Team Studio · 3 people.

1. Start with a customer, not a storefront
2. Shape an offer and understand the numbers
3. Build a product page that answers questions
4. Run a small launch and learn from it

An online store is a way to present an offer, accept an order and help someone receive what they expected. It is not evidence that anyone needs the product. Before choosing a name or arranging a homepage, describe one customer in one situation. A useful starting sentence is: When this situation happens, this person wants to complete this task, but this obstacle gets in the way. The situation makes your idea specific enough to investigate. Separate the problem from your proposed solution. Someone who struggles to carry lunch and work equipment might want a better bag, a smaller lunch box or a different routine. If every interview question mentions your preferred product, you may miss the real need. Ask about the last time the problem occurred, what the person did, what they tried before and what made the alternative inconvenient. Past behaviour gives you more concrete material than a polite prediction about future purchases. Fictional worked example: Maya is exploring Desktrail, an imagined shop for compact work accessories. Her first idea is a cable pouch for people moving between home and shared desks. Her problem statement says that these workers want to set up quickly, but loose chargers and adapters are difficult to find. In an invented practice interview, a worker describes using an old pencil case. Maya records that description as an observation. She separately records her hypothesis that internal pockets might help. The interview does not establish that the worker would pay for a new pouch. Maya prepares a learning sheet with four columns: observation, interpretation, unanswered question and next check. One observation can support several interpretations. A tangled cable could indicate poor storage, too many cables or an awkward bag. She chooses the cheapest useful check: showing a rough pouch layout and asking the person to arrange their actual items. She does not order stock on the strength of a compliment. Common mistake: describing a customer only as a broad demographic or treating enthusiasm as confirmed demand. A specific situation is more useful than an impressive audience size. A few conversations can reveal questions worth testing, but they do not represent everyone. Useful next step: write one problem statement and arrange a small set of voluntary conversations with relevant people. Explain the purpose, avoid collecting unnecessary personal details and keep exact observations separate from your conclusions. Finish by naming the assumption that could most change your decision to continue.

1. Choose the right task for an assistant
2. Give context, constraints and an output shape
3. Summarise without losing important details
4. Build a repeatable review workflow

Your goal is to choose work where assistance is useful and mistakes can be caught. ChatGPT can help organise supplied notes, develop a first draft or offer alternative explanations. A fluent response is still something to review. It can omit a condition, invent a detail or misunderstand your purpose. Start with a small task whose quality you can judge yourself, rather than asking an unfamiliar tool to make a consequential decision. Use three questions to assess a task. Is the input suitable to share? What would an acceptable result contain? Can I verify that result before anybody relies on it? If one answer is unclear, narrow the task or do it manually. Editing the wording of a fictional meeting agenda is a useful practice exercise. Deciding whether a real customer has a legal entitlement is not an appropriate beginner shortcut. Separate preparation, transformation and approval. Preparation means choosing the facts and removing information you should not provide. Transformation is the requested activity, such as grouping agenda items by topic. Approval means checking the result against the facts and deciding whether to use it. The assistant does not take responsibility for the final message simply because it drafted the wording. Fictional worked example: Maya coordinates a community workshop. Her notes say: 12 seats, beginner level, bring a notebook, room opens at 17:45, session starts at 18:00. She wants a short reminder. This is manageable because the source is small and the output can be checked line by line. She writes down five acceptance criteria before prompting: correct capacity, correct audience, correct item to bring, both times preserved, and no new promises. If a draft invents free parking, she removes it and checks whether other unsupported details appeared. Keep practice data fictional. For real work, follow your organisation's approved tools and data-handling rules; removing a person's name alone may not remove identifying details. Do not assume that a particular plan or setting makes every upload appropriate. Features and availability can change, so this course teaches a review process rather than a specific subscription or button sequence. A common mistake is choosing an enormous task because generation looks fast. A long document can take longer to verify than to write. Your next step is a task card: source material, desired transformation, acceptance criteria and reviewer. Keep the first experiment short enough to compare with the original in a few minutes.

