# Process is how a company progresses > Why process is how a company moves forward. One shared picture of the work helps with succession, AI adoption and exit, with the evidence and its limits. URL: https://inocta.io/lab/field-notes/process-is-how-a-company-progresses/ Our Lab ### How we make the company progress with humans and machines aligned Process is not a binder. It is one shared picture of the work that helps people pass it on, AI take hold and a buyer see a company that runs. Read the note By Mo Kahlain Co-author Nadia Inocta · AI agent Insight 11 Nov 2026 6 min read #### In short - A process is one picture of the work that all of us share, so the company can move forward together. - With that picture, people can pass the work on and AI is easier to put in place. A buyer sees a company that runs without us. - A process that helps has an owner, is used by the team and changes when we learn. Discuss it with your agent ChatGPT Claude Opens your assistant with this note and its link already in the prompt. Many of us see process as a burden. We think of a binder to write or a BPMN diagram to produce, and then it sits on a shelf. That is not what a process is for. A process is how we make the company progress. The word itself comes from a Latin idea of moving forward [6]. To progress together, we first need to agree on one picture of how we do the work. Without it, each of us works from a private version, and nothing moves in the same direction. #### Draw the picture with the people doing the work There is a scene in The Founder that makes this visible. The McDonald brothers mark out a kitchen on a tennis court. Staff rehearse the work, moving through the proposed layout. The brothers watch, rearrange it and try again before building it [1]. What matters in that scene is that the people who will do the work test the plan. A drawing becomes something they can challenge through use, before anyone commits to a layout that will shape their day. This is where BPMN earns its place. It is a standard way to draw the picture so that everyone reads it the same way. The drawing is not the goal. The goal is the conversation it starts when we walk through an order, a customer request or a new hire's first week. Where does work wait? Where does information disappear? Where does someone have to guess? #### A company that someone else can run The same picture helps when someone is away, when the team grows and when the company changes hands. If every exception still returns to the founder, we have not transferred responsibility. Bradley Lay buys companies. He describes reviewing a construction company where one operations manager held much of the practical know-how [7]. For an owner who plans to exit, this matters a great deal. The lesson for us is to protect and share that expertise, and never to treat the person as the problem. BDC's guidance on selling a business says the same thing in its own words. It recommends repeatable, teachable processes and less dependence on the departing owner [2]. Profitability, customers and capable people stay central. A shared method makes the handover easier to prepare. Franchising shows how a method travels. The McDonald's agreement with Arcos Dorados combines current operating manuals with training requirements and continuing obligations [3]. The manual does not run a restaurant. People learn the method and practise it. Brand, resources and operator experience matter too, so we cannot say documentation alone explains a network's success. Research on how practices spread points the same way. At Rank Xerox, a field study covered 15 western European countries. It found that using a template helped a practice get adopted and perform at the receiving unit [8]. A study followed 36 pizza stores franchised from one corporation. What a store learned reached other stores under the same owner, and not stores under a different owner [9]. A method spreads when we share it on purpose. #### Give AI work we understand "You can't automate what you don't understand." For us, that means knowing the outcome, the decisions and the limits before we delegate the work. A 2026 working paper followed 515 high-growth startups. All of them had AI access and training. The firms that were also shown examples of reorganizing work around AI found 44% more uses and completed 12% more tasks [4]. It is a working paper, and it does not test documentation alone. We read it as a hint, not as a rule. When the whole workflow sits on one picture, we can ask which steps should disappear, which should change and which a machine can safely do. Someone stays accountable for the outcome. We define what good work looks like, what a machine may decide and when it must stop for human judgment. We test real cases, exceptions included, before we give it more authority. #### Make the method live EOS names the adoption challenge Followed By All [5]. Writing down a core process is half of the work. People must learn it, use it and keep it current. In a randomized trial, textile plants in India received free consulting on management practices. Adopting them raised productivity by 17% in the first year, and within three years the firms opened more plants [10]. That trial tested a package of practices, not documentation alone. Our extension is simple. When people and AI work together, they work from the same current picture of the purpose, the rules and the expected result. Human adoption comes first. A process does not need AI to help the company. We can start with three to five processes that matter most. Each one gets an owner who answers for the outcome. We walk the steps with the team and make the handoffs and exceptions clear, including who takes over when someone is away. Then we use it. We look at rework, delays and mistakes, and we revise the method when the evidence changes. What we learn stays within reach of the next person who needs it. Process is how a company moves forward, with people and machines pulling in the same direction. With AI beside us, that progress does not have to stop. It only holds if the picture of our work stays alive, as a drawing people can read and as a file machines can read. Keeping it alive is our responsibility. From now on, our role is to direct the progress. #### Sources - The Founder production design interview. The scene is a film illustration, not independent historical or business-performance evidence. - BDC on preparing a business for sale. Practitioner guidance, not a measured valuation premium. - McDonald's and Arcos Dorados agreement. April 2025, sections 10.2 and 12.2. Contractual requirements are not causal evidence of franchise survival. - Mapping AI into Production, Kim, Kim and Koning, 2026 working paper. See also the HBS institutional account and HBR on mapping workflows. Randomized startup experiment, not a documentation-only treatment. The primary abstract and institutional account were checked, not the full methods and tables. The results should not be generalized to all businesses. - EOS Process Component. An external framework we cite, not a statement of affiliation or certification. - Etymology of process. Process and progress have different Latin histories. Forward movement is an editorial connection, not proof that every process improves results. - Bradley Lay's LinkedIn post, 25 March 2025. An account of due diligence by someone who buys companies, not independent evidence or a completed acquisition. The post's £5 million refers to annual revenue, not a purchase price. Its broader numerical claims are not used here. - Template use and the effectiveness of knowledge transfer, Jensen and Szulanski, Management Science, 2007. Eight-year field study at Rank Xerox. The abstract was checked, not the full text. It is not a restaurant study and not a documentation-only test. - The acquisition, transfer, and depreciation of knowledge in service organizations, Darr, Argote and Epple, Management Science, 1995. Weekly data from 36 pizza stores in southwestern Pennsylvania, franchised from one corporation and owned by 10 franchisees. The abstract was checked, not the full text. - Does management matter? Evidence from India, Bloom, Eifert, Mahajan, McKenzie and Roberts, Quarterly Journal of Economics, 2013. Randomized field experiment on large textile plants. The abstract was checked, not the full text. A package of management practices, in one industry and one country. Try IO Process, the free tool Get the next field note by email. Subscribe to The Signal #### Keep reading All of the Lab Field note ##### I Built a Team of AI Agents. Then I Became the Bottleneck. Fourteen AI agents work in our operations, and the consequential steps still wait for one person. What HBR and new research say, and what we are changing. Mo Kahlain 19 Nov 2026 5 min Read the note AI this week ##### Anthropic heads to Nasdaq, more AI output to review, and a kill switch for AI agents Anthropic updated its Nasdaq filing, a Miro survey finds more AI output to review, and CrowdStrike lets an administrator stop a running AI agent. inocta.io 6 Nov 2026 1 min Watch the episode Field note ##### Build AI skills that can handle a changing business An AI skill should carry expertise an agent can adapt to a new request, inside company rules. Here is what a CEO should decide before wider use. Mo Kahlain 3 Nov 2026 4 min Read the note