Replace the vendor with an internal platform and carry the maintenance. This scenario starts at $400K MRR with 22 engineers and 1.8% monthly churn — break-even arrives early and unit economics clear the 3:1 bar. Every control below is live; changing one re-runs the full five-year projection.
Healthy trajectory
Break-even arrives early and unit economics clear the 3:1 bar.
From $4.9M today — 3.4x over 5 years.
On $46.1M of cumulative revenue.
Operating profit turns positive in month 9.
Lowest balance $3.9M in month 8.
LTV $63.8K against $14K acquisition cost.
Net revenue retention 101.7% per month.
60 monthly periods, compounded. Hover for exact figures.
Definitions for every metric on this page — LTV, CAC payback, churn, runway, burn multiple — are in the glossary, and the FAQ on the full simulator explains how each is computed.
Metric glossary →Metric FAQ →In-depth guides →Build the equivalent internally. One-time capital cost, then permanent maintenance.
A plain-language reading of the projection above. Every figure updates as you move the controls. Current verdict: Healthy trajectory.
Starting from $400K in monthly recurring revenue, organic growth of 3.5% outruns 1.8% monthly churn by 1.7% a month before any paid acquisition. Re-investing 15% of operating profit at a $14K acquisition cost adds a compounding second engine on top. After 60 months the model lands at $16.5M of annual recurring revenue — 3.4x today's $4.9M.
In month one the business spends $367K. Engineering payroll is $266.8K of that — 73% of every operating dollar — across 22 engineers at $145K fully loaded, plus $3.8K of contract engineering at $95 an hour. Infrastructure starts at $25.6K and reaches $79.4K by month 60, because $55 of cloud spend rides on every $1K of MRR. With salary inflation at 4.5% and price inflation at 3.4%, the total monthly cost base ends the period at $453.1K.
Gross profit first covers the full cost base in month 9 (9 mo from now). Everything lost before that point is repaid by month 14. The cash balance bottoms out at $3.9M in month 8, which the opening $4M covers with room to spare, and finishes at $16.2M.
At $1.4K ARPA and 82% gross margin, each account contributes $1.1K of gross profit a month. Against 1.8% monthly revenue churn that implies a lifetime value of $63.8K, and an LTV:CAC ratio of 4.6x, which clears the conventional 3:1 bar, so acquisition can be scaled with confidence. Acquisition cost is repaid after 1y of gross profit. The burn multiple works out at 0.0x of net burn per dollar of new ARR.
On the build in-house path, the software itself costs $2.5M over five years — $0 in licences and $2.5M in build and maintenance. Holding every other assumption constant, "Buy SaaS" would end the period $3.1M better off, largely because the maintenance FTE on an internal system is a permanent cost that never amortises away.
The same inputs run three ways. The spread between the outer columns is the honest output of any five-year model — treat the middle column as a midpoint, not a forecast.
| Metric | ConservativeGrowth misses plan, churn runs hot, costs drift up. | RealisticYour inputs, taken at face value. | AggressiveAcquisition compounds, retention improves, costs hold flat. |
|---|---|---|---|
| Year-5 ARRRecurring revenue exiting month 60. | $3.9M | $16.5M | $57.5M |
| 5-year net profitCumulative profit after re-investment. | -$6.9M | $12.2M | $51.3M |
| Ending cashBalance at month 60. | -$2.9M | $16.2M | $55.3M |
| Break-evenFirst month operating profit is positive. | Never | Month 9 | Month 2 |
| Cash runwayMonths before the cash balance reaches zero. | 3y 5mo | 60+ mo | 60+ mo |
| Lowest cash balanceDeepest point of the cash trough. | -$2.9M | $3.9M | $4M |
| LTV : CACLifetime value against acquisition cost. | 3.4x | 4.6x | 6.1x |
Identical assumptions, three software paths. Select one to load it into the simulator.
