Net New In Year
Next year’s new accounts four ways: Straight Line, Seasonal, Cohort and Power Law, each reforecast as the months close.
Chapter 1 of 3 · 18 min read · A guide by Broad Eye Solutions · Version 1.0 · Updated November 14, 2026
You’ve just been made responsible for next year’s forecast. The number is due before the year closes, and everyone from the board to the sales floor will be held to it. How do you build it?
This chapter starts with net new in year: the revenue from accounts approved during the year, in the months they bill that year. It builds next year four ways. Each section is a complete forecast you can build on its own, and each asks more of your data than the one before.
- Straight Line. The year’s number from this year’s totals and a goal, drawn as a straight line through the twelve months.
- Seasonal. The same number, with each month carrying its usual share of the year.
- Cohort. The year built from the accounts that make it: each month’s new accounts, followed as they bill.
- Power Law. The cohort build, split by the size of the account.
Each section ends with a Reforecast: the same build made again once the year starts, at the 3+9 (three months closed, nine forecast), the 6+6, the 9+3 and every month. The book is a payments portfolio, and the numbers are sample data.
Here’s what you’ll be able to build by the end: next year by month, four ways. Switch between the methods under the title, and hover over a month for its new accounts and cost per new account.
Straight Line
Start with the simplest, fastest way there is: a forecast you can put together in an afternoon and defend in any room.
Starts from: this year’s totals and a goal.
What You’ll Need
Four numbers, for two years, month by month. The prior year (PY) in full, and the year you’re in (IY) as far as it has closed. Finance and the CRM can hand over every cell today. Revenue here, and all through the guide, is what each account billed in the month, as on its statement.
Every number in those two grids gets used. Each row also needs its total for the year. PY’s is on the books. IY’s revenue and new accounts are the months so far divided by PY’s share of those same months. Its leads and investment are that count at this year’s rates so far. One more number isn’t in any system: the goal. Here it’s 10% more revenue from new accounts. The investment and lead figures are illustrative.
Step 1: Start With The Number
The plan for next year’s new business starts at the end. Before anyone counts a lead or a rep, the company picks the revenue it wants from new accounts. Everything else gets built to reach it.
On this book, this year’s new accounts will bill about $3.76M before the year closes. The goal for next year is 10% more, $4.14M.
Then draw next year (NY) as a straight line. This year’s January, $21.8K, was 0.58% of this year. NY’s January starts at that share of the target, $24.0K. Every month after adds the same amount, $58.4K, so December lands at $665.9K and the twelve months add up to the target.
A straight line has no curves in it. It can’t tell one month from another except by how far into the year it is. That’s where the next section, Seasonal, comes in.
Nothing is wrong with that. A board can sign off on it, and every team can repeat it. It’s also the last time the plan looks at revenue directly. From here on, each step turns the number into something else, and each one does it with an average.
Step 2: Work Backwards To The Count
How many new accounts is $4.14M? The forecaster reaches for this year’s average. This year 3,268 new accounts brought in $3.76M, about $1,150 each inside the year. Hold that average, and 10% more revenue takes 10% more accounts: 3,595.
Sales and marketing now have a count. The count is what gets tracked every month, because it’s the easiest thing in the plan to see.
Step 3: Price It
The count gets a budget the same way, from this year’s averages. In this example the portfolio spent $1.63M to sign its 3,268 accounts, $500 each, blended across every channel. It took 81,700 leads, and 4.0% of them became accounts. That’s the conversion rate, CVR in the grids. (These spend figures are illustrative.)
At the same rates, 3,595 accounts cost $1.80M. The budget goes up 10% with the goal.
This is where the plan gets negotiated. Nobody wants to ask for 10% more budget, so someone changes a rate.
Lift the conversion rate from 4.0% to 4.4%, and the same 81,700 leads bring in all 3,595 accounts. The goal is met on this year’s budget. On paper. The plan now rests on a rate nobody has hit yet. This chapter carries Plan 1 forward, the goal with nothing else changed. The aside at the end of the chapter shows when a better rate shows up in the months, and what that does to the year.
