Set a marketing goal you can actually measure

You can spend a week posting, designing and checking numbers without knowing whether any of it helped. A useful goal gives that work a destination. This chapter will help you choose one customer action that matters, write a realistic target, decide how to count it and review the result without pretending that one number explains everything.

By the end, you will have a one-page goal card: an outcome, a clear measure, a baseline or a plan to establish one, a target, a deadline, a data source and a first action. You can do the exercise with a notebook. No analytics subscription is required.

Begin with the decision, not the dashboard

Imagine opening a dashboard that reports 2,000 impressions, 140 clicks, 35 form visits and 12 enquiries. Which number tells you that your marketing worked? There is no automatic answer. It depends on what the business is trying to achieve, which customer action it values, how the numbers were collected and whether it could serve the extra demand.

An impression is a recorded appearance of content on a screen; it does not mean a person remembered it. A click is an interaction with a link or button; it does not prove a purchase. An enquiry is a request for information, but its meaning depends on what counted as one. Numbers are clues. The goal tells you which clues matter first.

Start with a sentence in everyday language: “We want more suitable people to request an appointment online.” Then ask what behaviour would show progress: completed appointment requests, not simply views of a post. Finally ask what you can count reliably. This direction, from desired result to measure, helps prevent an attractive chart from choosing the goal for you.

This builds on the earlier chapters: you first clarify what your business offers, who might value it and what alternatives people have. If those questions are still open, use a small learning goal such as “speak with five people who face the problem” rather than inventing a confident sales forecast. A useful target reflects your current evidence.

Where this fits in the whole plan

PR Smith’s SOSTAC® planning framework asks six connected questions: Situation (where are we now?), Objectives (where are we going?), Strategy (how will we get there?), Tactics (which detailed methods?), Action (who will do what?) and Control (how will we know and adjust?). The current chapter concentrates on Objectives, while the baseline comes from Situation and the review belongs in Control. The framework is a map for planning, not a requirement to complete six long documents before taking a small step (SOSTAC, no date).
image
Figure 1 in words: Situation establishes the starting point. Objectives state the destination. Strategy, Tactics and Action connect the two. Control checks the result and informs the next Situation. In this article, we work mainly on the destination and the check.

Goal, objective, action and measure are different things

These words can overlap in everyday speech. To stay clear in this series, a goal is a broad desired change: “Make it easier for suitable customers to book.” An objective is a precise result within a period: “Increase completed online appointment requests from 12 in the previous four weeks to 16 in the next four weeks.” An action is work you choose to try, such as clarifying the booking page. A measure is the defined count, such as completed requests recorded by that page. A target is the value you hope the measure will reach: 16.

The difference matters. “Post three times every week” is an activity you control. It could be a useful task, but it is not evidence that customers got what they needed. “Reach 16 completed requests” is a desired result. You cannot force it merely by writing it down, so you also need a plausible action and a way to learn from the outcome.

If you want a broader view of how an objective fits a business's purpose, Dhruvi's existing Vision, Mission and Objectives in Strategic Management goes deeper. This chapter gives you the simpler working method you need here.
A04-02-goal-ladder
Figure 2
Figure 2 in words: Start with a customer need and a business outcome. Translate them into one measurable objective. Choose an action you can carry out. Collect the defined evidence. If the evidence disagrees with your hope, change your explanation or action instead of changing the numbers.

For a new business with no working booking page, a target of 16 requests may be premature. A first objective might be to publish a clear page and check whether five target customers can find the booking information. That is a learning objective: it produces evidence needed before you set a volume target. “No baseline yet” is an honest starting point, not a failure.

Use SMART as five questions, not a magic formula

SMART is a common way to test a proposed objective. In this chapter it means Specific, Measurable, Achievable, Relevant and Time-bound. You will also encounter versions where A means Actionable or Attainable. These variants are real; the point is to ask whether the objective is both feasible and useful for action, rather than argue about the spelling of a letter (Chaffey, 2025).