1. Define one audience and one objective
2. Write a message supported by evidence
3. Choose channels with a small experiment
4. Read results without jumping to conclusions

A small campaign becomes easier to plan when it answers two questions: who needs this information, and what observable action matters now? An audience is not merely a list of ages or interests. Describe a situation, a task and an obstacle. Someone may fit your audience but have no need to act today. Your message should help that person judge relevance, not assume that attention means purchase intent. Separate the offer from the objective. The offer is what someone receives, including its scope, price and limitations. The objective is the action you want to observe within a defined period. Awareness, enquiries and purchases are different objectives. Pick one primary objective and keep other measurements as supporting context. A planning target gives you a reference for review; it does not establish what a campaign will achieve. Fictional worked example: Clear Week is an imagined local workshop for people planning their first week of freelance work. Its audience brief says: people with several small client tasks who need a simple way to see competing commitments. The workshop offers a guided planning exercise and a blank weekly worksheet. It does not promise more income or fewer working hours. The campaign objective is to receive eight submitted information requests during a seven-day trial. Eight is a chosen planning target, not a forecast or evidence of demand. Make the counting rule explicit before publishing. For this example, an information request means one completed enquiry form submission. Test entries are excluded; repeat submissions from the same person are noted separately where the available records allow that distinction. A page visit is not an enquiry, and an enquiry is not a booking. Record the observation window and any missing data. Without an earlier comparable campaign, label the baseline unknown instead of inventing an improvement percentage. Write down what you know about the audience and what you still assume. A conversation describing an existing planning problem is evidence of that person's experience. It is not proof that everyone in the audience wants this workshop. Keep personal details out of the working brief when they are unnecessary. Use questions about current behaviour to refine the brief. Common mistake: choosing followers as the main measure when the actual goal is information requests. A larger follower count cannot answer whether people asked about this offer. Next step: complete a one-page audience and objective brief, including a counting rule, time period and one assumption you need to investigate.

1. Give every piece a useful job
2. Structure a clear opening, middle and ending
3. Create with accessibility in mind
4. Publish, reuse and improve responsibly

A useful piece of content answers a recognisable question for a particular person. Starting with a format, such as a post or video, can hide that purpose. Start instead with the audience's situation: what are they trying to do, what do they already know and where do they get stuck? A narrow answer can be more useful than a broad introduction that leaves the reader unsure what to try. Choose one takeaway that fits in a sentence. This is the idea the audience should be able to explain or apply afterwards. Separate it from your publishing goal. Publishing every week describes your activity; helping a beginner identify the next item they need describes a useful audience outcome. You can plan for that outcome without claiming everyone will achieve it or that the content has already helped real people. Fictional worked example: the imagined project Small Desk Notes creates a short guide for someone returning to a crowded home desk. The question is, How can I prepare this space for my next task? Its takeaway is, Keep the materials for one chosen task within reach and give unrelated items a temporary place. The guide is not a promise of better health, higher productivity or a permanently tidy home. It is a modest organising procedure the reader can adapt. Build a six-field brief: audience, question, takeaway, evidence, format and next step. For Small Desk Notes, the format is a short illustrated explanation, and the next step is to arrange one task's materials. Evidence consists of an original staged example showing the sorting process. That example can demonstrate the instructions, but it cannot prove broader performance benefits. Label observations from your own practice as observations, not universal facts. Collect only what the piece needs. Write down which statements require checking, which images you will make yourself and which outside materials would need permission or suitable usage terms. Avoid including private papers in a desk photograph. If another person appears or supplies an example, agree how their contribution may be used before publication. Keep unresolved questions visible rather than filling them with confident guesses. Common mistake: expanding a simple desk guide into advice about scheduling, storage purchases and motivation before teaching its one procedure. The extra subjects compete with the promised answer. Next step: complete the brief and write a short boundary sentence stating what the piece will not cover. Use the brief to judge every planned section: if it does not help answer the chosen question, save it for another piece.

