable. Performance brands do not have the regulatory forcing mechanism that compelled financial services to build governance — which makes voluntary governance even more important.
Lesson 4: AI amplifies capability — it does not create it
Lesson 5: decision quality beats technology spend
Financial services spends more on technology than almost any sector — yet McKinsey's analysis found that spend growing faster than revenue has not translated into productivity gains. For performance brands, this is liberating: they do not need banking-sized budgets. They need banking's hard-won discipline, adapted to their scale.
The performance brand maturity gap
The maturity gap is clearest when looking beyond headline technology adoption. Many performance brands already use modern tools. The issue is not whether they have software. It is whether they have the leadership capability to turn software into repeatable advantage.
The most dangerous version of the maturity gap is not visible from the outside. A brand can look sophisticated because its website, campaigns, app or athlete ambassadors are strong. Behind the scenes, the operating model may still be stitched together through spreadsheets, agency reporting, founder intuition and manual reconciliation.
The five places the gap appears
| Gap dimension | What it looks like in practice |
|---|---|
| 1. Strategic alignment | The leadership team does not share a single view of the most important technology question. |
| 2. Technology confidence | Executives are unsure whether they are making the right decisions or simply reacting to pressure. |
| 3. Infrastructure readiness | Back-office systems cannot support the pace, channel mix or operational complexity of front-office growth. |
| 4. AI clarity | The organisation feels pressure to adopt AI but lacks a clear view of what is worth doing, what is risky, and what should be ignored. |
| 5. Decision framework | Technology choices are made case by case, often influenced by vendors, urgency or internal politics rather than a repeatable method. |
Why the middle market is most exposed
The highest-risk segment is not the smallest startup or the largest global brand. It is the scaling business in the middle. A 20-person company can still improvise. A 2,000-person company can hire a CDO, CTO and architecture team. A 75-to-300-person performance brand often has neither the simplicity of a startup nor the institutional capability of an enterprise.
The first serious technology spend is often where the biggest mistakes happen. The brand is large enough for the decision to matter, but not yet mature enough to make the decision with confidence.
The firms halfway up the mountain
This report intentionally does not focus only on very large performance companies. Gymshark, HOKA, WHOOP, Strava, Garmin and Ironman are useful indicators of where the market is heading, but they are not the centre of the Corbelle opportunity. The more commercially useful group is the set of scaling brands about to encounter the Performance Ceiling.
A three-tier evidence structure
| Tier | Examples | Role in the argument |
|---|---|---|
| Tier 1: Market direction signals | Gymshark, HOKA, WHOOP, Strava, Ironman, Garmin | Show where technology, data and AI expectations are heading. |
| Tier 2: Scaling brand heartland | Precision Fuel & Hydration, Veloforte, Innermost, fourfive, Runna, TrainingPeaks, Threshold Sports, Human Race, Let's Do This | Show the real buyer context: founder involvement, COO pressure, partial data capability and rising complexity. |
| Tier 3: Financial services proof base | Capital One, JPMorgan, Monzo, Starling, Revolut | Show the historical pattern: high spend without discipline creates debt; disciplined operating models turn technology into advantage. |
Examples and implications
| Company / segment | What it illustrates |
|---|---|
| Precision Fuel & Hydration | PF&H combines products, sweat testing, an algorithm-based questionnaire and personalised hydration strategy. That is a strong data and personalisation foundation. The next challenge is not whether they have data — it is how that data informs retention, product, education, fulfilment, retail and customer experience decisions. That requires a single operating view, not more dashboards. |
| Science in Sport | SiS had abundant data but infrastructure that fell short of capitalising on it. Their Adobe Commerce and Adobe Experience Platform implementation is a direct example of the gap between data ambition and operating capability — and of what it takes to close it. |
| Huel | A useful example of a scaling nutrition brand moving from commerce success toward operating discipline: Snowflake on AWS for a single source of truth, Shopify Plus across global stores, and an AI and automation team with champions embedded across the business. |
| Runna / Strava (April 2025) | The acquisition demonstrates the strategic value of training intelligence, coaching, data and community. It also shows how quickly integration, platform strategy and product-roadmap questions emerge once scale arrives. The day after acquisition is when technology decision maturity starts to matter most. |
| Endurance events operators | Participation businesses manage registration, CRM, participant communications, timing data, customer support, merchandise, volunteers and event operations. Most are not technology-native. They are exactly the organisations at greatest risk of hitting the ceiling without recognising what has changed. |
A founder should not read this section and think: we are not Gymshark. They should think: we are approaching the point where Gymshark, Huel, SiS, Strava and others had to professionalise their technology decision-making.
The Performance Ceiling
The Performance Ceiling is the core concept in this report. It names a problem founders and COOs often feel before they can explain it. The business is still growing, but decisions feel heavier. The team has more dashboards, but less confidence. Tools have multiplied, but clarity has not.
