> ## Documentation Index
> Fetch the complete documentation index at: https://docs.getnovaplan.com/llms.txt
> Use this file to discover all available pages before exploring further.

# ARR Studio

> Steer ARR with driver levers, top-down targets, activation curves, and churn plans.

ARR Studio turns the revenue forecast into a model you can steer. Instead of typing a number, you pull levers (reps, quota, attainment, churn, targets) and watch ARR redraw. The past stays fixed; only the future responds. When a lever set is the direction you want, you apply it as a scenario to the plan.

<Card title="Watch the demo" icon="circle-play" href="https://getnovaplan.com/demos">
  Steer Your ARR, Live: pull driver levers and watch ARR redraw, rank levers by impact, goal-seek to a target, get Nova lever suggestions, then apply a scenario to the plan.
</Card>

## The Building Blocks

| Block | What it models |
| - | - |
| **Drivers** | Bottom-up capacity by cohort: productive reps times quota times attainment |
| **Reps** | Who counts as a sales rep, pulled from headcount, with per-period overrides |
| **Validity rules** | How drivers compose with pipeline per period |
| **Top-down targets** | Quarterly CARR add targets split by region and stream, distributed to months |
| **Churn plan** | A total annual churn pinned to a base balance, distributed across months |
| **Activation** | The lag from booking CARR to revenue going live (see [Recognized Revenue](/planning/recognized-revenue)) |

## Driver-Based Forecasting

The driver model builds ARR bottom-up. For a subscription cohort, new ARR in a month is productive reps times quota per rep times attainment. Rep counts come from your headcount plan (employees plus net open positions), filtered by a sales-role classifier you define, and ramp curves model the time a new rep takes to reach full productivity.

<Steps>
  <Step title="Define who is a rep">
    Set the sales-role classifier (by job title pattern, department, tag, or explicit list). Pull rep counts from headcount and override per period where needed.
  </Step>

  <Step title="Set cohort assumptions">
    Per segment and product cohort, set quota per rep, attainment, cohort mix, ramp curve, and retention.
  </Step>

  <Step title="Choose how drivers compose with pipeline">
    Per period, pick pipeline-only, drivers-only, or floor-fill (the larger of the two, never the sum).
  </Step>

  <Step title="Preview the cohort output">
    Review the bottom-up cohort table before you commit, then run the forecast.
  </Step>
</Steps>

<Warning>
  Floor-fill takes the larger of pipeline and drivers, not their sum. A rep's quota already implies they will close some of the deals named in their own pipeline, so adding both would double-count those deals.
</Warning>

## Top-Down Targets

When you plan top-down, set a quarterly gross CARR add target, split it across region and stream by percentage, and let the model distribute it across the months of the quarter. You can also enter one global target for a fiscal year, split it by a region-by-stream matrix, distribute it with a quarterly seasonality matrix, and cascade the result into per-quarter targets. The same activation and recognition that apply to bottom-up ARR apply to target-driven ARR, so the two paths land in the same place.

## Churn Plan

A churn plan pins your total annual churn instead of letting it emerge from a rate. Set it as a percent of a base-period ending balance or an absolute dollar amount, choose a basis (CARR or Live ARR), and distribute it linearly or by weights across twelve months. Weighted distribution normalizes so the annual total is always preserved. A churn plan suppresses the per-period rate-on-balance churn for the periods it covers, so the two never stack.

## Levers, Impact, and Goal-Seek

ARR Studio lets you pull a lever and watch the ARR trajectory redraw live, rank levers by how much they move the number, and goal-seek (ask "what lever value lands me on this target?"). Nova can suggest the levers most likely to close a gap. When a lever set is the plan you want, apply it as a [scenario](/planning/scenarios) so it layers on the base without disturbing it.

## Common Questions

<AccordionGroup>
  <Accordion title="Where do rep counts come from?">
    From your headcount plan. Active employees plus net open positions that match your sales-role classifier, counted direction-aware so a backfill does not double-count. You can override the count per period.
  </Accordion>

  <Accordion title="What does a ramp curve do?">
    It models the share of full productivity a new rep reaches each month after starting, so a rep hired in March does not contribute full quota in March.
  </Accordion>

  <Accordion title="Can I plan bottom-up and top-down in the same year?">
    Yes. Validity rules are per period, so one quarter can be driver-based and another target-only or floor-fill. The activation and recognition layers treat all of them the same downstream.
  </Accordion>

  <Accordion title="If I pull a lever, does it change my committed plan?">
    Not until you apply it. Steering is exploratory. You commit by applying the lever set as a scenario, which keeps your base intact for comparison.
  </Accordion>

  <Accordion title="My churn plan and a churn rate both seem to apply. Will they double-count?">
    No. A churn plan suppresses the per-period rate churn for the periods it covers, so total churn equals what the plan specifies.
  </Accordion>
</AccordionGroup>

## If the Numbers Look Off

* Driver ARR reads zero for a cohort: confirm the cohort has assumptions set and that rep counts resolve (the sales-role classifier must match real employees or positions).
* Pipeline and drivers appear to add up: check the period's validity treatment. Set it to floor-fill if you want the larger of the two rather than the sum.
* A target does not flow into months: confirm the quarterly weights for that quarter sum to one and the region-by-stream split sums to one hundred percent.
* Live ARR lags the booking you expect: that is the activation curve at work. See [Recognized Revenue](/planning/recognized-revenue).