An annual operating plan is the budget your board approves and your team runs against for the year. For a Series A or B company it takes about six weeks and moves in a set sequence: leadership sets the spending envelope from cash on hand, department leads build bottom-up inside it, the two are reconciled, the board approves, and finance tracks variance against the locked plan every month.
The sequence matters more than the spreadsheet. Most plans fail because a company runs only one half of it. Top-down alone produces a number nobody owns. Bottom-up alone produces a number the company cannot afford.
If you are starting from a blank file, there is a working model at the end of this post you can download and fill in.
What is different about planning for 2027
Three facts should shape the plan before you enter a single number.
Rates are going up, not down. The Federal Reserve raised its target range to 3.75% to 4.00% in September 2026, reversing the direction it had projected earlier in the year. The dot plot now puts the range at roughly 4.00% to 4.25% at the end of 2027, with most participants expecting rates to hold or rise rather than fall. Inflation projections moved up alongside it. Plan on the cost of capital staying where it is or climbing, not easing.
Venture dollars are at a record, and that number is misleading. Global venture investment hit roughly $510 billion in the first half of 2026, more than all of 2025. Deal count, though, remains well below the 2021 peak. That combination means concentration, not breadth: more money going into fewer companies. Artificial intelligence took about 86% of US venture dollars in the first half of the year, and two companies alone accounted for something like 43% of global startup funding. If you are not in that category, the headline number does not describe your market.
The Series A bar has moved. Roughly 15% of the 2022 seed cohort raised a Series A within two years, against about 31% of the 2018 cohort. The median gap from seed to Series A has stretched past 600 days. Companies clearing the bar are generally showing $2 million to $3 million in ARR with net revenue retention above 120%.
None of that means you should not build. It means the plan has to assume the next round takes longer, costs more in dilution, and demands more proof than your last one did. Build 24 months of runway into the plan rather than 18, and the rest of the year gets easier to manage.
The six-week sequence
Weeks 1 and 2: leadership sets the envelope. Cash on hand, target runway, total spend ceiling. Weeks 2 through 4: department leads build bottom-up inside a stated target, not in a vacuum. Week 4: reconcile. The bottom-up total will exceed the envelope. Deciding what comes out is the real work. Week 5: board review and approval. Week 6: lock, distribute, and set up variance reporting.
Start in October if you want the plan approved before January. Starting in December means approving a plan in February that the company has already spent two months ignoring.
Step 1: Set the envelope, working backward from cash
Before anyone builds anything, answer three questions.
How much cash do you have, and what is your current burn? If you cannot answer this to the dollar, stop and fix the close first. Our guide to burn rate and runway covers the four ways the standard calculation misleads.
What runway do you need to protect? Pick the minimum number of months you will not go below, and treat it as a constraint rather than an output. Given the funding data above, 18 months is thin for a company that will need to raise. 24 is the more defensible number for 2027.
What does that leave you to spend? Cash on hand, minus the runway reserve, divided by the months in the plan, adjusted for the revenue you expect to collect. That is the envelope. It is a rough number and it is supposed to be.
Now translate it into rough department targets and hand those to your leads. Not a blank page, and not a fixed allocation either. A target they are expected to build against and push back on.
Step 2: Build headcount before anything else
Payroll is 60% to 75% of operating spend at most Series A and B companies, so headcount planning is the plan. Everything else is rounding by comparison.
Three rules make the difference between a headcount plan that survives the year and one that does not.
Load the salary properly. Base salary is roughly 75% to 80% of what a person actually costs. On top of base you carry employer FICA at 7.65%, federal unemployment, state unemployment that varies by state and experience rating, benefits that commonly run $900 to $1,400 per employee per month, and bonus or variable comp. A plan built on base salary alone runs over every single month, and the gap compounds. If you are hiring across state lines, our multi-state employment compliance guide covers the registrations that come with it.
Model start months, not annual salaries. An engineer at $180,000 starting in July costs $90,000 of base in the plan year, not $180,000. Plans that load full-year salaries for mid-year hires overstate spend badly and then get quietly revised, which destroys the credibility of the whole document.
Tie each hire to a trigger, not a date. A start month in a spreadsheet is a guess. Write down what has to be true before the role is approved: a revenue threshold, a product milestone, a customer count, a closed round. Then a bad quarter means you delay three hires deliberately instead of scrambling. This is the single highest-leverage thing in the entire plan, and almost nobody does it.
Add recruiting costs while you are here. Agency fees run 20% to 25% of first-year salary, and a plan with 8 hires and no recruiting budget is short a real number.
Step 3: Build the revenue plan from the bottom up
Take your current ARR, then build forward with four separate lines: new bookings, expansion, contraction, and churn. One blended growth rate is not a plan, it is a wish, and no board member will accept it.
New bookings should tie to the sales capacity you just built in the headcount plan. Number of reps, ramp time, quota, and expected attainment. If the revenue plan assumes bookings that would require 3 more reps than the headcount plan funds, you have found a problem worth finding in October rather than April.
Expansion and churn come from your actual cohort data. If your net revenue retention has run at 105% for the past year, a plan built on 130% needs a specific reason. Our piece on the SaaS metrics that matter at Series B covers definitions and benchmark ranges.