Step 4: Put Each Plan On The Line
Every plan from Step 3 goes on Step 1’s straight line, with its own total: $4.14M for Plans 1 to 3, this year’s revenue for Plan 4. Switch between the plans under the title.
The second switch sets how a plan’s change arrives. Evenly, it’s there from January: the same leads, accounts and spend every month. Weighted, nobody expects a new rate on January 1, so it’s phased in: 10% under the plan’s rate in January, rising in equal steps to 10% over in December. Plan 1 changes nothing, so the switch leaves it alone. Weighted, Plan 2’s better conversion tilts its accounts up through the year, and Plans 3 and 4’s cheaper leads tilt spend down. Every line stays straight, and every plan still adds up to its Step 3 totals.
Straight Line counts every account at this year’s average, so Plan 2 still lands on $4.14M either way. The aside at the end of the chapter shows what that misses. This chapter carries Plan 1 forward: $4.14M, on the straight line.
Reforecast
Once the year starts, each close turns a forecast month into a fact, and the forecast gets built again with what has closed. Most teams do it three times: at the 3+9, the 6+6 and the 9+3. More and more do it every month.
Straight Line reforecasts the way it was built. Last year, the months closed so far were a known share of the year. Divide this year’s months closed by that share, and that’s the year. Take away what has closed, and put the rest on the straight line through the months still open: January’s share of this year, rising by the same step each month. Here’s 2026, the year in progress, as it stood at each close.
What This Build Can’t Tell You
A straight line fits new business better than it sounds. An account approved in January bills all year, and one approved in December bills once, so net new revenue starts near nothing and builds as the accounts add up: close to a straight climb. What the line can’t see is the time of year. The same account bills less in February than in December, and a line drawn from January’s share has no way to know it.
Seasonal
Same year, same count, same budget. What changes is which months carry it.
Starts from: Straight Line’s year, $4.14M, with its count and its budget.
What You’ll Need
- Each month’s net new revenue, back as many years as you have. The more years, the steadier the read.
Step 1: Read The Shape Of One Year
A careful forecaster starts from this year, month by month, and reads each month as a share of the year. January was 0.6% of it. December will be about 15.4%. Copy every share into next year, and only the size of each month moves.
Read down a month and the dollars change. Read along this year’s share row and the plan’s, and they match, month for month. The first quarter is 5% of the year in both. The last quarter is 42%.
Step 2: Read Every Year Together
One year’s shape can carry one year’s accidents. Line up every year you have, every January with every other January, and so on through December, and take the average. Then fill Straight Line’s year into the months by the average share.
On this book the shares hold. No month moves more than 2.4 points across seven years. Most of what the grid shows is new accounts adding up through the year, so one year’s shares carry most of it. The average is steadier, and it’s what this section uses.
The grid closes this year out at $3,896.2K, from how every earlier year finished. That number is only the yardstick for this year’s shares. Straight Line closed the year out at $3,763.4K from last year’s shares, and Seasonal keeps that number: the year stays the same, and only the months move.
That’s Seasonal’s next year: the same $4.14M as Straight Line, with each month carrying its usual share.
Reforecast
The same year as Straight Line at each close: the months closed, divided by last year’s share of them. The months still open each take their average share of the year, not the straight line’s share.
What This Build Can’t Tell You
The months now have the right shape, but the year is still built from an average account: this year’s revenue per new account, times a count. Revenue per account, cost per account, the conversion rate, each month’s share of the year: four averages, multiplied together, and each one blends very different accounts. In a book with a power-law shape, a minority of the accounts brings in a majority of the revenue, so each average moves whenever the mix moves. The plan can’t say which accounts it’s counting on.
The shape itself mostly comes from accounts adding up through the year. Build the year from those accounts, and the shape comes with it.
Cohort
This section rebuilds the same year from what it’s made of. Each month’s new accounts go on their own line and get followed, month by month, as they bill. It takes about a day in a spreadsheet.
Starts from: Straight Line’s count, 3,595 new accounts.
What You’ll Need
- The month each account was approved.
- What each account billed in every month since: one row per account per month.
- About two years of it, so the time of year can be read.
