Marketing Planning and Goals Playbook

A marketing plan starts from the company’s revenue goal, not from a list of channels. Work back from that goal to one marketing number, usually qualified pipeline, then choose one objective for the quarter with three key results you can grade, each with a baseline, a target, a date and an owner. Cost the plan channel by channel against a budget ceiling, check it weekly, monthly and quarterly with named owners, and score it at quarter end before you set the next one.

Most founders I talk to have goals but no plan that connects them. The board deck says $1.2 million of ARR; the marketing document says “launch a webinar series”. Nothing in between tells anyone whether the webinars are enough.

SMART goals and OKRs are two tools for writing that middle part down. Neither is a law. In my view, SMART works as a checklist for writing one goal and OKRs work as a format for choosing what to bet on, and any part of either that doesn’t help you decide something can go.

This playbook is the planning system itself: the arithmetic, the goals, the budget, the meetings and the reset. Why the yearly, quarterly, monthly and weekly clocks fit together, and why every target needs a guardrail, is in Marketing Planning Cadence and OKRs. Turning a quarter’s goal into week-by-week experiments is the 90-Day Growth Playbook.

The company’s revenue goal is the one number everyone already agrees on, so start there. Turn it into new revenue: the year-end goal, minus where you start, plus the revenue you expect to lose to churn and downgrades. Founders often leave out the churn line, and then the plan is short from the first day.

Divide the new revenue by your average first-year contract to get customers, then by your win rate to get opportunities. Opportunities, or the qualified pipeline they add up to, are the marketing number I prefer at seed: marketing can move it inside a quarter, sales agrees what counts, and it links straight to revenue. The Pipeline Coverage and Forecasting Playbook shows how much pipeline a target needs at your win rate.

Re-run the arithmetic every quarter from what is left of the year, not from the original split. A quarter that falls short moves its gap into the quarters still ahead, and the plan should show that.

An OKR is an objective, a short statement of what you want to change, plus a few key results that show whether it changed. Google’s guide, the version most founders meet first, calls objectives ambitious, grades OKRs on a scale of 0.0 to 1.0, and suggests three to five objectives with around three key results for each^[1]. That range suits a company of Google’s size; for a marketing team of one to three people, I set one.

One objective forces the choice a plan exists to make. List the candidate bets from last quarter’s review, place each by how much it could add to the marketing number and how much evidence you already have, and make the strongest the objective (Figure 2). The rest keep running as they are, wait, or get a small test from the budget reserve in Step 5.

Keep the number from Step 1 outside the OKR. The pipeline number is a commitment the business planned on, reported every week; the objective is the stretch bet you expect to help reach it, graded honestly at quarter end. If sales also has an objective, the company sheet has two, which is how the planning cadence article sets them up.

SMART is a checklist, usually traced to George T. Doran’s 1981 Management Review article “There’s a S.M.A.R.T. way to write management’s goals and objectives”, and its letters have been read several ways since. The version I use is specific, measurable, achievable, relevant and time-bound. I add three things the acronym leaves out: a baseline, a quality check and an owner (Figure 3).

The baseline does the most work. “Forty partner MQLs” means little until you know the number was 9, and a target with no baseline can’t be judged too easy or too hard. If you have no baseline, measure for two to four weeks before you set the target, or write the target as a labelled assumption and replace it after the first month.

The quality check is there because a hard number can be hit the wrong way. Ordóñez and colleagues’ 2009 review lists the side effects of specific, challenging goals, among them a narrow focus that neglects whatever the goal leaves out, and more unethical behavior^[2]. In a 2004 laboratory experiment by three of the same authors, 154 undergraduates scored their own word puzzles, and those who missed a specific goal overstated their results more often than those told to do their best, most of all when they had fallen just short^[3]. For an MQL target, the check I use is acceptance by sales.

Make the target hard. In Locke and Latham’s summary of 35 years of goal-setting research, specific, difficult goals led to higher performance than vague ones such as “do your best”, as long as people were committed to the goal and got feedback on their progress^[4]. The same holds for teams: a 2011 meta-analysis of group goals found that specific, difficult goals led to considerably higher group performance than nonspecific goals, a standardized difference of d = 0.80 across 23 comparisons^[5]. Locke and Latham also report that on complex tasks people don’t yet know how to do, a specific performance goal can backfire, and a goal to learn, such as finding workable strategies, does better.

That finding on complex tasks is why I handle “achievable” with care. For a channel you already run, set a stretch number. For a channel you have never tried, set a learning goal, such as “run three tests and find one with a cost per MQL under $500”, and turn it into a number next quarter.

