Initforthe

🔌 What happens if they change the rules? Ars Technica reported last week that T-Mobile is moving thousands of virtual machines off VMware, while fighting Broadcom over support for perpetual licences. The bit that jumped out at me wasn’t the legal argument - it was the dependency. Over 300,000 CPU cores, more than 1,000 applications, and a migration that sounds painful, expensive and technically awkward. Most businesses aren’t T-Mobile, but they do have the same problem in miniature. A SaaS platform gets sold, pricing changes, a feature disappears, the interface changes, and suddenly the way your team works no longer quite fits. An API that used to be usable becomes expensive, limited, or just awkward enough to break the process around it. And this is the bit people don’t understand: Your data might technically be “yours”, but can you actually get to it? Not just download a rough CSV, not just view it through their screen, not just run the report they’ve decided you’re allowed to run. Can you access it properly, query it, audit it, connect it to other systems, move it somewhere else, and use it in the way your business needs? Because if the answer is no, then you don’t really own your data in any meaningful sense. It’s the same with on-premise systems. People talk about SaaS as though it’s the only risk, but if you’ve got an on-premise system with a hidden database, no proper export, and the vendor has to be paid every time you want access, that isn’t much better. You’re still dependent, it just lives in your building. Dependencies are fine until they stop being fine. Most of the time, the system works well enough, the team knows how to use it, reports come out, invoices get raised, customers get served. Then the vendor changes the rules, and suddenly it’s not just an IT problem, it’s a margin problem, a customer service problem, a staffing problem, and a “how quickly can we get out without breaking the business?” problem. So I’d be asking: 💸 Which platform could put its prices up tomorrow and leave you with no realistic option? 🔐 Where is your business-critical data, and can you actually access it usefully? 🔁 Which processes only work because a vendor currently allows them to work that way? 📣 If a feature changed, would your customers feel it before you had a workaround? 🧩 What would be horrible to replace, even though everyone knows it isn’t quite right? This is why we don’t start by asking “what app do you want?” At Initforthe, we’re usually looking at what the business needs to control: data, workflows, integrations, reporting, customer experience, and the bits where too much power sits with one supplier. Sometimes the right answer is still SaaS, of course it is. There’s no prize for building everything yourself. But there’s a big difference between using a platform and being trapped inside it, and you normally only find out when the vendor changes the deal. #AI #Automation #BusinessImprovement #FutureOfWork

💸 AI still needs a cost control Microsoft has reportedly told thousands of engineers to stop using Claude Code by the end of June and move over to GitHub Copilot CLI instead. The Verge reported that the change affects teams across Microsoft’s Experiences + Devices group, including Windows, Microsoft 365, Outlook, Teams and Surface, with the 30 June deadline lining up with Microsoft’s financial year end. There will be all sorts of reasons behind that. Product alignment. Internal tooling. Security. Control. But the timing is hard to ignore, and Windows Central has also reported that financial considerations appear to be part of the decision. That should make every business leader pause for a moment. If Microsoft needs to rationalise AI usage, what does that mean for everyone else? One of the risks with AI is that it can feel like a cheaper alternative to adding people. Instead of hiring more staff, we give the team agents, assistants and clever tools that can work through tasks at speed. But if they are not governed properly, they are not really that different from throwing people at a problem. Except these ones do not know when to stop eating. They will keep generating, retrying, summarising, rewriting, searching, checking and calling other tools for as long as the workflow allows. Every prompt, every context window, every retry, every badly designed loop has a cost attached to it. That cost might not look like salary. But it is still margin. The question is not whether AI is useful. It clearly is. The question is whether it is being used with the same commercial discipline you would apply to any other part of the business. ❓ Is this AI workflow saving more than it costs? ❓ Could a simpler automation do the job? ❓ Are we using AI to solve the problem, or to avoid understanding it? ❓ Where could usage run away before anyone notices? At Initforthe, this is the gap we help businesses smooth out. Not by telling people to avoid AI, and not by pushing it into every corner because it sounds modern, but by looking at the work itself. 🔍 Where is the friction? ⏳ Where are people wasting time? 🧠 Where does judgement matter? ⚙️ Where would automation be enough? 📈 Where does AI genuinely create commercial value? The best technology does not just make work faster. It makes the business stronger. AI should absolutely be part of that conversation. But profit still needs to be in charge. #AI #Automation #FutureOfWork #BusinessTechnology #Profitability