The change from Straight Line and Seasonal is in the revenue. They needed one date on it: the month it was billed. Cohort needs two: the month the account was approved, and the month it billed. Here are last year’s first two approval months laid out that way.
Every row is one month’s new accounts. Every column is a month they billed. Without both dates on the revenue, the steps below can’t be built.
Step 1: Put Both Years On The Grid
Start with last year, PY. Group the revenue by approval month, one line for each, and follow each line through the months it bills.
Read the bottom row first. It’s the PY row Straight Line started from, the one the averages came from. Every row above it is what that one row averaged together. January’s 251 accounts billed $611K inside the year. December’s 278 billed $32K.
Do the same for the year you’re in, as far as it has closed.
Now the two years sit side by side, line for line. Take the first line. This year 273 accounts were approved in January, and they billed $405K through October. Last year’s 251 January accounts had billed $517K by the same point. More accounts, less revenue. Straight Line had one line for both.
Step 2: Read The Age Curve
Divide each line by its accounts, then slide it left so every line starts at its first month on the book. Every cohort now reads on the same scale.
A new account on this book bills about $110 in the month it’s approved, because most start partway through the month. By its third month it bills about $218. Then it eases. The bottom row is the curve: the last twelve approval months read together. Add it across and a new account bills about $2,200 in its first twelve months.
The curve replaces Straight Line’s single average. Straight Line said a new account is worth about $1,150 inside the year. The curve says it depends on when the account arrives. A January account gets twelve months of the curve. A December account gets one.
Four per-account numbers appear in this chapter, and each answers a different question. About $1,150 is what this year’s average new account bills inside the year (Straight Line, Step 2). $1,280 is last year’s (Power Law). $1,229 is next year’s on this curve (Step 4). About $2,200 is a new account’s first twelve months, wherever they fall.
Step 3: Read The Time Of Year
Age isn’t the only pattern. The same account bills differently in February than in December. To see it, compare what every cohort billed in each calendar month with what its age alone says.
On this book February runs 12% under an average month and December 7% over. Seasonal’s grid can’t show this, because inside one year every calendar month is also a different mix of ages. Two years of cohorts pull the two apart.
Step 4: Build Next Year
Next year is Straight Line’s count, placed by approval month, times the curve, times the time of year. Each cell is one multiplication.
These are the same 3,595 accounts. The cohort build puts them at $4.42M. Straight Line and Seasonal said $4.14M.
Most of the gap is a choice the averages made without showing it. Straight Line took this year’s average account, and this year’s average carries a weak start: this year’s January accounts have billed less each than last year’s had by the same point (Step 1). The cohort build prices next year’s accounts off the last twelve approval months instead, $1,229 each inside the year.
Neither is a fact. One assumption is hidden inside an average. The other is a line on a grid that anyone in the room can argue with. If the curve holds, the $4.14M target takes 3,369 accounts, not 3,595.
Reforecast
At each close, the accounts approved so far run forward on the age curve at their own level, read from what they’ve billed. The accounts not approved yet still rest on the plan’s count. Straight Line and Seasonal scale the months closed up by last year’s share of them instead, which is why the two start far apart at the March close.
What This Build Can’t Tell You
It still can’t say which accounts. January’s line is 273 accounts averaged together, and on this book they are nothing alike. Some bill $30 a month and some bill $1,900. Two cohorts with the same total can be made very differently: every account a little smaller, or a handful of large ones missing. Those are different problems with different owners. And every month still to be approved rests on the plan’s count: if fewer accounts are approved, every month comes in under.
Power Law
Every line on the cohort grid is still a few hundred accounts averaged together. This section splits every line by the size of the account. It’s the step Power Law forecasting is named for.
Starts from: the cohort build, $4.42M.
What You’ll Need
One more column: a tier for every account, cut from its monthly volume. Nothing else: the rest is already on the cohort grid. The cut used here is Power Law Nested Means. Take every account’s monthly volume. Split the book at the average. Then split the accounts above it at their own average, and the accounts below it at theirs. That makes four tiers, A to D, and each account keeps the best tier it has reached.