An outcome is something that changed outside the company: a buyer booked a call, a partner referred a client, an opportunity was created. An activity is the team’s own work: a webinar held, a post published, a kit sent. My test: if you could hit the number without anyone outside the company doing anything, it is an activity.

Key results should be outcomes, because an activity can be done perfectly and change nothing. Activities still belong in the plan, as the task list under each key result. Between the two sit leading indicators: weekly numbers that move before the outcome does, such as webinar sign-ups from partner invitations (Figure 4).

Each layer is reviewed differently. Outcomes are graded at quarter end, leading indicators are read every Monday, and activities are checked off. When an outcome misses and every activity was done, the plan was wrong rather than the people, and that is worth knowing.

Budget the plan, not last year’s number. Cost each channel from the outcomes you expect it to produce: planned MQLs times what an MQL has cost you there, or a fixed cost for things like an event. Add a small reserve for the one test the quadrant in Step 2 pointed to.

Then check the total against a ceiling worked out from the top down. At seed I budget against the round, not revenue: my working range is 10% to 20% of the round, spread across its runway, as I explain in How Much Should a Startup Spend on Marketing? That band is my view, not a published benchmark, and the Marketing Budget Calculator runs it for your stage.

If the bottom-up plan costs more than the ceiling, change the plan, not the arithmetic. Cut the weakest channel, lower the commitment and tell the board, or raise the ceiling on purpose. A plan that only adds up at a cost per MQL you have never achieved is not a plan.

A plan nobody looks at stops being a plan by week three. Three meetings are enough, and each needs one owner who books it, brings the numbers and writes down the decision (Figure 6). The planning cadence article covers what each clock is for; this step is about who runs each meeting and what it may decide. Google’s guide describes annual and quarterly OKRs, shared and graded at company-wide meetings each quarter, and notes that some teams revisit theirs a few times a quarter^[1].

The weekly check reads the commitment and the leading indicators, marks each key result on track, at risk or off track, and agrees one to three actions. The monthly review moves money between channels; my rule of thumb, not a benchmark, is to let it shift up to 20% of the quarter’s budget without going back to the board. The quarterly review scores and resets, which is Step 7.

The weekly half hour earns its place. A 2016 meta-analysis of 138 experiments with 19,951 participants found that prompting people to monitor their progress made them more likely to reach their goals, more so when the progress was physically recorded or reported publicly^[6]. Monitoring a behavior changed the behavior but not the related outcome, while monitoring the outcome changed the outcome, which is why the weekly check reads the commitment and not only the task list.

Write the decision rights down before the quarter starts. Who can pause a channel, who can release the reserve, and who tells the board when the commitment slips are the questions that otherwise get argued about in week eight.

At quarter end, grade each key result as actual divided by target, and report the commitment beside it as a plain share of plan. Google’s guide calls the 60% to 70% range the sweet spot and expects an average of 0.6 to 0.7 across all OKRs^[1], so a 0.7 on a stretch goal is a good quarter, not a failure. The same guide says OKRs “are not synonymous with performance evaluation”, so a grade tells you about the plan, not about the person who owned it. The commitment is held to a stricter standard, because the business planned on it.

Then reset each key result in order: was the work done, did it score, and does it still matter next quarter (Figure 8)? A missed goal with the work done means an assumption broke, so name the assumption. A goal that scored well and has become routine moves to the weekly dashboard and frees the slot.

Last, re-run Step 1 from what is left of the year, and send the board the grades, the commitment and the next quarter’s plan on one page. The Marketing KPI Board Report template lays that page out.

What changes by stage?

What mistakes should I avoid?

• Starting from a list of channels. Without the revenue bridge you can’t tell whether the plan is enough (Step 1).
• Leaving churn out of the bridge. The plan is short before the quarter begins (Step 1).
• Several objectives for a team of two. Nothing gets the attention a real bet needs (Step 2).
• Grading the board number like a stretch goal. A 0.7 is fine for a bet and a problem for a commitment (Step 2).
• Targets with no baseline. Nobody can say whether 40 is ambitious or easy (Step 3).
• A hard number for a channel you have never tried. Set a learning goal first (Step 3).
• Activities as key results. “Publish 12 posts” can be done perfectly and change nothing (Step 4).
• Copying last year’s budget. Cost the plan you have, then check it against the ceiling (Step 5).
• Meetings with no owner and no decision rights. The argument happens in week eight instead (Step 6).
• Carrying a miss forward without saying so. Re-run the bridge and tell the board early (Step 7).