🧐 A lot of growing businesses are not really scaling anymore. They are adding people to compensate for operational inefficiency. At first, hiring creates genuine capacity. More customers create more work, teams become more specialised, and coordination naturally becomes more complex. But after a while, businesses start accumulating roles that exist largely because work no longer flows properly through the organisation. People spend large parts of their week chasing updates, reconciling spreadsheets, checking statuses, moving information between systems, sitting in meetings to establish what is happening, or manually coordinating work that should already be visible. None of it looks particularly alarming in isolation, which is why it often goes unchallenged for years. The business keeps growing, more people get added around the friction, and eventually the operational overhead becomes part of the normal cost base. On their own, people are the most expensive part of running a business. And then you add the management overhead, communication overhead and dependency risk that comes with every additional layer of operational inefficiency. However, the answer is not fewer people. The right people in the right roles create enormous value. They stay longer, make better decisions and contribute with far more purpose and energy. But they cannot do that properly when the organisation around them creates constant friction. Does it sometimes feel like you are running barefoot on a hot sandy beach, and the sea never gets any closer? That is what badly designed operational systems feel like inside a growing business. Huge amounts of effort, but less progress than there should be. This is why most conversations about AI and automation feel too shallow to me. The real value is not replacing good people. It is building systems that remove unnecessary operational drag so good people can focus on judgement, problem-solving and meaningful work instead of carrying coordination overhead all day. 🧮 We have launched a calculator to help businesses estimate how much operational friction may actually be costing them internally, particularly where hiring has gradually become the default response to inefficiency. Because before adding more headcount, it is probably worth understanding whether the business genuinely needs more capacity, or whether too much of the existing team’s time is disappearing into work the organisation should have designed out years ago. 👇 Link to the calculator in the comments #AI #Automation #OperationalExcellence #BusinessGrowth #FutureOfWork

🤖 AI is not a plaster for poor process We are starting to see a strange pattern in business technology. A company has a messy customer journey, patchy documentation, unclear internal ownership, and a website that does not answer the obvious questions. Then someone suggests adding an AI chatbot. On paper, it sounds modern. In practice, it can become an expensive way to avoid fixing the basics. If customers are asking the same ten questions every week, do you need a conversational AI system burning through tokens to answer them? Or do you need a clear FAQ, better website content, cleaner onboarding emails, and a process that makes sure the right information is easy to find? That is not anti-AI. Far from it. AI can be incredibly powerful when it is used in the right place. It can help summarise messy information, support internal decision-making, speed up admin, triage support requests, and make people’s working lives easier. But AI should not be the first layer you add to a broken process. It should be the layer you add once you understand where the real friction sits. Otherwise, you are not automating intelligently. You are just making the mess more expensive. At Initforthe, this is the sort of gap we help businesses work through. Not by shoving AI into every corner because it sounds impressive, but by looking at how the business actually works today and asking better questions. ❓ Where are people wasting time? ❓ Where are customers getting stuck? ❓ Where is information being repeated, lost or misunderstood? ❓ What could be solved with a better page, a cleaner system, a smarter form or a simple automation before AI even needs to get involved? The best technology decisions are rarely about chasing the newest tool. They are about building the shortest, clearest path between the business you have today and the one you want tomorrow. AI belongs in that conversation. It just should not be allowed to dominate it. #AI #Automation #FutureOfWork #BusinessTechnology #DigitalTransformation