Here are last year’s new accounts, sorted that way.
The top two tiers are 16% of the new accounts and 72% of the revenue they billed. That’s a power-law shape, and it’s why the average misleads. The average new account billed $1,280 inside the year. An A account billed $13,420. A D account billed $227. No account on the book looks like the average.
Step 1: Sort The New Accounts Into Tiers
That grid is the step. Every account gets a letter, and every line of the cohort grid becomes four.
Step 2: Read A Curve For Each Tier
The cohort build read one age curve for all new accounts. Read it again, once for each tier. Read it from accounts that have had time to settle into their tier: the ones approved November 2024 to October 2025, twelve to twenty-three months ago. A new account’s tier can only rise as it bills, so a young account’s tier says little yet.
An A account bills more in its first month, $821, than a D account bills in its first twelve, $382. A single curve for all of them sits between the two and describes neither.
Step 3: Build The Plan By Tier
The plan was a count: 3,595 new accounts for $4.14M. Put the tiers on it, at last year’s shares.
Now the plan says what it needs. About 135 A accounts carry $1.63M of it. About 1,966 D accounts carry $0.40M. Ten A accounts short is about $121,000, and it takes about 590 more D accounts to make that up. A team can hit its count and miss its year by a few dozen accounts, and a count can’t show it.
Then fill in the months. Nobody knows a new account’s tier before it’s signed, so a year ahead the tiers can’t change the total. Take the cohort build, $4.42M from the same 3,595 accounts, and split each month by tier.
Reforecast
A new account’s tier can only rise as it bills, and it doesn’t settle until the account’s fourth month. From then on it runs on its own tier’s curve, at its own level. Anything younger runs on the cohort curve.
At the March close no new account has got there yet, so Power Law is the cohort build. At the June close the accounts approved in January, February and March have settled, and the two builds part.
What This Build Can’t Tell You
A year ahead, the tiers don’t change the total. The plan says which accounts it needs, tier by tier, but nothing in it makes them sign, and every month still to be approved rests on the plan’s count.
Where net new in year lands you
Four forecasts of the same year, each a complete build. Straight Line and Seasonal share one number and differ in the months. Cohort and Power Law build the year from the accounts, and Power Law splits it by tier.
And 2026, the year in progress, as each build stood at each close:
Here’s the chart from the top of the chapter again. You can now build every bar, four ways.
What This Chapter Can’t Tell You
All of it covers new accounts. On this book that’s about 12% of a year’s revenue. The other 88% comes from accounts already on the book when the year starts. Chapter 2 takes them through the same four sections. And which of the four to believe, and how far ahead, is what Chapter 3 tests.
Aside: when the better rate arrives
Straight Line’s Step 3 left four plans on the table, and a plan that changes a rate has to say when the change shows up. These grids fill each plan into the months by this year’s shape: each month’s part of this year, which Seasonal reads in its Step 1.
A better rate has to be scheduled too. One way is evenly: every month converts 10% better than it did this year. The other is weighted, because nobody expects a new rate to show up on January 1. January starts 10% under the goal rate, the rate climbs in equal steps, and December finishes 10% over it. Across the year it averages out to the goal.
Both plans sign the same number of accounts, within a handful. They don’t bring in the same revenue. Evenly, the plan lands on its $4.14M. Weighted, it lands on $3.99M, about $150K short. That’s growth of 5.9%, not 10%, from a team that hit its count and its average rate.
Open Plan 2, the plan with the better CVR, and flip the switch under the grid.
The reason is when the accounts arrive. On this book, an account approved in January billed $2,436 inside the year. A June account billed $1,217. A December account billed $115. The first quarter’s approvals were 26% of the year’s new accounts and 45% of its net new revenue. The weighted plan puts its weakest months where an account is worth the most.
Plans 3 and 4 change the cost per lead, and the switch schedules that the same way. Evenly, every lead costs $18 from January. Weighted, January pays 10% over that and December 10% under. The leads, the accounts and the revenue don’t move, and the year’s budget barely does. What moves is when the saving shows up. On Plan 3, the first half costs $45.1K more than the even plan said it would.