How to Write a Memo: 10 Steps That Drive Decisions

Most memos die on the vine because they’re either vague, too long, or missing a decision. Here’s a battle-tested, 10‑step memo process I use to move teams from argument to action without adding meetings or headcount.

Slack is great for nudges and noise. It’s terrible for decisions. A good memo forces the author to think, gives stakeholders the full context, and leaves a durable trail of what was decided and why. If you want speed, start with clarity, not more channels.

Memos scale judgment. They compress research, trade‑offs, and next steps into a single artifact anyone can read in ten minutes. That’s how you avoid two months of drift and a last‑minute “can we push to next sprint?”

And unlike live meetings, a memo respects async work across time zones and functions. Sales can comment at 7 a.m., Engineering at noon, Finance at 6 p.m. The document converges. People align. Deadlines survive.

Start with the decision. Not the background, not the drama. Write a literal one‑liner at the top of your draft: “Decision: Approve $120K budget to rebuild onboarding flow by Q2 to increase activation from 24% → 40%.” If you can’t write that one line, you’re not ready to write the memo.

Assign an owner. “Owner” means accountable human, not committee. Add their name and role next to the decision line. This cuts future confusion, especially when multiple teams contribute.

Define success in the same breath. Tie it to one or two metrics. Example: activation rate, payback period, churn reduction, lead velocity. The tighter you frame the decision, the less room there is for scope creep later.

Your audience dictates your tone and level of detail. If you’re writing to the exec team, keep it sharp and decision‑ready. If you’re writing for the revenue org, include customer quotes and pipeline impact. If it’s a cross‑functional build, include resource asks and milestone risks.

List stakeholders early in the memo. Not as a CC dump, but as “Impacted Groups” with 1–2 lines about how the decision touches their world. Now each group knows why they should care and where to focus their comments.

Call out who must review and who is FYI. Diffusing decision rights is the fastest way to stall. Get this wrong and you’ll watch your memo turn into a 47‑comment thread about naming conventions.

After the decision line, add one sentence that explains the purpose of the memo. This is your North Star when the debate starts to sprawl. Example: “This memo proposes a focused rebuild of onboarding to raise activation and reduce support load, based on user interviews and funnel data.”

Purpose is not a restatement of the decision. It’s the why, the forcing function that keeps reviewers from asking for unrelated features and pet projects.

If you can’t write a crisp purpose, pause. Go talk to customers, peek at the dashboard, or interview support. You’re still guessing.

Good memos surface constraints before they bite you in sprint three. Add a “Facts & Constraints” section: what the data shows, what’s blocked, what’s non‑negotiable (compliance, legal, SLA), and any budget or team limits.

Lay out options, not just your favorite one. Three is ideal. For each option, list expected impact, cost (money + time + opportunity), and risks. Keep it skimmable: bullets and short paragraphs beat walls of text.

Tie options to strategy. If your decision shapes market focus, positioning, or ICP, reference your go‑to‑market choices and show how the options align. If you need to level up this part of the work, here’s where a structured GTM strategy helps you avoid whiplash between “enterprise” on Monday and “PLG” on Friday.

Pick an option. Say it plainly: “Recommendation: Option B.” Then explain why, using the facts you just laid out. Resist the urge to re‑litigate everything; focus on trade‑offs, not perfection.

Put your neck on the line with an explicit bet statement: “We believe rebuilding just steps 2–4 (not the whole flow) will lift activation by 15–20% within 60 days because interviews show confusion at these two checkpoints.” Bets create learning loops even if the outcome underperforms.

Flag what you are intentionally not doing. This protects scope and preserves trust: “We are not touching billing logic this quarter.” Silence invites derailment.

Write the summary after you’ve squared away rationale, risks, and plan. One short paragraph and a bullet list is enough: decision, why, impact, cost, timeline, owner. Make it digestible for a busy exec scanning on a phone.

Front‑load numbers and dates. “+$220K ARR in 90 days” beats “We think this is strategic.” Put a single sentence that captures the consequence of saying no. That frames trade‑offs and prevents endless deferrals.

Remove adjectives in the summary. You’re not selling; you’re guiding a decision. Save any color for the body.

Use clear sub‑heads and short paragraphs. Think in scenes, not a stream of consciousness. Reviewers should find “Customer Insights,” “Funnel Data,” “Options,” “Risks,” and “Plan” in predictable places.

Bullet lists are your friend, but not bullet soup. If every paragraph is a list, you’ve dodged synthesis. Synthesis is why the memo exists. The job is to reduce the world, not copy‑paste it.