🔎 This week in AI Amazon has started tracking how thousands of its engineering teams are using AI. Not just who has access, but how often they use it, where it fits into their work, and whether it makes any real difference to output. That feels like an important shift. Most businesses have already given people AI tools. Some teams are using them well. Some are experimenting. Some are probably using them in ways leadership does not fully understand yet. The problem now is not access. It is integration. I see this playing out in a lot of businesses. One team uses AI to summarise calls. Another uses it to draft documents. Someone else builds a small automation for reporting. Then an agent gets added to a workflow with a broad instruction to “handle follow-up”, and suddenly nobody is quite sure where the work starts, where it ends, or who is responsible when something looks wrong. It looks busy. It does not necessarily make the business better. More tools create more overlap. More overlap creates more places for work to drift between people, systems and assumptions. The smarter move is to start with how work actually flows, then decide where AI can strengthen that flow. Sometimes that means a simple embedded automation that follows up, checks a status, or updates the right record at the right time. Sometimes it means a narrowly scoped agent with a clear goal, clear boundaries, and an obvious point where a human steps in. That distinction matters. AI agents are not senior people with years of context and judgement. They are much closer to junior team members who can move quickly and handle volume, but need structure around them. Most businesses understand this instinctively with people. They do not give a new starter a vague remit, access to everything, and no supervision. Yet that is exactly how some AI is being introduced. At Initforthe, this is where we focus our work. We help organisations look honestly at how work moves today, where handovers are weak, where data gets messy, and where automation can reduce friction rather than create another layer of confusion. That is how you smooth the gap between where the business is now and where it wants to be next. Because the companies that get this right will not be the ones using the most AI. They will be the ones moving faster, taking market share, and making everyone else wonder why they suddenly feel so slow. Where in your business is AI creating activity, but not yet creating progress? #AI #Automation #BusinessOperations #FutureOfWork #Leadership

⚙️ Week 11: Construction Projects With Visible Coordination Construction is one of the clearest examples of how businesses end up hiring more people just to keep things moving, and not because the work itself demands it. It’s the coordination around the work that starts to take over. Walk onto most sites and you’ll see it straight away. Materials haven’t arrived when they should have, teams are working from drawings that have already changed, and someone is on the phone trying to piece together what’s actually happening across suppliers, site managers and trades. None of it is unusual, and most teams deal with it well, but it takes effort to keep everything aligned. As projects increase, that effort grows with them. More sites mean more people chasing updates, more time spent checking what’s right, and more layers added just to stay on top of things. Work still gets done, but each additional project carries a bit more weight than the last, and over time that starts to show up in cost, delays and rework. The organisations that handle this better tend to look at it slightly differently. Instead of treating coordination as something people need to stay on top of, they design it into the system so the information moves on its own. Plans, materials, schedules and progress are connected, so when something changes it is visible everywhere it needs to be. If a delivery slips, you can see the knock-on effect early. If something on site moves ahead or falls behind, it is reflected without someone needing to chase it down. People are still making decisions, but they are not spending their time trying to reconstruct what is going on. That is where technology starts to earn its place. Not in replacing the judgement of the people doing the work, but in removing the constant need to check, confirm and reconcile. The system carries more of that load, which frees people up to focus on the parts that actually need experience and judgement. Over time, the difference becomes quite noticeable. The same team can oversee more projects without the same increase in coordination effort, simply because they are not having to hold everything together manually. Fewer surprises appear late, and when things do move, they are easier to deal with because the impact is already visible. Most organisations respond to complexity by adding more process and more people around it. It feels necessary in the moment, but it also makes the system heavier. When coordination depends on people stepping in all the time, growth tends to come with drag. When the system takes on more of that work, things start to feel a bit different. Projects move with less friction, people spend more time making decisions rather than chasing information, and growth does not immediately translate into more overhead. #Construction #Automation #AI #OperationalExcellence #ProjectManagement #FutureOfWork