A04-03-smart-checks
Figure 3
Figure 3 in words: Specific asks who will do what. Measurable asks what you will count and where. Achievable asks whether capacity and evidence make the target plausible. Relevant asks why the result matters to customers and the business. Time-bound asks when the result will be assessed.

S — Specific: name the action and its boundary

“Grow online” could mean more visitors, more enquiries, more purchases or more return visits. Each calls for a different decision. Specify the action, channel or location and the people concerned. “Completed online appointment requests from the booking page” is more useful than “website engagement.” It names an observable action. If you can explain in one sentence what belongs in the count, you are getting close.

Specific does not require a page of conditions. An objective should be clear enough that another person could count the same thing in the same way. If someone submits twice because a button appears unresponsive, will that count as one request or two? Decide before the review. Do not quietly change the rule to make the target look reached.

M — Measurable: decide what evidence you can actually get

You can count completed form submissions, confirmed bookings or customer questions in a simple log. You cannot count “trust” directly with a click counter. If trust matters, use several imperfect signs, such as what customers say, whether they understand the offer and whether they choose to return. Qualitative notes—what people say and why—can explain a count without pretending to turn every feeling into a precise percentage.

Write down the source, the counting rule and the review interval. For example: “The booking inbox; count one valid request per person after excluding tests and obvious duplicates; compare consecutive four-week periods.” A measure without a rule can change halfway through an experiment and make comparison misleading.

A — Achievable: check resources and uncertainty

Ask whether the business can respond to the target number of requests and fulfil the resulting appointments. Look at the prior period, available time, seasonal changes and any planned change to the website or promotion. The illustrative rise from 12 to 16 requests is not a forecast. It is a proposed target to test against capacity and evidence. Even 16 requests may be too many if only four appointment slots are free.

Achievable does not mean easy. It means the target has a defensible reason and does not require resources you lack. If you have no baseline, collect one first. If a small baseline fluctuates, do not promise an exact growth rate as though it were certain. An alternative use of A, Actionable, adds a useful check: will this number help you decide what to do next? (Chaffey, 2025).

R — Relevant: connect it to a customer and business result

Increasing followers can be useful for some purposes. It is weak as the sole objective when the business needs completed bookings, especially if the new followers are unlikely to need the service. Relevant asks: “If this target is reached, what changes for the people we serve and the business?”

A target can be measurable yet point the wrong way. A rise in form submissions may reflect confusing information that makes people ask basic questions. If the real aim is smooth booking, look at the quality and outcome of those requests too. Relevance is a reason for selecting a measure, not a label you attach afterward.

T — Time-bound: use a real comparison window

“Soon” is not a deadline. Choose a period long enough to observe the action but short enough to review. The example compares one completed four-week period with the next four-week period. Those windows must use the same counting rule. Write their dates on the goal card when you run the exercise. Check weekly for problems, then judge the main target at the end of the agreed window. A single unusually good day should not be presented as a four-week trend.

If a business has strong weekly or seasonal patterns, matching two neighbouring periods may still be unfair. Record holidays, changes in opening hours or a large promotion. You may need a longer view before claiming improvement. The deadline makes the question answerable; it does not remove context.

Turn a vague wish into a goal card

Here is a made-up calculation for teaching. It describes no real business and is not drawn from a slide. Suppose a local service's booking form recorded 12 valid completed appointment requests in a previous four-week period. The business can handle more requests and wants to make the booking page clearer. It proposes 16 valid completed requests in the next four weeks, with the same form and counting rule. The difference is four requests; relative to 12, that is about 33% more. This percentage is arithmetic, not evidence that the goal is likely to be achieved.

The resulting objective could read:

During the next four-week period, increase valid completed appointment requests through our website form from a baseline of 12 to a target of 16, using the same counting rule, while responding to every valid request within two working days.

The response condition is a guardrail: a second check intended to stop the business from celebrating more requests while its service worsens. It is not proof that every customer is satisfied. If response time is hard to measure, start by recording the date each valid request arrived and the date the first useful reply was sent. Use a consistent meaning of “working day.”

image
Figure 4
Figure 4 in words: The example result is valid completed appointment requests. The baseline is 12 in the preceding four weeks; the target is 16 in the next four weeks. The source is the website form record, counted once per valid request with tests and duplicates removed. The guardrail is a useful response within two working days. A real goal card should also contain the actual start and end dates, owner and first action.