What Stage 3 sounds like in practice
Decision drag — the third stage — is the moment founders and COOs most often recognise themselves. It rarely announces itself cleanly. It sounds like this:
"Every board deck takes three days to reconcile. We've got four dashboards and they all say something different."
"We know we should be doing more with AI, but nobody can agree on where to start or who owns it."
"The vendor demo was compelling. We bought the platform. Six months in and we're still arguing about what a customer actually means in our business."
"I'm still the person who knows where everything lives. That's not a compliment anymore."
These are not symptoms of bad leadership. They are symptoms of an operating model that has not kept pace with the complexity it has created.
The five stages
The symptoms
| Area | What it looks like |
|---|---|
| Commercial | Customer acquisition becomes more expensive, retention logic is unclear, personalisation is blunt, channel performance is hard to compare. |
| Operational | Manual reconciliation increases, teams build parallel trackers, processes rely on specific individuals. |
| Leadership | Board decks take longer to prepare, executives challenge the numbers, founders feel less confident in technology decisions, COOs absorb ambiguity. |
| Technology | Tool sprawl increases, integrations become fragile, data definitions vary, vendor dependency rises, decommissioning rarely happens. |
| AI | Teams experiment with tools, but use cases are not prioritised, governance is unclear, ROI is hard to evidence. |
The Performance Ceiling Index
The index scores eight dimensions from 1 to 5. A score of 1 means the capability is informal, reactive or absent. A score of 5 means the capability is embedded, owned and repeatable. The diagnostic produces a practical conversation, not a theoretical maturity score.
What the scores actually look like
Here is what a 2 and a 4 look like for the three dimensions where founders and COOs feel the most acute pain.
AI clarity
| Score | What it looks like |
|---|---|
| Score 2 — Exposed | The leadership team has seen vendor demos. There are a few ChatGPT experiments in marketing. Someone has put "AI strategy" on the board agenda. But there is no agreed list of use cases, no defined success metrics, no governance over what data AI can access, and no clear owner. When the board asks what the company is doing with AI, the answer is a list of experiments, not a decision. |
| Score 4 — Progressing | The organisation has a prioritised shortlist of AI use cases linked to specific business outcomes. Each use case has a defined owner, a data requirement that has been assessed, and a 90-day trial scope. The leadership team can say no to AI proposals that do not meet the threshold — as clearly as it can say yes to those that do. |
Data ownership
| Score | What it looks like |
|---|---|
| Score 2 — Exposed | Five people asked the same question get five different numbers. Every board deck requires a reconciliation process. The finance team and the commercial team run parallel exports. Nobody is confident which dashboard is right, but nobody wants to be the one to say it out loud. |
| Score 4 — Progressing | The business has defined its five to seven operating metrics, assigned a named owner to each, documented the source system and calculation rule, and agreed a review cadence. Disagreements about numbers now have a resolution path rather than an endless negotiation. |
Decision confidence
| Score | What it looks like |
|---|---|
| Score 2 — Exposed | The last three major technology decisions were made under time pressure, led by a vendor proposal, or approved by the founder without a written rationale that the COO could defend eighteen months later. |
| Score 4 — Progressing | Every major technology commitment is preceded by a one-page decision brief: the problem, the cost of inaction, the alternatives considered, the owner, the success measures, and what will be stopped or removed as a result. The brief is the thing that stops the team buying the demo. |
The full index
How to interpret the score
Take the diagnostic
Sixteen scenario questions scored across eight dimensions. Takes under five minutes. Your results include a personalised radar chart, dimension scores, and a prioritised action list — generated from what you actually answered, not a generic template.
Mistakes to avoid
The following mistakes are adapted from financial services, but translated for scaling performance brands. They are not theoretical. They are the predictable failure modes that appear when technology investment runs ahead of decision maturity.
| Mistake | How it shows up | Better decision |
|---|---|---|
| Buying the tool before defining the exam question | A founder asks which AI tool to buy before defining the business problem, workflow, owner, data requirement and success metric. | Start with the decision or process that needs to improve. Technology comes last. |
| Letting the vendor set the roadmap | The platform demo becomes the strategy. The business starts solving the vendor's problem rather than its own. | Create an internal decision brief before seeing vendor proposals. |
| Mistaking dashboards for clarity | The company has more reporting but leaders still argue over the numbers. | Define metric ownership, source of truth and decision cadence. |
| Over-automating weak processes | AI or automation is applied to a workflow nobody has properly mapped. | Fix the process before automating it. |
| Ignoring decommissioning | Every new tool is additive. Nothing is removed. Cost and complexity compound. | Make decommissioning part of every business case. |
| Treating data quality as an IT issue | The team blames systems when the root cause is unclear ownership, weak definitions or inconsistent behaviours. | Assign data ownership to business processes, not only systems. |
| Building business cases on uncertain AI economics | Savings are assumed before usage cost, governance, review and retraining are understood. | Model best, expected and downside case. Revisit after 90 days. |
| Scaling without operating cadence | The leadership team reviews technology episodically, usually when something breaks. | Create a quarterly technology decision review linked to strategy, risk and performance. |
Actions for founders and COOs
The practical response is not to slow down ambition. Performance brands win because they move with conviction. The response is to increase the quality of decisions before spend accelerates. Treat technology as an operating discipline, not a purchasing category.
