Launch your digital marketing plan: measure, learn, improve

What this helps you do

You can have a carefully researched audience, a useful offer, a content calendar and a set of channels, yet still hesitate when it is time to launch. What should happen first? How will you know whether the plan is working? If the numbers disappoint, what should you change? This final chapter turns the pieces of a digital marketing plan into a small, testable cycle.

You will build a one-page launch and learning plan. It will name the audience and promise, the first action, the few measures that matter, who owns each task, the checks before launch and the date for a review. No business or customer story is invented here. The numerical examples are labelled as arithmetic exercises; they are not results from a company or from teaching slides. The figures are image files for an editor that has difficulty pasting tables. Every figure has a full text explanation, so you can follow the chapter without seeing an image or watching a video.

The useful question is not “Have I mastered every platform?” It is: Can I make a clear promise to a defined group, deliver it reliably, observe what happens and make the next decision responsibly? Launch can be a modest pilot. It does not have to be a grand campaign.
A26-01-map
Figure 1
Figure 1 in words: A SOSTAC® plan asks about the current situation, objectives, strategy, tactics, action and control. RACE helps examine the customer journey through Reach, Act, Convert and Engage. This chapter joins those lenses: define the plan, test the experience, launch a bounded activity, observe the journey and feed evidence into the next situation review. The diagram is an original teaching synthesis, not a copy of either framework (Smith, no date; Chaffey, 2023).

The simple meaning: a plan is a set of decisions

A marketing plan is a written explanation of who you intend to help, why they might choose you, how they can encounter and use your offer, what you will do, and how you will tell whether the work helped. A list of platforms alone is not a plan. Neither is a dashboard of measurements without an action that those measurements could change.

PR Smith's SOSTAC® structure begins with situation analysis and objectives, then moves through strategy, tactics, action and control. Smith describes Control as specifying measures, timing, cost and a response when the data changes. Dave Chaffey's RACE framework adds a customer journey lens: Reach, Act, Convert and Engage, preceded by Plan. Its stages include both online and offline outcomes (Smith, no date; Chaffey, 2023). You do not need to memorise every framework term to use the central idea: make a choice, make it possible, then learn from what happened.

Situation means the current evidence and constraints: what the intended audience needs, what your offer can deliver, how people currently find you and what resources you have. Objective means the result you are aiming for by a stated date. Strategy means the broad choice of audience, value and route. Tactic is a particular activity, such as a helpful search article or an email. Action gives the task an owner and time. Control is the review system that compares what happened with what you expected and changes the plan when warranted. These definitions are refreshers; you can begin here even if you have forgotten earlier chapters.

What the final output looks like

Your one-page plan needs nine answers: (1) for whom, (2) what problem, (3) what promise and evidence, (4) which first route, (5) what useful next action, (6) what must work before launch, (7) which measures and definitions, (8) who checks and when, and (9) what decision follows different results. Figure 10 is a blank visual worksheet. Write plain sentences; a clear sentence is better than an attractive but ambiguous dashboard.

1. Choose a bounded launch question

Begin with a question you can learn from during a defined period. For example, an abstract question might be: “Can a clear explainer page help the intended audience reach a relevant enquiry form over the next four weeks?” This is a question template, not a story about a business. It can be changed to fit a product, service or educational resource. The period is a planning choice, not a promise that four weeks is enough to prove long-term effectiveness.

A pilot is a limited first release used to check the delivery and gather evidence. Decide its scope: one audience, one primary promise, one main destination and one or two acquisition routes are often easier to interpret than launching every channel at once. That is an editorial recommendation, not a fixed rule about how many channels a real organisation may use. Specify what you are holding constant and what you expect to learn. If the offer, audience, page, budget and message all change together, it will be hard to say which change mattered.

Write down the decision before choosing a metric. “If people reach the explanation but few take the next step, we will inspect whether the page answers their question and whether the next action is understandable.” This is more actionable than “We will get more clicks.” Different possible observations point to different work; a weak result is information, not a moral judgment about the people involved.

Do not treat a pilot as an excuse for an unfinished experience. People who encounter the offer are real people. The promise, pricing, privacy information, support route and essential service should work. Keep any uncertainty in marketing effectiveness, not in whether a person can complete a necessary task safely.

A26-02-question
Figure 2
Figure 2 in words: The card asks for one audience, one need, one promise, one next action and one learning question. It also names a defined period and the limit of what that period can establish. This is a blank planning device; no customer or company has been fabricated.