Now separate the objective from the hypothesis. The objective says what result you hope to reach. A hypothesis says why a chosen action might help: “If we make appointment availability and the form button easier to find, more suitable visitors may complete a request.” That explanation can be wrong. The point of a small test is to find out.

You may also decide that 16 is an unreasonable target after seeing capacity or typical variation. In that case, revise it before the test and record the reason. After the period has begun, preserve the original target in your notes even if the result falls short. You can set a new target for the next period, but changing the old one erases learning.

A04-05-baseline-target
Figure 5
Figure 5 in words: The first four-week window is an illustrative baseline of 12 valid requests. The second is a target of 16, not an observed bar. The change required is four additional requests. Actual results should only be entered after the second window ends.

Choose a small set of measures that can answer different questions

A metric is any defined measure you track. A key performance indicator, or KPI, is one of the few measures chosen because it shows progress towards an important result. The distinction is about its purpose, not about a special type of number. KPI.org describes KPIs as critical quantifiable measures of progress towards a desired result and recommends comparing them over time while considering leading and lagging indicators (KPI.org, no date).

In our example, valid completed requests could be the primary KPI. Form visits and button clicks are supporting metrics: they can help explain where people leave, but more clicks are not automatically the outcome. Response time is a guardrail for service quality. Confirmed appointments could be a further outcome check. You do not need every available platform number on the first scorecard.

A04-06-measurement-chain
Figure 6
Figure 6 in words: People may see booking information, visit the form, complete a valid request, receive a useful reply and book an appointment. The first steps can help diagnose friction; a completed request is the example's primary KPI; response and appointment checks show whether the extra requests remained useful. This sequence is a reasoning aid, not a claim that everyone follows one fixed path.

The scorecard image below offers four different questions. It is an example to adapt, not a list of numbers every business must collect.

image
Figure 7
Figure 7 in words: Use a volume count such as valid requests to assess the main objective. Check quality through relevant requests or customer feedback. Check value through confirmed appointments or sales where that outcome is available. Check cost through the time or spend needed to produce the result. Use sources you can actually maintain. An increase in any one column does not prove a profitable or better customer experience.

For a solo business, a notebook may be enough: record weekly counts and one sentence explaining anything unusual. If the website reports “conversions,” find out precisely what event it labels that way. A form view, button tap and successful submission are different events. Compare like with like; do not add a platform’s clicks to your own completed requests as though they represented the same people. Nor should you assume a social post caused an enquiry merely because both happened in the same week.

What if the only numbers you have are followers and views?

Use them as clues about possible awareness, then create a path to a more meaningful action. Can a person understand the offer and find a clear next step? Can you count a valid question, a booking request or another action appropriate to your business? A follower count alone cannot show whether the intended customer received value. Conversely, if your purpose is genuinely awareness, you might choose an appropriate reach measure—while checking whether the audience and message are relevant. The objective comes first.

What if your sample is tiny?

Treat changes cautiously. Going from one request to two is a 100% increase, but it is also one additional request. Report both the count and the context. Ask what the people who visited needed, whether the measurement worked and whether the result repeats. Avoid announcing a general trend from a handful of observations. The goal is a practical decision, not statistical theatre.

What if you have no analytics at all?

You can record a baseline manually, as long as the method is consistent and respectful of customers' information. For example, count valid booking enquiries each week and note their channel only when the customer volunteered it or the channel is evident from the record. Do not collect extra personal data just to decorate a dashboard. If the count is unknown, write “baseline unknown—measure four weeks using this rule”. Your first objective can be to establish a trustworthy baseline and remove an obvious customer obstacle.

Review without forcing a success story

Schedule a short weekly check and one end-of-period review. At the weekly check, confirm that the form works and that requests get replies. Note any change in hours, price, page design or promotion. Do not redesign the entire plan because one Tuesday was quiet. At the end, compare the actual count with the 16-request target and the 12-request baseline, applying the same definition in both periods.