1. Write the promise before choosing colours
2. Build a small, usable visual system
3. Find a voice that makes information clearer
4. Keep the brand consistent across the journey

A brand is the expectation someone builds from repeated experiences with a business. Its name and appearance help people recognise it, but the product, explanation and response to problems determine what that recognition means. Start by deciding what customers should reliably expect. A promise should guide real choices, not simply provide an attractive line for the homepage. Describe the audience by a situation and a need. Then write the useful difference you intend to provide. A working formula is: For this audience in this situation, we offer this product with this specific focus. The difference can be modest. Clear sizing information, a tightly selected range or easy-to-understand instructions can provide direction without claiming the business is the best at everything. Support the difference with evidence or a concrete operational commitment. Fictional worked example: Paperpath is an imagined shop selling paper planning pads for people coordinating small community projects. Its draft promise is, Simple planning pages that keep shared tasks visible. The promise does not say that buying a pad will remove stress or make every project successful. Instead, the fictional product concept includes space for a task, its owner and its next review date. These design choices connect the promise to something a buyer could inspect. The founder, Nia, chooses three attributes: clear, practical and considerate. She translates each into behaviour. Clear means showing a filled-in sample alongside an empty page. Practical means explaining what belongs in each field and what can be left blank. Considerate means avoiding language that shames a person for unfinished tasks. If a proposed advertisement says never miss a deadline again, it fails the promise test because the pad cannot control the team's actions. Test a promise against the whole experience. Can the product support it? Can the description show it? Can support behave consistently when something goes wrong? A useful brand choice also excludes something. Paperpath might avoid a huge catalogue of unrelated decorative goods because maintaining a focused planning range better supports its present purpose. This is a working choice, not a rule that the business can never change. Common mistake: selecting generic adjectives such as premium, innovative and trusted without specifying what anyone would observe. Another is copying a competitor's promise while lacking the same product or capabilities. Useful next step: write one promise, three attributes and one observable behaviour for each attribute. Add one claim you will deliberately avoid. Review the set against a product sample and one difficult support situation before considering names, logos or colours.

1. Capture commitments before organising them
2. Choose priorities with explicit trade-offs
3. Plan with capacity, not wishful thinking
4. Review the system and make one improvement

A useful planning system starts with a trustworthy inventory, not an elaborate app. Before deciding what matters most, collect the work that already exists. Look at notes, recent messages, your calendar and promises you remember making. The purpose is to see commitments clearly enough to make decisions, not to produce the longest possible list. Separate four kinds of item. An action is something you can do, such as drafting an opening paragraph. A project is an outcome requiring several actions, such as publishing a portfolio. A waiting item depends on someone or something else. An idea is a possibility you have not committed to. Mixing these together makes a list feel larger and more urgent than the actual work. Rewrite each actionable item with a verb, an object and a stopping point. 'Portfolio' does not tell you where to begin. 'Choose three completed projects for the portfolio draft' does. You do not need to define every future step immediately. Identify the next visible action and note what it depends on. A task should help you start working instead of making you plan the same task again. Fictional worked example: Sam is preparing a small design portfolio while handling routine work. The captured notes are 'portfolio', 'ask Jo for photos', 'reply to workshop organiser' and 'learn animation someday'. Sam treats the portfolio as a project. The next action is to choose three projects. Requesting photos is an action until the request is sent, then becomes a waiting item with a follow-up date. Replying to the organiser is a separate action. Animation goes on an optional ideas list, not tomorrow's schedule. Keep enough context to act without collecting unnecessary information. A waiting item might contain a topic and a follow-up date, not a copied private conversation. Store work in an appropriate place and respect boundaries between personal notes and shared team systems. A public task list should not become an accidental store of customer information. A common mistake is organising every item before checking whether it still matters. Delete obsolete personal reminders, clarify unclear commitments and ask the relevant person before changing a shared obligation. Capture first, then process a manageable number of items. Perfect labels do not compensate for work that has no defined next action. Your next step is one small inbox and a regular time to review it. If capturing an item takes several minutes, simplify. The system should be easy enough to use on a busy day and clear enough to trust when you return after a break.