Then pressure-test it. If revenue comes in 30% under this plan, what happens to runway, and what would you cut? Write the answer down now. That is the downside case, and building it in October is free. Building it in June, under pressure, is not.
Step 4: Non-payroll spend, by cost center
Everything that is not people: software, hosting, rent, legal, accounting, insurance, marketing programs, travel, recruiting fees.
Build it by cost center rather than by category, so each line has an owner. Two categories deserve specific attention because they are consistently underestimated.
Cloud and infrastructure scale with usage, not with your budget. Model it as a percentage of revenue and check it against what you actually spent over the past six months.
G&A creeps. Legal, accounting, audit, insurance, and software renewals all rise as the company grows, and D&O premiums in particular jump when you add an institutional board. Our G&A budget guide covers what a realistic startup budget allocates to G&A by stage, and where the line usually breaks.
Step 5: Reconcile, which is the part that matters
The bottom-up total will come in over the envelope. In our experience it lands 20% to 40% high, and that is normal rather than a sign anyone did the work badly. Every manager optimized for their own function and none of them owned the total.
The reconciliation meeting is where a budget stops being a spreadsheet and becomes a set of decisions. Run it with the whole leadership team in the room and work in this order:
- Confirm the envelope is real. If protecting 24 months of runway is genuinely a constraint, say so plainly and do not negotiate it in the meeting.
- Rank every incremental hire and program against the plan’s goals. Not by department. One list, ranked.
- Draw the line where the envelope runs out. Everything above the line is funded. Everything below is explicitly deferred, with the trigger that would bring it back.
- Write down what was cut and why. This document is worth more than the budget itself six months from now, when someone asks why marketing never got the headcount.
The temptation is to split the difference and shave every department by 15%. Resist it. Across-the-board cuts fund your lowest-priority work at the expense of your highest, and they teach every manager to pad next year’s submission.
Step 6: Take it to the board
Boards approve the annual operating plan, usually in the Q4 meeting. Send the package a week ahead and bring five things:
- The plan itself: revenue, spend by department, headcount, ending cash and runway
- The prior year’s plan versus actual, so they can judge how good your forecasting has been
- The assumptions, stated explicitly, with the two or three that most affect the outcome flagged
- The downside case and the specific triggers that would put it into effect
- The list of what you cut and why
That last one is the item most founders leave out, and it is the one that builds the most credibility. A board that sees you made hard choices deliberately trusts the number in front of them. A plan that funds everything looks like a plan nobody stress-tested.
Our guide to what goes in a startup board reporting package covers the standing format.
Step 7: Lock it, then track variance monthly
Once the board approves, freeze it. The approved plan is the baseline for the year and it does not get quietly edited to match actuals.
Every month, report plan versus actual by department with a one-sentence explanation for each material variance, favorable and unfavorable. Explaining only the misses signals that the beats were luck.
Re-forecast quarterly, and keep the re-forecast visible alongside the original plan rather than replacing it. The board should be able to see both what you committed to and where you now expect to land. Those are different questions and they both matter.
If your close is not reliable enough to produce this within a week of month end, that is the thing to fix before the next planning cycle. Our Series A finance stack covers the infrastructure that makes monthly variance reporting routine rather than a fire drill.
Download the 2027 budget model
We built the model we use with clients into a spreadsheet you can download and fill in. No email required.
It has eight working tabs: assumptions, a headcount plan that computes fully loaded cost per person and flows it to the right cost center, a revenue build, non-payroll operating expenses, and a linked P&L, balance sheet, and statement of cash flows with monthly burn and runway at the bottom. The employer tax rates are broken out so you can set your own state unemployment rate. The balance sheet carries a check row that has to read zero.
To run a downside case, save a second copy, cut the new ARR line by 30%, and read the runway row.
Download the 2027 startup budget model (Excel)
Common questions
When should we start the annual operating plan?
October, for a plan approved before January. The build takes about six weeks for a Series A or B company, and December disappears faster than anyone expects.
How detailed should a startup operating plan be?
Monthly, by department, with headcount modeled person by person. Below that level the detail costs more than it produces. Above it, nobody maintains the model past February.
Should the plan be top-down or bottom-up?
Both. Leadership sets the envelope, departments build inside it, and the two get reconciled. Running only one produces the two classic failures: a number nobody owns, or a number the company cannot afford.
What runway should we plan for in 2027?
24 months is the defensible target for a company that will need to raise. The gap between seed and Series A now runs past 600 days at the median, and capital is concentrated in a narrow set of companies. Planning on 18 months assumes a market that is not the one in front of you.
Who builds the operating plan?
Finance owns the model and the reconciliation. Department leads own their own numbers. The CEO owns the envelope and the final trade-offs. At most Series A and B companies that finance role is a controller or fractional CFO rather than the founder, and our comparison of fractional CFO vs. controller vs. VP Finance covers where the line falls.
What if we are pre-revenue?
The sequence is the same, but the envelope does all the work. With no revenue to offset spend, runway is purely a function of cash and burn, and the hiring triggers should tie to product and customer milestones rather than revenue thresholds.
Building your 2027 plan and want a second set of eyes on the assumptions before it goes to your board? Start a conversation.