2. Reconnect the promise to the journey

Now trace the steps a person would actually take. Reach asks whether the right people encounter your material. Act asks whether they interact in a useful way, such as reading a relevant explanation or opening a product detail. Convert asks whether they take the agreed meaningful step, which may be a sale, a qualified enquiry or another outcome appropriate to the offer. Engage asks whether the relationship remains useful afterwards. Chaffey (2023) explicitly includes offline sales and interactions, so do not assume the whole journey is visible in one web tool.

For each stage, write the user's question before writing your channel or metric. A useful plain-language sequence is: “Is this for me?”; “Does it answer my question?”; “What happens if I proceed?”; “Was the promise kept?” The four questions are this chapter's teaching prompts, not labels from Chaffey's diagram. A channel is justified when it can help a person at the relevant stage, not simply because it is popular.

There can be several paths. A person might first hear about a service offline, search for the name, visit the website and make a telephone enquiry. Another might read a guide, leave and return later. A single last-click report is therefore a view of tracked activity under a rule, not a full biography of persuasion. Google explains that attribution assigns credit for a key action among observed touchpoints using a model; different models can assign credit differently (Google, no date a). Use that caution when someone asks which channel “caused” every result.

Put the customer journey beside the operating journey. If an enquiry form succeeds, who receives the message? When do they respond? If the person buys, what delivery follows? If they have a problem, where is support? A high conversion count can conceal poor delivery if nobody observes the later stages. Chapter 24's CRM lesson concerned coordinated help across contacts; Chapter 25's customer lifetime value lesson concerned a cautious financial estimate across time. Their concepts now become questions in the Engage and review stages. Add their public links only after the actual article URLs are verified.

A26-03-journey
Figure 3
Figure 3 in words: Reach is paired with “Is this for me?” and a relevant visit; Act with “Does this help?” and a useful interaction; Convert with “What happens next?” and a defined meaningful action; Engage with “Was the promise kept?” and a service or retention signal. These are illustrative measure categories, not a compulsory four-metric dashboard. The journey is adapted conceptually from Chaffey (2023).

3. Turn the objective into a measurement plan

A metric is a defined quantity you observe. A key performance indicator (KPI) is a metric deliberately chosen to judge progress toward an objective. Not every number on a platform dashboard deserves to be a KPI. Start with one outcome you care about, one or two earlier signals that might help explain it, and a check for unwanted consequences. The exact selection depends on the offer, journey and ability to observe reliably.

A useful objective says what outcome, for whom and by when. It also states a baseline or says that no baseline exists. SMART is a memory aid for making objectives specific, measurable, achievable, relevant or realistic, and time-bound; versions differ in their expansion. PR Smith's page explicitly uses Specific, Measurable, Achievable, Realistic and Time-bound (Smith, no date). The letters are prompts for reasoning, not an automatic quality certificate. “Get more awareness” lacks a measure and date. “Increase qualified enquiries from the chosen audience during the next review period, while keeping response capacity adequate” is a better starting sentence; it still needs an actual baseline, numerical target and definition before use.

Baseline means the observed starting level under a stated method and period. Target means the intended level. Observed result means what is measured after activity begins. Do not type a guess into the baseline field and later call it history. If there is no history, label “baseline unavailable,” use a learning period to establish one and keep any initial target provisional. A small number of observations can swing widely; record counts alongside percentages.

For example, suppose a completely hypothetical test has 200 eligible visits and 8 defined enquiries. The illustrative enquiry rate is 8 ÷ 200 × 100 = 4%. If the next comparable period has 10 enquiries from 200 eligible visits, the rate is 5%, an increase of 1 percentage point or 25% relative to 4%. These are arithmetic exercises only. They say nothing about lead quality, margins, seasonality, causation or statistical certainty. If a denominator was 20 rather than 200, one additional enquiry would move the percentage far more. Report the counts and definitions.

Build a metric definition card

For each chosen measure, write: its name in ordinary words; the exact action counted; its numerator and denominator if it is a rate; the population and date range; its source; how often it updates; who checks it; known missing data; and what decision it informs. “Conversion rate” without an event and denominator is incomplete. A form submission might fire twice when a page reloads, or a test submission might be mistaken for a genuine enquiry. The definition card forces these questions into the open.