Three outcomes are possible. Above target: check whether the requests were relevant, responded to and turned into useful appointments; ask what else changed. Near but below target: identify where people may have struggled and whether the target or action was plausible. Well below target or no valid data: check the measurement and the customer path before drawing a dramatic conclusion. A failed or uncertain test can still tell you what to change next.

A04-08-review-loop
Figure 8
Figure 8 in words: Define the goal, carry out one action, record what happened, compare it with the baseline and target, then keep, adjust or stop the action. The resulting evidence becomes the Situation for the next planning cycle. This is the Control connection in PR Smith’s SOSTAC® model (SOSTAC, no date).

Avoid the sentence “Our new page caused four extra bookings” unless you have evidence that separates the page change from other influences. A simple before-and-after comparison shows what happened in two periods. It does not, by itself, prove why. You can still make a sensible small decision: “The form is clearer, requests rose and replies stayed timely; we will keep it and continue measuring.” State uncertainty plainly.

Try it: write one goal you could check next month

Use this exercise for your own idea. If you are not operating a business yet, choose a learning goal and do not invent a sales baseline.

  1. Name the customer action. Complete: “I want the right people to ___ because it helps them ___ and helps the business ___.”
  2. Write the counting rule. What qualifies? What will you exclude? Which single source will hold the record? Can someone else repeat the count?
  3. Find your baseline. Use a comparable past period. If none exists, write “unknown” and choose a short baseline collection period. Do not use a guessed number as an observed result.
  4. Choose a target and date. Explain why the target might be possible given capacity, time and evidence. Write exact start and end dates before testing.
  5. Choose one action. State what you will change or try. Keep it separate from the hoped-for result.
  6. Add a guardrail. What would tell you that more volume came with poorer quality, worse service or unreasonable cost?
  7. Plan the review. Decide when you will check that the measurement works and when you will make the keep, adjust or stop decision.
Your finished sentence can be: “From [observed baseline] in [baseline dates], we aim for [target] [defined customer actions] during [future dates], counted from [source and rule], while maintaining [guardrail]. We will try [action] and review on [date].” If the baseline is unknown, write a measurement objective first. That is more useful than a beautifully worded target built on invented evidence.

Quick self-check before you move on

Can you explain your goal without using “engagement” as a substitute for a customer action? Can you point to the exact source of your count? Would you still consider the result good if the guardrail failed? Is the target a proposal rather than a promised outcome? If you answered no to any of these, revise one part of the goal card and try again.

Watch, if video helps you

The article stands on its own. These two optional videos show the goal and its wider planning context.

  • Start with a worked explanation: How to Set SMART Goals for Your Business (Guide), HubSpot Marketing (2024). Its description promises step-by-step prompts, examples and common mistakes. HubSpot sells marketing software, so treat this as a practical explanation rather than independent research. The wording of SMART may differ slightly from the version in this article.
  • See the wider planning map: SOSTAC Planning in 3 minutes by PR Smith, PR Smith Marketing. It is by the framework's creator and introduces all six stages; return to the Objectives and Control boxes in Figure 1 after watching. It is a brief overview, not a full lesson on measuring your own goal.

Where to go next

Your goal card is a working decision, not a promise of results. Keep it near the evidence you gather. The next unit in this series looks more closely at the people you want to help; those customer insights may change which objective is most relevant. Until a neighbouring chapter has a verified public URL, use the Digital Marketing from Zero series title to find the guide in Knowledge.

References

Chaffey, D. (2025) ‘How to define SMART marketing objectives (with example RACE KPIs)’, Smart Insights, 29 July. Available at: https://www.smartinsights.com/digital-marketing-strategy/define-smart-marketing-objectives/ (Accessed: 2 October 2026).
KPI.org (no date) ‘What is a key performance indicator (KPI)?’. Available at: https://www.kpi.org/kpi-basics/ (Accessed: 2 October 2026).
SOSTAC (no date) ‘About us’. Available at: https://sostac.org/about-us/ (Accessed: 2 October 2026).