Limit exhibits to what’s directly relevant: a funnel chart, a user journey, a P&L snippet. Link to deeper docs for the curious. If you need to show CRM fields, definitions, and handoffs, reference your revops source of truth and keep the memo focused. If that source of truth doesn’t exist, fix your foundation with a real CRM implementation before you argue about MQL definitions.

Write the pre‑mortem: “It’s 90 days later. This failed. What happened?” List the three likeliest failure modes. Now say how you’ll prevent or detect each one early. This shows you’re not naive about execution risk.

Document your rollback plan and kill criteria. “If activation doesn’t move after cohort 3, stop work and re‑interview users.” Or, “If CAC rises 30% after two weeks, pause creative set X.” Decisions feel safer when exit ramps are visible.

Clarify dependencies and sequencing. Risks love ambiguity. Spell out which teams need to ship first, what assets must be ready, and what “done” means for each step.

Line up a quick red team: two or three people who disagree with you or see the world differently. Ask them to find holes, not to rewrite. Time‑box their pass to 24–48 hours. Incorporate what matters; ignore the bikeshedding.

Define your approval path in the memo footer: “Approvals required: VP Product, CFO. Review: Sales, Support (FYI).” That simple line cuts three days of Slack ping‑pong. Put deadlines next to names so nobody is surprised by launch dates.

Version control matters. Put a version and date at the top. If you update the recommendation after feedback, note the change. This prevents the “which memo are we talking about?” spiral.

Pick one system of record (Docs, Notion, Confluence) and link it in Slack/Email with a short note and a deadline. Don’t paste your whole memo in chat. You will lose context, comments, and the thread of decisions.

Use a follow‑up cadence that fits the decision: 24–48 hours for review, a decision date, and a checkpoint date. Add the checkpoint to your calendar and the team’s. “Silence means consent” only works if you set the rules upfront.

For go‑to‑market memos (campaigns, outbound), treat the memo as the master brief and attach playbooks as appendices. When sales asks, “What are we sending on day three?”, they shouldn’t need to assemble it themselves. If outbound is the decision at hand, tie the memo to a concrete sending plan and infrastructure, not vibes. If you want a production‑grade approach, build from a proven cold email outbound system.

You don’t need fancy templates. You need consistency. Here’s a light structure you can copy into your doc tool of choice. The goal is a 2–4 page readable document, not a novel.

Keep headings predictable across teams. That way Finance knows where to look for cost, Product knows where to look for trade‑offs, and Sales can skim for the impact on pipeline and messaging.

Use this skeleton and then tailor depth to the decision size:

• Title + Version + Date
• Decision (one line) + Owner + Success metric
• Executive Summary (one paragraph + bullet list)
• Purpose: why this memo exists
• Facts & Constraints: the data, the limits
• Options: 2–3 choices with impact, cost, risks
• Recommendation: chosen option + rationale + non‑goes
• Plan: milestones, owners, resources, cost
• Risks & Pre‑mortem: top 3, mitigations
• Rollback & Kill Criteria
• Approvals & Reviewers + deadlines
• Appendix: charts, interviews, deeper docs

Quick context: I run Anton Chisliuk · AI Marketing Ops - an AI‑first outsourced marketing operator model. One senior operator ships GTM, SEO content, outbound, and automation in a 90‑day build cycle. Memos are how I keep cross‑functional work aligned without adding meetings. For example, with Cleverence (B2B SaaS) we scaled from 0 to ~80K organic visits/month and ~2K inbound leads/month after memo‑driven decisions on ICP, content architecture, and workflows. If you’re weighing whether the operator model beats a classic agency layer for your stage, read a short overview about the company and how the economics work.

Example 1 - Product activation: Decision to rebuild onboarding steps 2–4 to lift activation from 24% → 40% in 60 days. Facts: 57% of drop‑off at ID verification; 41% confusion on plan selection. Options: quick copy/design pass vs. rebuild flow vs. defer. Recommendation: rebuild steps 2–4, keep billing untouched, run A/B on copy. Risks: identity provider SLAs; mitigation: shadow traffic with parallel integration. Result: decision in 48 hours; scope locked; sprint unblocked.

Example 2 - Paid acquisition: Decision to shift 40% of Meta budget to Google Search for higher intent. Facts: Meta CPL $18 with 1.1% SQL rate; Google CPL $34 with 6.3% SQL rate; blended CAC rising 19%. Options: hold, shift 20%, shift 40%. Recommendation: shift 40% with 2‑week checkpoint on ROAS and SQL rate. Risks: limited search volume; mitigation: expand match types and add 3 new SKUs. Result: exec approval same day; team moves budget Monday.

