New MRR

In a subscription company, "revenue went up $6,500 this month" tells you less than it sounds like. Whether that $6,500 came from customers who just arrived or from customers who were already there adding seats is packed into the same sentence.

Without the distinction, some decisions can't be made at all. Suppose marketing spent $15,000 that month and revenue rose $6,500. If all of the $6,500 was three existing accounts expanding, the advertising did no work whatsoever. If all of it came from new customers, spending more of the same budget is the obvious move.

In practice it shows up like this. The CEO asks how many new customers there were and hears twelve; she asks what that is in dollars and someone has to go reopen a spreadsheet. Twelve accounts at $40 a month and twelve accounts at $1,600 a month are entirely different events for the company, and they get reported as the same number.

New MRR pins that distinction to a single figure. Separate out the share of revenue growth that came from new customers and it can be compared directly against what was spent to bring them in.

Add up what the customers who paid for the first time this month have committed to pay every month. Extra money from customers who were already there does not belong here.

Four rules hold it together.

1. Only accounts that actually paid. A free-trial signup enters in the month the billing starts. Counting them during the trial inflates the number fourfold when one in four trials converts.
2. Convert to a monthly figure. A $30,000 annual contract adds $2,500 — the twelfth — to that month's new MRR. The $30,000 on the contract is not in the unit that monthly recurring revenue (MRR) uses.
3. Use what is actually billed. A $400 plan sold at 30 percent off for the first six months counts as $280. In the month the discount ends and the price returns to $400, that $120 difference becomes expansion MRR.
4. Leave out one-time money. Setup fees, training, and data migration charges don't arrive again next month.

Say eight accounts paid for the first time this month. Six picked the $400 plan for $2,400, one picked the $1,600 plan, and one signed a $30,000 annual contract worth $2,500 a month. New MRR for the month is $6,500.

One more thing has to be settled: where a returning customer goes. Some companies count them as new and some pull them out as reactivations. Either works, as long as the rule doesn't change — change it and that month alone jumps, which makes the comparison against last month meaningless. The same goes for the date. Pick the contract signature or the first charge and keep using it.

The figure alone says nothing about good or bad. New MRR of $6,500 is a different result depending on whether it cost $8,000 or $30,000 to produce, which is why this number is read in a pair with customer acquisition cost (CAC).

The second thing to check is whether one deal carried it. If $2,500 of the $6,500 came from a single large contract, the other seven accounts produced $4,000 between them. Months like that are hard to use as an average, so many teams record the figure with the large deal removed as a separate line.

Three reference lines go together with it.

1. Spread rather than attainment against target. Six months at $6,300, $6,700 and $6,500 support next month's plan; $2,500, $12,000 and $5,000 average out to the same $6,500 and guarantee nothing.
2. Size against churned MRR. However large new revenue is, the company shrinks in any month where more disappeared than arrived. The gap between the two is what net new MRR reports.
3. Share of total growth. An established subscription business gets roughly a third of its revenue growth from existing customers paying more. When all of the growth is new business, growth speed is tied directly to the advertising budget.

"New MRR was $6,500 and $2,500 of it is one account. Take that out and we're where we were last month."

One large deal lifted the number while the flow of customers stayed flat.

"Even for annual contracts, divide by twelve before it goes into new MRR."

Don't mix total contract value with a monthly metric.

"New MRR beat the target and acquisition cost doubled."

The number was hit, but it cost far more to hit it.

At Pagedeck this is a figure Ravi and Sofia look at together. Ravi owns the small teams that come in through ads and content and put a card in themselves; Sofia owns the enterprise accounts signed in person. The new MRR the two of them produce differs in size and in character, so reading the $6,500 total as one line hides which side is working.

The number therefore gets split by channel every month. If the ad share drops while the sales share rises and the total holds, each deal now takes longer to close, so keeping the same number next quarter means more people in sales.

The most common error is booking the total contract value as that month's new MRR. The $30,000 on the contract really was won this month, so writing it down feels right. But this metric is denominated in revenue that repeats monthly, so only the $2,500 belongs. Enter the full amount and the month spikes twelvefold, and a target built on that spike gets reported as a miss no matter how well the team does.

Treating one month as next month's expectation is the other one. Last month produced $6,500, so this month should too — natural enough. New business swings by a factor of two depending on whether a large deal lands, though. Plan hiring and budget around $6,500 and the month that brings $3,000 without a big contract throws off everything downstream. Plans come from the trend across several months together with the figure that excludes large deals.