I spoke to a room of organisation leaders recently, from across business, university and college. The topic was the culture of organisations using the right balance of creativity and technology. The conversation that followed shaped this piece.
TLDR
Business is harder. Customers are hesitating. AI is reshaping how people research and decide. The companies growing through this period are doing five things deliberately - putting customers first as an operating discipline, using creativity to out-play not out-spend, applying emerging technology to visibility, experience and efficiency, thinking entrepreneurially regardless of size, and enabling their staff to thrive in the new reality. None of it works without culture. And the 70/20/10 split - across what's working, what's emerging and what's experimental - is how you focus the effort to do it.
Full article (9 minute read)
Business is harder than it has been for a long time. Costs are up. Teams are stretched. Customers are more cautious. Decisions take longer. Add a wave of new technology moving faster than any leadership team can comfortably keep up with, and you have a context where most companies are running just to stand still.
We've had over fifteen years of low interest rates. The last three have been different, and the pressure is showing. UK SME confidence dropped from 66% in Q3 2025 to 51% in Q1 2026 - back to where it was in late 2023. That's a structural shift in how cautious the market feels.
The pressure isn't evenly distributed. Large organisations need to act small, move faster, decide quicker, ship in weeks rather than quarters. Smaller organisations face the opposite pressure. They need to professionalise. Take governance, data and AI seriously. The standard their customers expect has gone up.
The companies getting left behind watching it, rather than leaning into technology. Treating AI as a trend to monitor rather than a capability to create advantage from.
The bigger shift, and the one most leadership teams are still underestimating, is happening on the customer side.
Customers are overwhelmed, and hesitation is where growth disappears
People are researching across more channels than ever. More options. More noise. When information is abundant and clarity is absent, the natural response is hesitation.
Gartner's research is striking on this. 77% of B2B buyers describe their last purchase as complex or difficult. Challenger's data is more striking still: 60% of complex deals end in no decision at all. The biggest competitor most businesses face isn't another supplier. It's their customer deciding to wait.
At 7DOTS we call this the Confidence Gap. The space between intent and action. The moment someone is genuinely interested but not yet confident enough to move forward. In 2026 it's getting wider.
There's a second shift underneath it that's only just becoming visible. Over 41% of B2B buyers now begin their research in AI, not Google. ChatGPT, Claude, Perplexity, Gemini. They're outsourcing the research itself, and increasingly the decision itself, to an agent.
If your business isn't visible, credible and well-described where AI is learning about your category, you're being filtered out of conversations before a human ever sees you.
The five things the winners do
Looking across our client base and the wider market, the businesses growing through this period share a pattern. Five things, done deliberately.
1. Think customer-first, operationally
Every business says they're customer-first. The ones that genuinely are can describe their customer's decision-making process better than their own org chart. They know where the customer hesitates. They know what the customer is afraid of. They know what would make the customer feel safe enough to act.
Bain's research is worth keeping in mind. They asked 362 companies whether they delivered a superior experience. 80% said yes. They asked the customers. 8% agreed. The gap between what we think we deliver and what people actually experience is where growth quietly disappears.
The leadership teams doing this well make customer outcomes the first thing discussed in every leadership meeting. The actual human outcomes, not the metrics. That single ritual reorganises an organisation faster than any restructure.
2. Use creativity to out-play
When everyone has access to the same channels, the same data, the same playbooks and the same AI tools, creative thinking becomes the only durable advantage.
I'm not talking about aesthetics. I'm talking about finding the sharper angle. The more useful story. The proposition that shifts how someone feels about moving forward. When everyone has the same tools, the differentiator is the thinking pointed at them.
McKinsey put 16 years of data behind this. The most creative companies are 70% more likely to grow faster than their peers. That's a category of competitive advantage most leadership teams treat as nice-to-have.
3. Use emerging technology as operational leverage
The winners aren't running pilots and slide decks about AI. They're embedding it into how they actually operate. The work breaks down into three things, and the businesses pulling ahead are doing all three.
Optimise for LLM and AI visibility.
Over 41% of B2B buyers now start their research inside an AI, not Google. If the answer that comes back about your category doesn't mention you, or worse, mentions you inaccurately, you're losing deals before a human has even looked at you. The work here is making sure your business is visible, credible and well-described in the places where AI is now learning. That means structured content, clear positioning, accurate third-party sources, and a deliberate approach to what's now being called GEO, generative engine optimisation. SEO used to be about being findable. GEO is about being recommendable.
Improve your digital experience through emerging tech.