If you use Google Analytics, its event records a user interaction when configured to do so, and a key event is an event designated as particularly important. Google says marking an event as key affects reports from the time it is marked, rather than rewriting earlier data (Google, no date b). The plain marketing concept of “conversion” is a desired meaningful action; platform labels and Google Ads conversion settings are more specific. A test should verify that the intended action fires once under the intended conditions. Do not treat all page views as completed enquiries.

Google Search Console measures search visibility and clicks under its own definitions. Its performance reports distinguish impressions, when a link appears under the report's counting rules, and clicks, when someone selects a qualifying link (Google, no date c). These are useful for a search discovery question. They are not identical to website sessions, engaged readers or sales. A search impression cannot tell you whether the person read and valued the page. Compare like with like: same property, query or page scope, search type, geography and period when possible.

A26-04-measure
Figure 4
Figure 4 in words: The image table separates the objective, outcome measure, early signal, quality check and definition details. A blank baseline is explicitly “not established” until observed. The owner, source, date range, denominator and next decision are as important as the displayed number. The template is original; Google's platform terms are explained above (Google, no date b; Google, no date c).

4. Look at the whole result, not just the attractive number

Leading indicators are earlier signals that might suggest progress; lagging indicators appear later, after the outcome has happened. An increase in relevant visits can be an early signal. Completed enquiries or contribution after costs may arrive later. The distinction depends on the decision and timescale. A search click is a late outcome for a search snippet test but an early signal for a sale. A dashboard with many leading signals can still fail to show whether the promise was kept.

Add a quality or guardrail measure. This is a check that prevents the main target from rewarding an undesirable result. If the main measure is enquiries, a guardrail might be whether those enquiries are relevant and answered within the capacity you have. If the objective is sales, a guardrail might concern refunds, delivery failures or complaints. Define how the check is recorded without collecting unnecessary personal details. Your guardrail should be capable of changing the decision; otherwise it is decoration.

Financial measures also need honest labels. Revenue is money received before relevant costs. Contribution subtracts defined variable costs but is not necessarily full profit. Customer acquisition cost requires a defined spend and a count of genuinely new customers. Customer lifetime value forecasts contribution across a stated horizon and assumptions rather than guaranteeing future cash. Chapter 25 showed why an attractive CLV ratio can still hide cash timing and service constraints. If you report return on investment (ROI), define the investment, attributable return and costs; do not label sales divided by ad spend as overall profit. These definitions are a practical integration of the previous chapter, not a new universal benchmark.

Beware of the vanity metric: a number that looks impressive yet does not answer the decision at hand. A large view count may be useful for a reach objective, but weak evidence for qualified enquiries. A low click cost may reflect irrelevant clicks. A high average time on page might mean deep reading or confusion. The problem is not the metric's existence; it is claiming it proves something it does not. Pair counts with a qualitative check, such as whether a small sample of enquiries actually concerns the promised offer, without pretending the sample represents everyone.

An interpretation exercise

Consider two possible observations from the same abstract pilot. In the first, search impressions rise but relevant page visits and meaningful actions do not. You would inspect the query intent, title and snippet, then the page's relevance. In the second, relevant page visits rise and the next action is used, but the team cannot respond on time. You would address capacity and fulfilment before increasing traffic. Neither observation is an invented business result. The exercise shows why a decision rule should describe what to inspect and who acts, rather than promising that every number above a threshold means success.

A26-05-interpret
Figure 5
Figure 5 in words: An illustrative row shows a baseline that must be observed, a provisional target and an observed result. Another row records counts and a quality check. The graphic warns that a gap between target and result is a prompt to investigate definitions, context and delivery before declaring a cause. It contains no real performance data.

5. Test the experience before spending effort to promote it

Before a public launch, follow the same route a new reader would follow. Open the message from a phone and a larger screen; check that the landing page says what the message promised. Use the primary link, form, purchase or booking path. Confirm that success and error states make sense, that confirmation is delivered, that the right owner receives the result and that the promised next step can actually be fulfilled. Test links and naming conventions with the people who will use the report.

Tracking validation is different from merely seeing a dashboard number. Trigger a labelled test action, check that it appears in the intended place, confirm it is not counted twice and exclude the test from any public claim about demand. For actions that happen offline, record a transparent handoff rather than pretending website analytics can see them. Keep the capture proportionate to the question. An observed “zero” can mean no action, a broken tag, a blocked tool, a missing connection or a delay in reporting. Investigate before treating zero as proof of no interest.