The first thing to notice about this profile is that the company is comfortable. Four hundred thousand of MRR, eighty-two percent margins, churn under two percent and four million in the bank — this business will absorb the build-versus-buy decision either way and still finish the five years with a great deal of cash. That comfort is precisely the hazard. Decisions that cannot hurt you in the current quarter get made on preference, on engineering appetite, on a strong opinion about a vendor held by someone senior. The model will not stop you. It will only show you, afterwards, what the preference cost.
The build cost is the number everyone argues over and it is almost never the one that decides the outcome. Eight hundred and fifty thousand over twelve months is large, visible, one-time, and it gets proportionally smaller every year the system stays in service. Set the maintenance FTE to zero and the five-year cost of the in-house path is exactly that capex and nothing else. Set it back to two and a half — the honest number for a platform of this kind — and the five-year total is roughly three times the build. The maintenance is close to twice the build cost on its own, it never gets smaller, and it rises with salary inflation for as long as the system exists. Every additional maintenance engineer beyond the first two and a half adds around three-quarters of the original build budget again over the same period.
Run the comparison at these inputs and buying wins decisively, which is not what the strategy control being set to "Build in-house" implies. Twenty-two engineers becomes roughly forty seats once the adjacent staff who always end up with a licence are counted, and at the default per-seat price the five-year licence bill is a small fraction of the in-house path. The buy path breaks even immediately, the build path takes most of the first year, and the difference in ending cash is around a fifth of everything the buy path earns across the five years. A fifth of five-year net profit is a serious sum to spend on a preference, and it is small enough relative to the cash balance that nobody in this company would necessarily feel it happen.
The crossover is arithmetic, and it is worth knowing where it sits rather than arguing about it. At this headcount, buying stops being cheaper somewhere around nine hundred and sixty dollars per seat per month. Below that line, no build estimate — however disciplined, however experienced the team — changes the answer on a five-year view. Above it, the case for building starts to make itself. Most organisations arguing about this are nowhere near that number and have never worked out where it is, which is why the discussion tends to be conducted in adjectives.
Headcount cannot close the gap either, and it is the argument most often reached for. Push the engineering headcount all the way to the top of its range and the licence bill still does not overtake the in-house path, because the build capex and the maintenance load are both fixed while seats scale linearly from a low base. Build economics improve with scale, but the scale required here is far beyond anything this company is planning. If the case for building depends on a headcount you do not have yet, it is a case for buying now and revisiting later, not a case for building.
One honest caveat about the duration control. Stretching the build from twelve months to twenty-four lowers the five-year total, because the maintenance clock only starts when the build ends — the model is measuring a fixed window, not the steady state. Do not read that as an argument for slower delivery. The run-rate after handover is identical either way, and it is the run-rate that decides the sixth year and the seventh. That leaves the Hybrid path as the option most worth a serious look before committing to either extreme: license the commodity layer, build only the slice that is genuinely differentiated, and carry a smaller permanent maintenance burden. It lands between the two on cost and on break-even, and it is usually nobody’s preferred outcome in the argument, which is exactly why it gets skipped.
| Year | Ending ARR | Revenue | Operating cost | Net profit | Ending cash |
|---|---|---|---|---|---|
| Year 1 | $5,880,754 | $5,365,521 | $4,498,145 | -$105,066 | $3,894,934 |
| Year 2 | $7,409,882 | $6,665,631 | $4,258,091 | $1,026,567 | $4,921,501 |
| Year 3 | $9,524,573 | $8,499,097 | $4,533,604 | $2,070,307 | $6,991,808 |
| Year 4 | $12,448,151 | $11,029,541 | $4,855,295 | $3,560,589 | $10,552,398 |
| Year 5 | $16,502,962 | $14,534,511 | $5,238,783 | $5,677,588 | $16,229,986 |
Small team, high growth, short runway. Every dollar of burn matters.
$12K MRR → $2.4M ARR
Growth is funded, headcount is climbing, tooling spend is compounding.
$140K MRR → $21.4M ARR
No outside capital. Low burn, modest growth, positive from month one.
$60K MRR → $2.5M ARR
Acquisition is working, retention is not. Watch the runway collapse.
$95K MRR → $460.3K ARR
