You have a great month. Revenue is up, maybe you closed a big deal or a product finally caught on, and someone (maybe you) does the math: “if we keep this up, we’ll do six figures this year.” That’s a run rate. It’s one of the most common calculations in small business, and it’s also one of the easiest ways to talk yourself into a number that isn’t real.
This isn’t a warning to stop using run rate. It’s a useful, fast way to size a business. The problem is what happens when you treat it as a promise instead of a snapshot.
What run rate actually means
Run rate takes a short period of performance, usually a month or a quarter, and scales it up to a full year. The simplest version: take your monthly revenue and multiply by 12. If you did $8,000 last month, your “annual run rate” is $96,000.
For subscription businesses this is usually called ARR, annual recurring revenue, and it’s built from MRR (monthly recurring revenue) times 12. The subscription version is a bit more defensible than the generic version because recurring revenue is, by definition, revenue you have reason to expect will show up again next month. A one-off product sale doesn’t have that property.
The math itself is never wrong. Multiplying a number by 12 is not where the mistake happens. The mistake happens in the assumption sitting underneath it: that this month is a fair representative of the other eleven.
Why one month rarely represents the other eleven
Any single month carries a mix of things that will repeat and things that won’t. A run rate calculation can’t tell the difference between them, it just takes whatever number you feed it and multiplies.
Say you run a small online store and December brings in $15,000 because of a holiday promotion, compared to a normal month of $6,000. Annualising December gives you $180,000. Annualising a normal month gives you $72,000. Neither number is “the” run rate of the business. They’re both true for the month they came from and false as a description of the year.
The same thing happens on the expense and effort side, which run rate calculations usually ignore entirely. A month where you land three new clients might also be a month where you worked sixty hour weeks doing manual onboarding that doesn’t scale. The revenue is real. The idea that you can sustain the pace behind it for eleven more months is a separate question that the run rate number says nothing about.
Seasonality is the most common trap
Seasonality is just a predictable pattern of ups and downs tied to the calendar. Retail has its holiday season. Tax and accounting services have their spring rush. B2B software often sees a slowdown in the summer and a budget-flush spike in Q4 as companies spend down what’s left before it disappears.
If your business has any seasonal shape to it at all, and most do, then any single month is going to sit somewhere on that curve, either above or below the average. Annualising a peak month systematically overstates the year. Annualising a trough month understates it. The only way run rate gets close to accurate for a seasonal business is if you’re calculating it from a period that happens to land near the yearly average, which you usually can’t know in advance and rarely check after the fact.
A practical way to guard against this is to look at the same month from a prior year, if you have the data, and compare it to the yearly total from that year. If last December was 20% of the year’s revenue rather than a flat 8.3% (1/12th), you now have a rough sense of how much a holiday month overstates things when you multiply it out.
One-time revenue disguised as a trend
The second big trap is mistaking a one-time event for a new baseline. This shows up constantly in service and consulting businesses. You land one large client, or one client pays for a big project upfront, and suddenly the month’s total looks like a new normal.
The test here is simple and worth actually asking out loud: will this specific dollar amount show up again next month without you doing anything new to cause it? A retainer client who pays the same amount every month passes that test. A one-time project fee does not, no matter how much it helped the month.
This is exactly why MRR exists as a separate metric from plain revenue in subscription businesses. MRR is built specifically to filter out one-time charges, so that what you’re annualising into ARR is the part of revenue that has some evidence of repeating. If you’re running a business without recurring revenue, an online store or a project-based service business, you don’t get that filter for free. You have to apply the same logic yourself, by asking which part of the month’s total is likely to recur and which part was a one-off.
Churn is the other side of the coin no one annualises
Run rate calculations almost always get built from a growth month, and they almost never get built from a month with a lot of cancellations. That’s a selection bias worth naming. Nobody sits down after losing three customers and says “let’s calculate our new lower run rate to plan around.” But that month contains just as much information about the trend as a good month does.
If you’re going to use a good month to project forward, it’s worth also checking what an average or bad month projects, and treating the real forecast as somewhere between the two, not at either extreme. A business that lost 5% of its customers last month has a different trajectory than the pure revenue number for that month suggests, even if new sales happened to offset the loss and the top-line number looked fine.
A steadier way to build a run rate estimate
None of this means run rate is useless, it just means a single month is too small a sample to build a forecast on by itself. A few adjustments make the number more honest without adding much work:
Use a longer window. Averaging the last three or six months and annualising that average smooths out both the spike months and the slow months, and gets you closer to a number you could actually defend if someone asked how you got it.
Separate recurring from one-time revenue before you do any multiplication. If a chunk of the month’s revenue came from something that isn’t going to repeat, pull it out of the base number first.
Check the trend, not just the level. A run rate tells you where you’d land if nothing changed. It says nothing about whether the underlying number is climbing, flat, or sliding, which matters more for planning than the snapshot itself.
Sanity check against last year, if you have it. Comparing this month to the same month a year ago tells you something a straight multiplication can’t: whether you’re actually growing, or whether you’re just in a predictably strong part of the calendar again.
What run rate is genuinely good for
Used carefully, run rate is a decent quick gauge of scale, useful for comparing yourself to a past version of your own business, or for a rough gut check when someone asks “how big is this thing.” It’s a snapshot, not a plan, and it works best as one input alongside trend data, seasonality, and a clear-eyed look at how much of last month’s number was one-time.
What it can’t do is guarantee that the next eleven months look like the one you just had. Treat it as a starting estimate to interrogate, not a forecast to build a budget on, and it’ll serve you a lot better than it will if you take it at face value.
If you want more plain-English breakdowns of the metrics that actually matter for running a small or one-person business, come take a look at what else we’ve written[/) on Data Research Analysis Collection.
Get new guides and videos first — join the Telegram channel.