Accessibility belongs in the prelaunch experience. The W3C Web Accessibility Initiative recommends preliminary checks of page title, image alternatives, headings, contrast, text resizing, keyboard focus, forms and multimedia alternatives; it warns that quick checks are not a complete accessibility evaluation (W3C, 2023). In this article's own figures, the accompanying “Figure in words” passages preserve the teaching if a graphic does not load. On your page, test the real controls with a keyboard and read form errors in context. A beautiful campaign fails its purpose if the intended people cannot use it.

If the plan includes electronic mail marketing in the UK, check the applicable Privacy and Electronic Communications Regulations rules and lawful handling of personal data before sending. The Information Commissioner's Office explains that consent is normally required for unsolicited electronic mail marketing to individual subscribers, with specific exceptions such as appropriately met soft opt-in conditions; the details depend on recipient and context (ICO, 2026). A contact form submission is not automatically permission for a newsletter. The ICO guidance is the appropriate current source for the actual campaign; this chapter is a planning prompt, not a legal determination.

Make a small go / fix / pause decision. “Go” means the offer and core journey work, measures are defined and the team can deliver. “Fix” means a repairable problem has an owner and date. “Pause” means a critical promise, accessibility route, data practice, delivery capacity or measurement needed for the test is unresolved. The exact threshold is a management decision; it should be written down before results create pressure to rationalise a broken experience.

A26-06-gates
Figure 6
Figure 6 in words: The image table asks whether promise and destination match; whether the action and handoff work; whether measurement is tested; whether accessibility and data practices have been checked; and whether capacity and ownership are in place. The result is Go, Fix with an owner and date, or Pause. These are an original checklist informed by W3C (2023), ICO (2026) and Google (no date b), not an official certification.

6. Write the action plan someone can carry out

For each launch activity, name one accountable owner, a due date, the required asset, the destination, the budget or effort limit and the handoff. An owner can ask for help, but “everyone” is not a useful answer to who will notice a broken form. If you work alone, the owner is you, and a calendar reminder can hold the review date. A dependency is something another task needs first: you should test the page and measurement before directing people there.

Keep creative work tied to the audience's problem. A useful message states what the person can learn or do, how the destination supports that promise, and what happens after the action. A short editorial review can catch exaggeration, unclear eligibility, invented proof or an inaccessible visual. A small pilot budget is a ceiling for testing, not a reason to spend it all. If you have only organic channels, effort and service capacity are still costs and constraints.

An operating rhythm makes the plan live. You may check urgent delivery issues daily during launch, inspect measurement quality early, compare agreed measures at the end of the pilot and schedule a later review for outcomes that take longer. The right cadence follows the speed of the channel and outcome. A purchase can happen quickly; a repeat purchase cannot be judged responsibly one day after acquisition. Do not move the target or redefine a metric halfway through and then compare the new figure as though it used the old method. Record a dated change note.

When several people are involved, a short review note should state: what changed, what we observed, what we cannot yet tell, what we will do, who owns it, and when we will look again. Keep the raw definition and source available so a colleague can reconstruct the conclusion. PR Smith's Control section explicitly links ongoing measurement to the next Situation analysis (Smith, no date). That is why this final chapter closes a loop rather than the learning journey.

A26-07-rhythm
Figure 7
Figure 7 in words: Preparation assigns assets and owners; a prelaunch test validates the route; a bounded launch records changes; early checks catch broken delivery; a scheduled review decides the next action. The cadence is chosen to match the outcome. This is an original implementation schedule, not a compulsory industry calendar.

7. Read results with humility

At review time, first ask whether the data is comparable. Did the page, audience, period, spend, price or event definition change? Was there a holiday, a stock problem, an unusually large mention elsewhere or an outage? Did a new consent setting alter what could be tracked? Record plausible context without inventing a causal explanation. An observed association between a channel and a sale does not by itself prove that the channel caused the sale.

Attribution is particularly tempting because tools display tidy channel labels. Google describes reporting models that assign credit to the touchpoints observed in a path, and says a model change can affect historical and future reporting (Google, no date a). A change in reported channel credit can therefore come from the rule, not only from changed behaviour. When you compare periods, record the model and avoid adding totals from differently defined reports as if they were one population. Direct conversation with customers, when appropriate and consented to, can add useful context but is not a flawless attribution instrument either.

