Marketing Plan

You turn a business goal into a plan the team can execute and the owner can hold you to: numbers, channels, budget, calendar and the measures that show whether it is working. A good plan is mostly arithmetic and choices. It says what you will not do as clearly as what you will, and every target traces back to the revenue goal.

When to use

• "Write our marketing plan for next quarter" or "for next year".
• "We have this much budget. Where should it go?"
• "How many leads do we need to hit the sales target?"
• "Marketing feels random. What should we focus on?"
• At the start of a quarter, or when a target or budget changes mid-year.

What you need

• The business goal: revenue, new customers or another target, with the period. Ask the owner if it is not written down.
• Unit economics: average order value or contract value, gross margin, repeat rate or churn, sales cycle length.
• Funnel history for the last 6 to 12 months: traffic, leads, qualified leads, opportunities, customers, by channel. Sources: Google Analytics (connected app or the owner's signed-in browser), the CRM (HubSpot, Salesforce, Pipedrive), the store or billing system (Stripe), the email tool (Mailchimp). If none are available, ask for exports via clarify and say which numbers are estimates.
• Last period's spend by channel, and the budget for the coming period.
• Positioning and best-fit customer. If none exists, flag it and use the owner's best description (see also positioning-and-messaging).
• Team capacity: who writes, designs, runs ads, and how many hours each has.

Method

1. Situation in one page. What worked last period, with numbers; what did not; the main competitors and what they are doing (public sources only); constraints (people, budget, seasonality). Use a simple SWOT only if it produces decisions.
2. Work the funnel backwards from the target. Use real conversion rates from history, not industry averages. Example for a sales-led business:
   • New revenue target / average deal value = customers needed.
   • Customers / close rate = opportunities needed.
   • Opportunities / lead-to-opportunity rate = qualified leads needed.
   • Qualified leads / visitor-to-lead rate = visitors needed. For an online shop: revenue / AOV = orders; orders / conversion rate = sessions. Compute it with execute_code and show the formula. Template in references/plan-template.md.
3. Find the gap. Compare required volumes with the current run rate per channel. The gap tells you whether the plan needs more traffic, better conversion, higher order value or better retention. Often a conversion fix is cheaper than more traffic: a lift from 2 to 2.5 percent conversion is 25 percent more customers at the same spend.
4. Choose objectives. Three to five, each with a number, a date and an owner. Mix lagging measures (revenue, customers) with leading ones (qualified leads, trial starts, organic clicks).
5. Pick channels on evidence. Score each candidate channel on: where best-fit customers already spend attention, past cost per customer, time to results (paid search is days, SEO is months), and whether the team can run it well. Keep the number of channels small; two done well beat six done thinly.
6. Split the budget 70/20/10. Roughly 70 percent to channels with proven returns, 20 percent to promising ones being scaled, 10 percent to experiments. Check every paid channel against unit economics: cost per acquired customer must sit below the margin you earn from that customer within an acceptable payback (see references/plan-template.md). If the owner asks what share of revenue to spend, look up a current benchmark with web_search (for example the annual Gartner CMO Spend Survey), cite it, and say it is a reference point, not a rule.
7. Build the calendar. Quarter by month: campaigns, launches, content themes, seasonal moments, and when each experiment starts and is judged. Add dependencies (site changes, creative, tracking).
8. Define measurement. A KPI tree from the revenue target down to channel metrics; where each number comes from; how often it is reviewed (weekly for leading metrics, monthly for the plan); what triggers a change (for example a channel 30 percent behind plan for four weeks).
9. State risks and assumptions. Every conversion rate and cost you assumed, the source, and what happens to the target if it is 20 percent worse. Run that sensitivity with execute_code.
10. Agree and schedule. Present to the owner, adjust, then set a cronjob for the weekly metrics check and a monthly plan review. Share the final plan with brief_team.

Output

A plan document of no more than 4 to 6 pages plus a spreadsheet:

1. Summary: goal, the three biggest bets, budget total, what you will stop doing.
2. Situation (one page).
3. Funnel maths: required volumes by stage, current run rate, gap.
4. Objectives with owners and dates.
5. Channel plan: per channel, the role, the target, the budget, the first three actions.
6. Budget table: proven, scaling, experimental.
7. Calendar by month.
8. Measurement: KPI tree, data sources, review rhythm, triggers.
9. Risks, assumptions and sensitivity. Show the funnel maths and budget split on a show_card.

Checks before you finish

• Every target traces back to the revenue goal through visible arithmetic.
• Conversion rates come from the owner's own data, or are labelled as assumptions with a source.
• Each paid channel's expected cost per customer is below the margin it earns within the stated payback.
• No more channels than the team can staff; hours are accounted for.
• Each objective has a number, a date and an owner.
• The plan says what will not be done this period.
• Benchmarks quoted have a source and date.

Pitfalls

• Top-down wishes. "Double leads" with no funnel maths is not a plan. Show the volumes each stage needs and whether the channel mix can supply them.
• Industry-average conversion rates. They mislead because funnels differ. Use the owner's data and label anything borrowed.
• Spreading thin. Six channels at 10 percent effort each usually lose to two at full effort.
• Ignoring time to results. SEO and content take months; do not count their leads in month one.
• Vanity objectives. Impressions and followers are inputs. Tie objectives to pipeline and revenue.
• No exit criteria. Every experiment needs a date and a threshold for scale, change or stop.
• Committing spend. You plan the budget; the owner approves it. Never commit spend, sign contracts or start paid campaigns without the owner's explicit go-ahead.