1. Match one useful page to one search need
2. Make important pages discoverable and indexable
3. Write page signals that help people choose
4. Measure outcomes and improve the page responsibly

Search engine optimisation begins with a person and a useful page, not a trick for gaining rankings. Choose a specific question that your organisation can answer from real knowledge. Then decide what the searcher is trying to do: learn, compare, find a place, or complete an action. Similar phrases can hide different needs. Someone searching for “balcony herb care” may want instructions, while “balcony herb starter kit” suggests a product comparison. Write a page purpose before choosing wording. A useful purpose states the audience, situation, answer and next step. It prevents one page from becoming a collection of unrelated phrases. Google’s Search Essentials recommends helpful, reliable, people-first content and using the words people use in prominent places. That is not an instruction to repeat a phrase unnaturally. It is a reminder to describe the subject plainly. Fictional worked example: Northstar Garden Club wants to publish a guide for apartment residents. Its broad idea is “herbs”. The team reviews member questions and chooses “How can a renter keep three kitchen herbs alive on a shaded balcony?” The page will explain light checks, watering observations and movable containers. A separate shop page may later describe a starter kit. Keeping the guide and offer distinct lets each page do one job. Northstar is fictional, and the example is not evidence that a page will rank or produce sales. Create a small evidence file. Record the questions you have actually heard, relevant terminology, sources for factual claims and gaps you cannot answer. Search results can show formats and related language, but copying a competitor’s outline does not make your page useful. Bring original experience, a worked example, a clear explanation or data you are allowed to publish. Choose a primary phrase only as a writing aid. Add natural variants where they improve clarity. The page title and main heading should make the subject recognisable, while the introduction should confirm that the reader has arrived in the right place. Avoid exaggerated promises such as “guaranteed first place”. Search systems change, competitors publish and results differ by context. The practical outcome is a one-page content brief. It should name the reader’s question, the page’s answer, supporting evidence, intended action and boundary. If you cannot explain why this page deserves to exist, research the need before drafting more copy.

1. Build an audience around clear permission
2. Write one email with one useful next step
3. Run a careful pre-send and delivery process
4. Read campaign evidence without chasing one metric

Email marketing starts with a reason for a person to hear from you. A list of addresses is not automatically a marketing audience. Record how each contact joined, what they expected, when that happened and how they can change their choice. Platform labels differ, but the underlying distinction matters: subscribed contacts have chosen to receive marketing; customers or imported contacts may not have done so. Write a simple audience promise. State who the messages are for, what kind of information you will send and a realistic frequency. A narrow promise is easier to honour than “updates about everything”. Use a signup form that explains the value without pre-checked ambiguity. Keep confirmation and preference records in the tool appropriate to your organisation. Applicable requirements depend on location and context, so consult current official guidance or qualified advice rather than treating this introductory course as legal advice. Fictional worked example: Fieldnote Kitchen is an invented community cookery project. Its form offers “one practical seasonal recipe and one class update each month”. It asks only for an email address and an optional cooking interest. A previous event spreadsheet contains attendees, but the team does not silently add every address to marketing. Instead, it sends any permitted event communication for its original purpose and provides a separate way to choose the newsletter. Design useful preferences. Someone may want recipes but not class announcements, or fewer messages rather than none. Preference options can reduce irrelevant email, but there must still be a clear way to stop marketing entirely. Suppress unsubscribed addresses from future campaigns and avoid re-importing them as new subscribers. Never purchase a list or invent consent. Collect less data. A birthday, postcode or job title should have a specific use and an appropriate basis before it appears on a form. More fields create more responsibility and can make signup harder. Protect exports, restrict access and document deletion or retention decisions. External email services and accounts are not included with this course. Your first deliverable is an audience register with source, promise, status, preference and last-confirmed date. It is a working control, not permission to send. Before any campaign, check the rules and provider policies that apply to the intended recipients. Trust grows when the content matches the expectation that created the relationship.