This is where AI starts to change what customers feel. Smarter onboarding. Personalised journeys. Customer-facing assistants that answer questions in your brand voice, on your content, so people don't get sent off to ChatGPT to find out about you. We're building customer-facing LLMs for clients to do exactly this, trained on their proposals, case studies, pricing logic, tone of voice. The point isn't novelty. The point is that the experience feels easier, faster and more useful at the moments that actually matter to the customer.
Increase operational efficiency.
This is where the unit economics shift. AI Enablement workshops at 7DOTS are designed for this part, leadership sessions where we map every function in the business and ask one question: where could AI take a real cost out, a real delay out, or a real point of friction out? Then we prioritise three and build them. The internal LLMs we build for clients live in this space too, trained on their proposals, case studies, pricing logic, tone of voice and client history, so teams move in minutes rather than days. A small team that uses AI well now operates like a team three times its size. A large team that doesn't is spending more money to move slower.
The adoption numbers tell a useful story. 88% of companies are using AI in at least one function. Only 5.5% are seeing meaningful bottom-line impact. MIT ran a separate study and landed on the same number, 5% of companies are genuinely built to use AI. Everyone has the same tools. The difference is whether you've redesigned how you work around them, or just bolted them onto the existing model.
The example I keep coming back to on operational efficiency is Medvi. Matthew Gallagher. Forty-one years old. Two months of building. Twenty thousand dollars in start-up capital. One employee… his brother. Launched a telehealth business in September 2024. By the end of 2025 it had done $401 million in revenue with 250,000 customers. On track for $1.8 billion this year. The code, the website, the marketing, the customer service, all built using ChatGPT, Claude, and a handful of other tools. He didn't replace doctors and pharmacies. He plugged into platforms that already provided them. What he built was the brand and the experience, at a speed and a cost that would have been impossible eighteen months ago.
The honest second half matters too. Medvi is now facing an FTC investigation request, a 100,000-person class action over alleged spam, and serious questions about its advertising practices. The leverage that built it that fast is the same leverage that made the governance fragile.
Two lessons in one story. The ceiling on what a small, well-equipped team can build has moved by an order of magnitude, and your competitors know it. And the faster you go, the more deliberate your governance has to become. Speed without judgement is just risk at scale.
If Medvi is the cautionary version, Octopus Energy is the patient one. Founded in 2015. By 2025 the largest household energy supplier in the UK, overtaking British Gas, which had a 200-year head start. Nearly 10 million customers worldwide. £7.8 billion in revenue. They thought customer-first in a sector that had spent decades treating customers as billing records. They led with creative propositions like Zero Bills, which offered homeowners ten years of free energy if they bought a fully-electrified property, and Agile, which lets customers run a load of laundry for free when there's surplus on the grid. And they built an internal AI platform called Kraken to run the business. Kraken now handles 62% of their customer service, with AI-answered emails scoring higher customer satisfaction than human-answered ones. Then the punchline. Kraken worked so well internally that they spun it out last year at an $8.65 billion valuation. Other utilities now pay them to use it. The internal AI capability became a second business.
Octopus is the rarer story. A business that's hit all three, visibility, experience and efficiency, and stayed the course. The constraint isn't the technology. It's the thinking applied to it, and the governance around it.
4. Think big, act entrepreneurially
This one cuts both ways. Small companies need to think like big companies. Big companies need to think like small ones.
The small-company shift is about professionalising. Data security. Guardrails. AI policy. Governance. The standard customers, regulators and AI agents now expect has gone up sharply, and the businesses that built fast and loose are finding the cost of that catching up with them. Medvi is the cautionary version of this - extraordinary growth, fragile foundations. Smaller, growing businesses need to take the boring infrastructure stuff seriously, or it limits how big they can become.
The big-company shift is the opposite problem. Large organisations carry complexity, legacy and internal friction that slows everything down. The ones pulling ahead have stopped trying to plan the whole journey before they start it. They run smaller bets, more often. They give teams permission to ship something imperfect rather than wait six months for something polished. They behave like a series of small companies inside one bigger one.
What both sides need is the same. Clear direction from leadership, so people know where they're heading even when they don't know exactly what to do next. The agility to move quickly when something works, and quickly when something doesn't. And the discipline to learn, fast, from both.
McKinsey's work on this is clear. Organisations that crack agility lift decision-making speed by five to ten times, and 65% of them report their financial performance improves as a result. The biggest unlock isn't strategy. It's speed of learning.
It's the discipline most leadership teams find hardest, large or small. It requires being comfortable with imperfect decisions made quickly, rather than perfect decisions made too late.
5. Enable your staff to thrive
This sits underneath the other four. It's the one most leadership teams are getting wrong.
You can have the sharpest customer focus, the best creative thinking, the most ambitious AI strategy and the most agile operating model. None of it works if your people aren't equipped, supported and trusted to deliver it. Staff enablement is the multiplier. Without it, the other four are slide decks.