⚙️ Week 9: Financial Operations That Carry Their Own Compliance Most finance teams don’t struggle because the work is inherently complex. They struggle because the system can’t be trusted to get it right on its own. You see it at the same point every month. Numbers don’t quite reconcile. Someone checks them. Then someone else checks them again. Not because the process is unclear, but because no one is entirely comfortable relying on the system without a second look. So more checks get added. More approvals. More controls layered on top. It looks thorough. And in many cases, it is. But it also creates a different problem: the work starts to depend on people verifying it at every step. As volume grows, so does the effort required to stay confident that everything is correct. More transactions mean more reconciliation. More regulation means more oversight. And before long, growth starts to feel like a hiring plan. That is where things begin to get expensive. High-performance organisations approach this differently. They do not start with the question of how to check the work. They start with how to structure it so it does not need checking in the first place. Transactions are created correctly at source. Rules are applied as the work happens. Exceptions are surfaced immediately, not discovered weeks later when someone has time to review them. Audit trails exist because the system produces them, not because someone has to assemble them under pressure. You stop relying on people to enforce discipline. The system takes on that role. This is where AI starts to become genuinely useful. Not in replacing finance teams, but in removing the need for constant verification. The repeated checks, the manual controls, the reconciliation work that exists because the system cannot be relied on to carry the load. When that layer reduces, the work changes. Finance teams spend less time proving that the numbers are correct and more time understanding what they mean. Conversations move away from whether something is right and towards what to do next. That is where the value starts to show up. And commercially, it makes a difference in ways that are easy to miss at first. You are no longer adding people just to keep up with compliance. The same team can support more growth, more complexity and more scrutiny without needing to expand at the same rate. Costs stop rising in line with activity. Margin stops getting squeezed every time volume increases. That is where the leverage comes from. A lot of organisations invest heavily in controls, but very little in the design of the system those controls sit within. They optimise tasks rather than stepping back and asking whether the system should need those tasks at all. But if your operation only works because people are constantly checking that everything is correct, then it is not really under control. It just hasn’t broken yet. #Finance #Automation #AI #OperationalExcellence #FutureOfWork

Hospitality does not break under pressure. It breaks under unpredictability. You’ve seen it play out. A table is not ready, guests are waiting, the kitchen is backed up, and someone is deciding whether to seat, delay or apologise. The plan made sense earlier in the day, but now it is being rebuilt in real time. What follows is familiar. Staff step in, managers intervene, and the operation is held together through experience rather than anything systematic. And yet many operators look at this and say, “it works”. Guests are served, tables turn, and the day gets through. The problem is that it only works because people are constantly compensating. They rebalance tables, chase updates, and smooth over issues before they escalate. On the surface it looks stable, but underneath it depends on continuous intervention to stay that way. That effort does not show up clearly on a P&L. It shows up in overstaffing, inconsistent service and lost capacity. Over time, that is where margin starts to disappear. High-performance hospitality organisations take a different view. They ask whether the system can keep up with reality without constant intervention. That changes how the operation is designed. Bookings, walk-ins, pacing and kitchen throughput are treated as live inputs. When a table overruns, the next booking adjusts. When demand builds, staffing shifts. When service slows, someone knows early enough to act. The operation responds as things happen, rather than waiting to be corrected. This is where AI starts to become useful. Not as a layer on top, but as part of how the system works. Rules are applied consistently, patterns are surfaced, and repeated decisions no longer depend on someone remembering to make them. As a result, you are no longer scaling headcount to manage variability. The same team can handle more volume and unpredictability without a corresponding increase in effort. That is where the leverage comes from. You start to see it in how the operation runs. Service becomes more consistent, space and staff are used more effectively, and performance no longer depends on who happens to be on shift. Hospitality should feel effortless to the guest. Most of the time, that effortlessness is created by people working hard behind the scenes. But if your operation only works because people are constantly stepping in to fix it, then it is not a system under pressure. It is a system that has never really been designed to cope. #Hospitality #Automation #AI #OperationalExcellence #CustomerExperience #FutureOfWork