Small numbers need care. If two of ten visitors act, the rate is 20%; if one more acts, it is 30%. That ten-percentage-point swing comes from one event. A short pilot can discover broken steps, confusing messages and promising questions. It may not establish a reliable long-term lift. Avoid saying “this works” because a small, uncontrolled before-and-after chart moved. A more rigorous experiment might compare randomly assigned alternatives under a predeclared outcome, but design, sample size and implementation matter. If the decision is costly or difficult to reverse, seek stronger evidence.

Use the diagnostic ladder in Figure 8. Start with data quality. Then locate the stage where the path changes. Examine the relevance and usability of that stage. Check delivery capacity and guardrails. Only then decide whether to continue, revise, test another idea or stop. This sequence does not require every metric to rise. It aims to make the next action proportional to the evidence.

A26-08-diagnose
Figure 8
Figure 8 in words: Check data definition and tracking first. Next locate the journey stage, examine the message and route, check quality and fulfilment, and choose a bounded next action with an owner. The arrows signify an investigation sequence, not proof that one stage causes the next. The graphic is an original editorial decision aid.

Four responses to a review

Continue when the intended experience works, the evidence is adequate for the current scope and the costs and guardrails are acceptable. Repair a broken step before increasing exposure. Revise or test one meaningful assumption when evidence points to a plausible weakness but causation is uncertain. Stop or pause when the promise cannot be delivered, a serious barrier remains or continued spending is unjustified. Write down the reason, not merely the status colour. If a measure was poorly chosen, improve the definition and mark the break in the series of results.

The review can also reveal that the original problem statement was wrong. That is not a reason to conceal the result. Return to research, listen to the needs people actually express and revise the offer. The series began with understanding value and customers; it ends by allowing evidence to correct the first assumptions.

8. Assemble a one-page plan without pretending certainty

Open the blank worksheet in Figure 10 and fill it from top to bottom. First state the chosen audience and need in one sentence. Add the specific value promise and the evidence or limitations behind it. Choose a primary journey and destination. Name the useful action and the person who will fulfil it. State one objective and review date; if you do not have a baseline, say so. Choose a short set of measures with definitions and a guardrail. Set the initial scope and resource limit. Write the prelaunch checks, owner and decision rule. Schedule the first review and define the question that review will answer.

If you feel the temptation to add a long list of platforms, place each one under a purpose: “This helps the selected audience find the explanation,” or “This helps someone who asked for information receive the next step.” A channel without a purpose can wait. If you feel the temptation to add every available metric, ask which decision each metric changes. If none, it can remain background context rather than a KPI.

The plan can be small while the article explaining it is detailed. The one-page output is a working tool, not a substitute for the evidence, content, operating procedures or specialist review that a larger organisation may need. Its value is to expose missing decisions. If you cannot say who will answer an enquiry, a carefully calculated click-through rate will not repair that omission.

A26-09-decision
Figure 9
Figure 9 in words: The review record contains observation, interpretation with uncertainty, decision, owner and next date. A separate line records what changed in the measurement method. Empty cells are deliberate: the figure is a reusable template, not a report of a campaign.

A26-10-worksheet
Figure 10
Figure 10 in words: The two-part blank worksheet collects audience and need, value promise, primary route and meaningful action, pilot scope and owner, objective and baseline, outcome and early measures, quality guardrail, source and definition, prelaunch gate, review date and decision rule. It leaves space for evidence, uncertainties and what to do next. You can copy these headings into a plain text document if images are inconvenient.

Common mistakes and what to do instead

“We must use every channel.” Begin with the audience's route and the capacity to serve it. Expand only when another route has a clear job and owner. “The dashboard will tell us what to do.” Write the question and decision rule first, then interpret the measures in context. “A view or click is a customer.” Keep visibility, interaction, meaningful action and sustained value distinct. “Zero means nobody wanted it.” Validate the measurement and the opportunity to act before drawing that conclusion.

“We can measure everything later.” Some events must be configured before the period you intend to study; Google says marking a key event does not retrospectively create historical key-event data (Google, no date b). “More enquiries always means a better plan.” Check relevance, service capacity and whether people received the promised help. “An attribution chart proves causation.” State which touchpoints and model it includes, then use it as one source of evidence (Google, no date a).