Example 3 - Outbound reboot: Decision to rebuild targeting and messaging for mid‑market ops leaders; goal: 20 meetings in 30 days. Facts: current reply rate 0.4%; ICP unclear. Options: keep blasting, narrow ICP + rewrite, pause outbound. Recommendation: narrow to two pain clusters, add case‑led intros, stagger sending 600/day with warmup. Risks: domain health; mitigation: new domains + gradual ramp. Result: measurable plan with clear KPIs, not wishful thinking.

Mistake: Treating memos like status updates. Fix: frame a decision, not a diary. If there’s no decision, write a one‑pager “research note,” not a memo.

Mistake: Hiding trade‑offs. Fix: show options and what you are not doing. Stakeholders trust decisions that reveal what you gave up.

Mistake: Writing for yourself. Fix: tailor depth to the audience. Finance cares about payback; Sales cares about messaging and pipeline; Support cares about ticket deflection and timelines.

Process metrics: decision made by date X, number of reviewers and time to approve, number of comments resolved. These prove the memo actually moved the organization.

Outcome metrics: the one or two success metrics you set in Step 1. For commercial work, that’s usually pipeline, win rate, activation, CAC, or LTV. For product, task completion or adoption. For ops, cycle time or error rate.

Learning metrics: even when wrong, a good memo tightens your loop. Did you capture why the bet missed? Did you close the loop in the doc with what you learned? If you’re not versioning and annotating outcomes, you’re just writing pretty prose.

AI can speed the grunt work: summarizing interviews, drafting option sets, cleaning data. It should not decide your trade‑offs. Use it to generate candidate outlines, tighten language, and surface alternative framings you might miss under time pressure.

Feed AI structured inputs: ICP, goals, constraints, raw notes. Then make the human call on scope, sequencing, and kill criteria. That’s the difference between a memo that reads fine and a memo that ships.

Final pass is always yours. Read it out loud. If a sentence makes you wince, cut it. If a claim lacks a number, add a number or delete the claim.

Cadence beats heroics. Slot one memo per week into your calendar: Monday draft, Tuesday red team, Wednesday decision, Thursday kickoff. Smaller decisions can share a memo; large ones deserve their own artifact.

Use memos to gate work. No memo, no build. It sounds rigid until you see your backlog halve and your cycle times fall. Teams like finishing things more than they like guessing at priorities.

Close the loop every Friday: annotate outcomes in the memo header (green/yellow/red), log what changed, and link to the next decision. Memory compounds when your docs talk to each other.

Memo: non‑reversible or expensive decisions, cross‑functional work, anything that benefits from slow thinking and an audit trail. Use when you need commentable text and a single source of truth.

Slides: external storytelling and executive updates where visuals carry the point. Use sparingly for internal decisions; decks often hide the footnotes where the real debates live.

Loom/video: demos, walkthroughs, onboarding. Great as an appendix to the memo, not a replacement. Video isn’t searchable, and searchability is what makes your org smarter over time.

A strong memo is a decision engine. It clarifies the ask, names the owner, exposes trade‑offs, and sets a clear plan with risks and rollback. Do this consistently and you’ll cut meetings, shrink cycle times, and raise your hit rate on bets.

If you want outside eyes on your growth levers and the decisions blocking them, book a free Growth Audit. We’ll map one or two high‑impact decisions you can lock in within a week, with data, risks, and a simple rollout plan.

Written by Anton Chisliuk - operator, not agency. Leads, sales, content, and systems - one senior operator.

- How long should a business memo be?

Two to four pages is the sweet spot for most decisions. If you need more, push detail into an appendix. If you need less, it might not be a memo decision - it might be a quick note or a task.

- What’s the difference between a memo and a brief?

A memo frames a decision with options, trade‑offs, and a recommendation. A brief turns an approved decision into an execution plan (creative, channels, timelines). Sometimes they live in the same doc; just keep the sections distinct.

- Who should write the memo: manager or IC?

The person closest to the work should draft; the person accountable for the outcome should own the decision line and recommendation. Often that’s an IC drafting with a manager shaping and signing.

- How do I get execs to actually read the memo?

Lead with a tight summary, put the decision date in the subject, and set explicit expectations for approval. Add a 10‑minute read time at the top. Busy people respect documents that respect their time.

- Should I include financials in every memo?

Include a lightweight model whenever money or headcount moves. Even a simple payback and sensitivity range beats hand‑waving. For purely technical or UX decisions, tie to the nearest commercial proxy (activation, support load, churn).

- Can I use AI to draft the memo?

Yes - for outlines, first passes, and clarity edits. Don’t outsource trade‑offs, risk calls, or kill criteria. Use AI to save time, not to abdicate judgment.