-
1Create a technology decision brief
Before any major spend, write one page defining the problem, cost of inaction, desired outcome, owner, dependencies, risks, success measures and what will be stopped or removed.
-
2Establish a quarterly technology decision forum
Bring founder, COO, finance, commercial, operations and relevant technology or data owners together quarterly to review priorities, risks, spend and decommissioning.
-
3Build a single commercial truth layer
Identify the 5 to 7 metrics that actually run the business. Define owners, source systems, calculation rules and review cadence.
-
4Separate experimentation from operating commitments
AI pilots should be cheap, bounded and reversible. Platform commitments should require a stronger governance threshold.
-
5Appoint an accountable technology decision owner
This does not always mean hiring a CTO or CDO. It may mean assigning decision accountability to a COO with external support and clear governance.
-
6Use optionality as a decision criterion
Ask how easy it would be to exit, integrate, change, scale or decommission each major tool.
-
7Measure decision quality, not just project delivery
Track whether decisions improved speed, confidence, cost, risk, revenue or customer experience. Delivery on time is not enough.
-
8Review AI through operating readiness
Only scale AI where process clarity, data quality, accountability and risk ownership are sufficient.
Before signing the next platform contract, ask: would we still make this decision if we had to defend it to the board eighteen months from now?
Conclusion
Performance brands are entering a decisive period. The sector has strong products, loyal customers, rich communities and growing data. It also faces increasing complexity. AI will accelerate that complexity. So will international expansion, channel diversification, personalisation, subscription models, retail partnerships and investor expectations.
The brands that win will not simply be the brands that buy the most technology. They will be the brands that make better decisions about technology. They will know what problem they are solving, which capability they are building, which risks they are accepting, which costs they are creating and which decisions can be reversed.
Financial services has already learned this lesson. The tuition fee was enormous. Performance brands do not need to pay it again.
Before your next board meeting, ask yourself: where are we most likely to waste money in the next twelve months because our decision quality is not yet good enough? If you can answer that question with confidence, you are already ahead of most scaling brands in this sector. If you cannot, the work starts here.
Source notes
- McKinsey & Company, Managing bank IT spending: Five questions for tech leaders, 18 October 2024. Reported that banks' IT spending reached 10.6% of revenues and 20.0% of expenses in 2022.
- McKinsey & Company, Unlocking value from technology in banking: an investor lens, 23 October 2024. Reported global technology spending in banking of approximately $650bn in 2023 (original data: Gartner, 2024 Enterprise IT Spending Forecast for Banking and Investment Services, 2023).
- PwC, 8th Global Sports Survey, 2024. Reported that 59% of sports organisations surveyed did not yet have a GenAI strategy.
- Sport England / ukactive, Digital Futures 2024 report, 28 November 2024. Reported a 51% average digital maturity and effectiveness score for the UK sport and physical activity sector, categorised as Digital Experimenter. Note: this survey covers fitness and leisure operators, national governing bodies, active partnerships and sports clubs — a strong sector proxy, though not a direct measurement of performance brands as defined in this report.
- Data & AI Leadership Exchange / Wavestone, 2025 AI & Data Leadership Executive Benchmark Survey. Reported that 92% of respondents viewed people and organisation change as the primary barrier to establishing data and AI-driven cultures.
- Bain & Company, How banks can parlay technology into a competitive edge, 2023. Analysis of the 42 largest global banks found that technology leaders deliver 5 percentage points higher total shareholder return through deft use, not simply higher spend.
- The Pixel, Science in Sport personalisation case study. SiS's challenge capitalising on abundant data and implementing Adobe Commerce and Adobe Experience Platform.
- Snowflake / Shopify / n8n, Huel customer case studies. Described Huel's data infrastructure, global commerce stack and AI and automation programme.
- Strava press release, Strava to acquire Runna, 17 April 2025.
- Monzo Annual Report 2026. Reported 15.2m customers, 10.4m monthly active users and £73.0bn card spend.
- Revolut Annual Report 2025. Reported revenue of £4.5bn and profit before tax of £1.7bn.
- Precision Fuel & Hydration public materials and Sports Business Journal coverage.