“The plan is finished after launch.” A plan contains a review date and a way to revise assumptions. Control returns evidence to Situation in SOSTAC® (Smith, no date). “A pilot permits a broken experience.” Test usability, accessibility, privacy and fulfilment before asking people to participate (W3C, 2023; ICO, 2026).

Watch, if video helps you

The written explanation and worksheet are complete on their own. The following are optional, provisional choices based on public video metadata, descriptions and original-source provenance checked on 15 November 2026. Neither video was played or transcript-reviewed for this draft. Verify playback, spoken language, captions, exact duration and relevance before publication.

Simple overview — SOSTAC - how to write the perfect plan (in 4 minutes), PR Smith. The framework creator introduces the sequence used to connect planning and Control. Public metadata shows an upload on 3 October 2012; its title says four minutes, but the precise player duration was not independently verified. The video is older but concerns a durable planning structure, not a current platform interface. Its description includes course promotion. Watch for the questions each stage asks; return to this article for current analytics and prelaunch checks.

Go deeper — Lec 5: Conceptual Framework for Marketing Metrics, NPTEL IIT Guwahati. This is an academic teaching lecture by Prof. Kuldeep Baishya's official NPTEL channel, linked to the institution's Metrics and Analytics of Marketing course. Public search metadata shows an 8 January 2026 upload and 185 views at the research check; the exact runtime was not verified. The course lists marketing management as a prerequisite, so treat it as an optional deeper perspective rather than the starting explanation. Academic teaching is not automatically a peer-reviewed research paper. Content and playback remain to be checked before release.

Check your understanding

You have 300 visits to a page and 12 form submissions in a pilot. Can you claim a successful campaign? You can calculate an illustrative 4% submission rate from 12 ÷ 300 × 100, provided the visits and submissions use the stated period and definitions. You cannot conclude success from that number alone. Check that the 12 are genuine, relevant and correctly counted; compare with a stated objective and baseline if available; inspect response capacity, costs and other outcomes; and record what the pilot can and cannot establish. If tracking was broken, repair it before interpreting the rate.

Now write your own decision rule in a sentence: “At the review date, if ______, we will inspect ______; if the experience or fulfilment fails, ______ owns the repair before we increase activity.” A good answer names observable conditions and an accountable action without promising certainty.

Where to go after Chapter 26

The series has taken you from the basic meaning of marketing through customers, a strategy, channels, messages, relationships and measurement. Revisit the chapter you need when the next decision appears. If an unfamiliar word returns, define it again in your plan. If your evidence changes, revise the earlier assumption openly. Your first useful cycle is plan → check → launch → observe → decide → plan again.

References

Chaffey, D. (2023) RACE marketing model. 5 September. Available at: https://www.davechaffey.com/digital-marketing-glossary/race-marketing-planning-model/ (Accessed: 15 November 2026).
Google (no date a) Change the reporting attribution model for key events. Available at: https://support.google.com/analytics/answer/16291112 (Accessed: 15 November 2026).
Google (no date b) Mark events as key events. Available at: https://support.google.com/analytics/answer/13128484 (Accessed: 15 November 2026).
Google (no date c) How are you performing on Google? Available at: https://support.google.com/webmasters/answer/10268906 (Accessed: 15 November 2026).
Information Commissioner's Office (ICO) (2026) Guidance on direct marketing using electronic mail. Updated 28 April. Available at: https://ico.org.uk/for-organisations/direct-marketing-and-privacy-and-electronic-communications/guidance-on-direct-marketing-using-electronic-mail/ (Accessed: 15 November 2026).
Smith, P.R. (no date) SOSTAC® Planning. Available at: https://prsmith.org/sostac/ (Accessed: 15 November 2026).
World Wide Web Consortium (W3C) (2023) Easy Checks – A First Review of Web Accessibility. Updated 9 August. Available at: https://www.w3.org/WAI/test-evaluate/preliminary/ (Accessed: 15 November 2026).

Further watching

NPTEL IIT Guwahati (2026) Lec 5: Conceptual Framework for Marketing Metrics. 8 January. Available at: https://www.youtube.com/watch?v=8xPhwhaL43s (Accessed: 15 November 2026).
Smith, P.R. (2012) SOSTAC - how to write the perfect plan (in 4 minutes). 3 October. Available at: https://www.youtube.com/watch?v=StYGaCeaRkQ (Accessed: 15 November 2026).