1. Connect the campaign objective to a real outcome
2. Build truthful creative around one message
3. Set a bounded audience, budget and launch plan
4. Review results, uncertainty and the next experiment

A Meta advertising campaign begins with a business outcome, not a favourite button in Ads Manager. The objective tells the delivery system what kind of result to seek. Current objective families include awareness, traffic, engagement, leads, app promotion and sales, but available options can vary by account and change over time. Choose the closest observable outcome you can support rather than the metric that looks cheapest. Write the path before opening the platform: audience situation, message, destination and final action. A campaign seeking workshop enquiries needs a usable enquiry process and a way to record qualified responses. Sending traffic to a page that cannot answer basic questions creates clicks without a complete journey. Define one primary result and a few diagnostic measures. Fictional worked example: Alder Street Pottery is an invented studio offering a beginner taster evening. Its outcome is completed enquiry forms from adults who can attend locally. The campaign uses a lead-oriented objective and sends people to a page with date, location, accessibility information, price and form. Link clicks and landing-page views help diagnose the path, but neither is counted as an enquiry. This example is not a recommendation for a particular account or a forecast of bookings. Describe the measurement boundary. If the site records a form submission, test the event and document what it means. A submitted form may still be spam, duplicate or unsuitable, so separate platform-reported results from reviewed enquiries. Attribution settings describe how a platform assigns credit; they do not prove that an ad was the only cause. Preserve a simple source record and compare it with the organisation’s own outcomes. Check feasibility before spending. Does the destination load and work on a phone? Can someone respond to enquiries? Is the chosen event frequent enough to observe during the test? If not, narrow the learning question or use a nearer measurable step without pretending it is the final outcome. Set a spending limit you can afford to lose. No performance or revenue is guaranteed. This independent course is not affiliated with or endorsed by Meta. A Meta account, advertising access, tracking setup and ad spend are not included. Interfaces and policies change, so confirm current options in official Meta documentation before launching.

1. Map the workflow before adding AI
2. Design structured inputs, outputs and tool boundaries
3. Add human approval, privacy and safe failure
4. Evaluate, monitor and improve the whole workflow

Automation should begin with a process that people can explain. Pick a repeated, bounded task and map its trigger, inputs, rules, outputs, owner and exceptions. Then decide which parts are deterministic and which require interpretation. Ordinary code is better for exact totals, identifiers and fixed business rules. An AI model may help classify or draft uncertain language, but its output remains something the surrounding system must validate. Draw the happy path and at least three failure paths. Ask what happens when an input is missing, duplicated, late or malicious. A useful automation does not merely produce an output; it leaves the organisation able to understand what happened and recover. Choose a low-consequence first workflow whose result can be reviewed before it affects a customer or irreversible record. Fictional worked example: Harbour Desk, an invented co-working space, receives general enquiry emails. The proposed workflow captures a permitted message, removes unnecessary personal details, asks an AI step to suggest one of four routing labels and drafts a short summary. Deterministic code checks that the label belongs to the allowed set. A staff member reviews the original message, label and summary before routing. Billing disputes and safety concerns always go to a person without automated action. Define success in operational terms. Harbour Desk wants to reduce manual sorting time while keeping misroutes visible and reversible. It will test on fictional or approved historical examples, measure label agreement and record correction effort. “Use AI everywhere” is not a goal. Neither is a vendor demonstration on three ideal messages. Include ordinary messages, ambiguous ones and cases the automation should decline. Set the boundary in writing. The model cannot approve refunds, promise availability, send external replies or create access credentials. The surrounding application owns authentication, authorisation, logging and execution. If a model suggests a tool action, the application still decides whether the action is allowed. Current function-calling systems use structured tool definitions, but tool availability is not permission to execute without checks. This independent course is not affiliated with or endorsed by any AI or automation provider. External accounts, API access, usage fees, workflow software and deployment are not included. Products and interfaces change; verify current official documentation before implementing. The deliverable here is a workflow map, not a production integration.

1. Choose a useful content promise and repeatable themes
2. Script a clear hook, body and close
3. Shoot and edit for a vertical, accessible frame
4. Publish responsibly and learn from useful signals