🔎 This week in AI There is an interesting shift happening in the background of the AI conversation this week. Most of the headlines still focus on models, features and capabilities. But if you look at how businesses are actually talking about AI, the emphasis is starting to move. It is less about what AI can do in isolation, and more about how it fits into the day to day running of an organisation. We are seeing more discussion around AI agents that act on behalf of users rather than waiting for instructions. At the same time, companies are restructuring and investing heavily in automation, which suggests this is no longer about experimentation. It is about how work gets done going forward. That changes the role AI plays inside a business in a way most organisations are not set up for. Up to now, most organisations have treated AI as something you use when you need it. A tool you open, a prompt you write, an answer you review. Useful, but still separate from the flow of work. What is starting to happen now is different, and it is easy to underestimate. AI is beginning to sit inside the workflow itself. Following things up, updating systems, keeping processes moving without someone having to remember to do it. The part that gets missed in all of this is simple. Most businesses still rely on people to hold everything together. Work progresses because someone chases it, checks it, or nudges it along. I see this in almost every organisation we work with. When you introduce automation into that environment, you are not just making things faster. You are exposing how the system really works. If ownership is unclear, if data is inconsistent, if processes stop at team boundaries, automation does not fix that. It simply makes the gaps more visible, and sometimes more costly. This is why so many AI initiatives feel underwhelming. The technology is capable, but the operating model around it has not changed. At Initforthe, this is the space we focus on. Not adding more tools, but helping organisations understand how work actually flows across the business and what needs to change for automation to land properly. That means looking at where coordination breaks down, where decisions really happen, and how to design processes that can run reliably without constant human intervention. The interesting question this week is not how advanced AI has become. It is whether businesses are ready for work to move forward without being pushed at every step. The organisations that get this right will not necessarily look more advanced. They will feel more stable. Less chasing, fewer gaps, and more of the right things happening when they should. That is where the real value sits. #AI #Automation #BusinessOperations #FutureOfWork #Leadership

⚙️ Week 6: Field Operations With Intelligent Dispatch Most field operations don’t struggle because people are underperforming. They struggle because the system cannot keep up with reality. Schedules are built in advance, routes planned on assumptions, and then the day begins. Traffic shifts. Jobs overrun. New requests come in. Priorities change. What looked solid at 8am starts to break by mid-morning. So people step in. Dispatchers reshuffle jobs. Teams call in. Managers jump between systems trying to understand what is actually happening. By the end of the day, success depends less on the plan and more on how hard everyone worked to recover it. It feels like control. It is actually constant correction. This is where organisations lose leverage. Every disruption creates work. More jobs mean more exceptions. More exceptions mean more coordination. And more coordination means more people just to keep things moving. Growth starts to demand headcount, not because the work requires it, but because the system cannot adapt. High performance organisations design this differently. Scheduling is not a one-off activity. It responds continuously. Routes adjust to live conditions. Job durations reflect real data. New work is prioritised and assigned dynamically. When something changes, the system absorbs it. A delay updates downstream schedules. A cancellation creates usable capacity. A high-priority job is inserted without disrupting everything else. People are not chasing the plan. They are working within a system that reflects reality. 🤖 Automation and AI remove the coordination burden. The system monitors, adjusts and optimises in real time. Teams focus on delivery, not recovery. 📈 The commercial impact is simple. When people manage disruption, cost grows with complexity. When systems manage disruption, complexity becomes cheaper to handle. You can increase jobs without increasing coordination effort at the same rate. Utilisation improves. Missed appointments drop. Margin holds. That is the difference between scaling activity and scaling capability. If your day depends on people fixing the schedule, you do not have an advantage. You have a coordination problem. #FieldService #Automation #AI #OperationalExcellence #FutureOfWork