Three things matter here.
The first is giving people the tools.
That sounds obvious. The gap between leaders saying "use AI" and actually equipping the team with the right tools, licences, training and worked examples is enormous. I see businesses every month where the executive team has ChatGPT and the front-line team doesn't. That's a culture signal whether it's intended as one or not.
The second is giving people the skills.
AI is changing what good looks like in almost every role — marketing, sales, customer service, operations, finance, legal. The skills your team had eighteen months ago aren't the skills they need now. Companies investing in proper, structured, role-specific learning are pulling ahead quickly. Companies that aren't are watching their best people figure it out themselves and grow frustrated that the organisation isn't keeping up.
The third is addressing the elephant in every team meeting.
There is genuine anxiety about what AI means for jobs. McKinsey's most recent State of AI work found 32% of organisations expect AI to decrease their workforce in the coming year. People know that. Pretending otherwise erodes trust faster than anything else a leadership team can do.
The cultures getting this right are direct. They tell their teams: we're not shrinking the team, we're expanding what each person can do. We need you to learn this. We'll invest in helping you. The roles will change, and we'll change them together. That's a different conversation from "use AI to be more productive," and it's the one that actually unlocks the experimentation you need.
The companies doing this well treat their own people as their first AI customer. Before they roll AI out to clients, they roll it out internally. Before they ask the sales team to use a new tool, the leadership team uses it themselves. The team can see whether the executive layer believes in what they're being asked to do.
None of this works without culture
You can't tell people to be more entrepreneurial and then punish them for the first thing that goes wrong. You can't ask for creative thinking and then approve only the safe option. You can't roll out AI tools and then leave your team guessing about what they're allowed to use them for.
The cultures that get this right do three things in parallel.
They set clear direction at the top, so people know where they're heading even when they don't know exactly what to do next. Leaders model the behaviour. They use the AI tools themselves. They tell customer stories in every meeting. The culture is what leadership demonstrates, not what it declares.
They give teams permission to experiment, inside guardrails. Time-boxed pilots. Defined budgets. Clear success criteria. The freedom to try, the discipline to stop.
They make it safe to be wrong. Honest mistakes made in good faith are separated from careless ones. Cultures that confuse those two end up with neither speed nor safety.
A practical way to focus all of this: 70/20/10
The natural question for any leadership team hearing this is: how do I do all of this without breaking the engine that's already producing revenue? The answer is a portfolio approach. And it's as much about focus as it is funding. Where leadership attention goes, money and energy follow. Get the focus right, and the funding question largely answers itself.
Within your marketing, sales and customer experience effort, we recommend a 70/20/10 split.
70% on what's working and evolving it
The channels, journeys, messages and tools already producing results. The demand engine, refined and kept sharp. This is where most of the team's time should sit, with a constant focus on evolution and improvement.
20% on what's proven elsewhere but new to you
AI enablement, GEO, internal LLMs, agentic readiness, new channels, new segments. Time-boxed pilots with explicit success criteria. A meaningful slice of leadership attention, not a side project.
10% on bets that might not pay off
Customer-facing agents, category-defining IP, partnerships with emerging platforms before they're mainstream. Most won't work. The point isn't the average return. The point is optionality. The bets that do land are the ones that define your next decade.
Most businesses, when they audit honestly, are running closer to 95/5/0. Almost everything going into what worked last year. A small "innovation" line that quietly disappears. Nothing genuinely unproven. And the deeper issue isn't the budget — it's the focus. Leadership attention defaults to the urgent and the familiar, which means the 20 and the 10 never get the airtime they need to mature into anything real.
Culture isn't what you say at the offsite. Culture is what your budget and your calendar reveal. Show me last year's marketing, sales and CX allocation and where your leadership team actually spent its time, and I'll tell you whether your team has permission to experiment. Most allocations say no.
A look to the future
We've spent the last two years talking about AI as a tool. Something you use. A faster way to write a document, summarise a meeting, draft an email.
The next few years will be about agents. AI that takes action on someone's behalf. Researches for them. Compares suppliers for them. Books, negotiates, buys for them.
This is AI agents in practice. Your customers are already starting to outsource decision-making to them. Increasingly, the entity evaluating your business won't be a human. It will be an AI working for a human, applying that person's criteria across a hundred options in seconds.
The big question. Are you building agents yet?
The ones that do, early, will have an enormous advantage over the ones that don't. That's what being a Category of One means in the agentic age. The brand people choose, consistently and confidently, in a world where the people doing the choosing are increasingly not people at all.
If you're a leader thinking about how to navigate this for your own business, we'd love to help.