A TikTok content plan begins with a reason for a particular viewer to stay, not a promise to go viral. Describe the audience situation, the useful change each video can create and the perspective you can support. A content promise might be “small-space repair demonstrations for first-time renters”. It is specific enough to guide ideas without claiming that every post will reach a large audience. Create three repeatable themes, sometimes called content pillars. One can answer questions, one can demonstrate a process and one can show decisions or mistakes. Themes reduce blank-page pressure while allowing different stories. They should come from real audience needs and your ability to produce evidence, not only from current trends. Trends can inspire a format, but the message still needs to fit the audience and your rights. Fictional worked example: Patch & Peg is an invented repair workshop. Its audience is renters who want reversible ways to care for a small home. Its promise is “show one safe, removable household fix at a time, including when to ask the property owner or a qualified professional.” Its three themes are tool basics, before-you-start checks and one-minute demonstrations. The workshop does not promise savings, property approval or a result in every home. Build an idea bank with question, theme, evidence, visual action and boundary. A question such as “How do I stop a rug edge curling?” may support a demonstration, while electrical work is outside the project’s beginner scope. Record the source of factual or safety information and remove ideas you cannot demonstrate responsibly. This course does not provide trade, safety or legal advice. TikTok recommends creating for the platform’s vertical context, but a native format is not permission to copy another creator. Use your own footage or appropriately licensed material. Credit or permission requirements depend on the source and intended use. Current Community Guidelines also address unoriginal content, fake engagement and manipulated media. Review the rules that apply before publishing. This independent course is not affiliated with or endorsed by TikTok. A TikTok account, production equipment, music licences, advertising access and promotion budget are not included. Start with six honest ideas you can actually make, then choose one based on usefulness and production feasibility rather than predicted reach.

1. Plan the edit and organise source media
2. Build a clear rough cut on the timeline
3. Refine captions, text and audio for comprehension
4. Export, quality-check and preserve a usable master

Video editing starts before the timeline. Define the audience, one takeaway, intended frame and approximate duration. Then write a simple beginning, middle and end. This gives every clip a job and prevents effects from becoming the structure. A beginner project should be small enough to finish and review, such as a 40-second vertical demonstration made from six original clips. Create a project folder outside the editor with source video, audio, graphics, script, rights notes and exports. Keep original files unchanged. Rename working copies with useful descriptions and dates rather than “clip final 7”. Confirm that you created the material or have appropriate permission for the intended use. Stock assets, templates, typefaces and music can have separate terms; availability inside software does not automatically establish every use right. Fictional worked example: Lantern Lunchbox, an invented recipe project, plans a short demonstration called “Pack a crisp salad without mixing the dressing early”. Its six shots are ingredients, dry container, dressing jar, layering, lid check and final packed box. The source log records who filmed each shot, whether a hand is recognisable, the music source and the recipe-fact reviewer. The example does not give dietary advice or promise a particular storage outcome. CapCut is available in different versions, and controls can move. The stable workflow is to create a project, import chosen media, place it on a timeline and preserve the originals. Set the intended aspect ratio before detailed text placement. Verify current official CapCut guidance for the device and version you use instead of depending on one brittle menu path. Review each clip before import. Reject footage that exposes private details, lacks permission, cannot show the required action or has unusable sound. Note technical differences such as orientation, resolution and frame rate, but do not upscale poor footage and call it improved evidence. Back up material appropriate to its sensitivity and delete accidental personal data through the organisation’s process. This independent course is not affiliated with or endorsed by CapCut or ByteDance. A CapCut account, paid features, cloud storage, stock licences, music rights and publishing accounts are not included. The first deliverable is an edit brief and source log, not a finished viral video.

1. Turn an objective into a test
2. Assign work without losing accountability
3. Build a testable creative brief
4. Launch with a bounded budget
5. Read the evidence and plan the next experiment

Start with a decision, not a channel. Example: a small stationery shop wants to know whether a product comparison page helps first-time visitors choose a notebook. The test is not “grow revenue”; it is “does a comparison page help qualified visitors complete a useful next action?” Write the audience, one intervention, the observable action, the measurement window and the decision you will make. Separate leading indicators (page interactions) from outcomes (completed purchases). A rise in clicks alone does not establish that the business improved.

1. Choose a workflow worth assisting
2. Create an input contract
3. Evaluate outputs with a small test set
4. Put approval at the boundary
5. Operate and improve the pilot

Select a repeatable, low-risk task with visible inputs and outputs: turning approved release notes into a support draft, for example. Map the current process before adding AI. Record what must stay under human control. Do not start with autonomous refunds, legal decisions or account changes. If a manual process has no agreed acceptance criteria, automation will usually make its ambiguity faster.