Key Takeaways

• The source split carries the signal, not the total. On a $240,000 book, $6,208 built from $3,570 of upgrades and cross-sells means something different from the same total driven by a price uplift.

• ChartMogul's Growth Levers analysis of 6,525 software companies found expansion supplied 15.4% of net-new MRR for eventual $20M ARR companies while they sat at $1M ARR, and 34.7% once they got there. 86% of them raised expansion's share by more than 10%.

• Benchmarkit's 2025 B2B SaaS Performance Metrics Benchmarks put expansion ARR at 40% of total new ARR, rising past 50% above $50M ARR.

• A price increase counts as expansion in every major analytics tool, because they classify on the direction of the MRR change, not on whether the customer got anything new.

Which sources of expansion produce a team's expansion MRR?

Expansion arrives through distinct levers, and any real month mixes them. Splitting the total by lever shows whether growth came from customers wanting more or from the price list going up. One September on a $240,000 book:

That's a 2.59% rate against opening MRR. Group the rows and the story sharpens: sales-led expansion (upgrade, cross-sell, add-on) contributed $3,570, product-led expansion (seats and usage) $2,130, and pricing actions $508. Report only the $6,208 and nobody can tell those apart. MRR movements holds the wider set of categories.

Does a price increase count as expansion MRR?

Yes, and the analytics tools agree even though the aggregator pages don't. ChartMogul's documentation is explicit: when a customer's overall MRR increases, it classifies the change as Expansion, and names an expiring discount as one cause. Direction decides the category, not the reason behind it.

Several glossary pages claim a rise with no added value isn't expansion. It's a fair-sounding distinction that doesn't survive contact with a billing system:

• No billing system knows whether a customer received more value. It knows the invoice went up.

• A discount rolling off is arithmetically identical to a price increase, so counting one and not the other breaks the tie-out.

• Pricing-driven expansion still needs its own row, because it doesn't repeat next year the way seat growth does.

• Currency movement is the real exclusion. ChartMogul creates no MRR movement for exchange-rate changes, which keeps a weaker dollar from posting as growth.

How much of net-new revenue should expansion MRR produce?

It climbs with scale, from roughly 15% of net-new MRR early on toward 40% later. ChartMogul's Growth Levers report, built on 6,525 software companies, tracked the same businesses: those that reached $20M ARR took 15.4% from expansion while they were $1M ARR businesses, and 34.7% at $20M.

What the benchmark data supports:

• Contract size pulls the number up. ChartMogul's platform data has companies with ARPA above $500 a month taking 43.7% of added MRR from expansion.

• Expansion is half the comparison, and has to clear contraction MRR and churn before any of it reads as growth.

Why do two dashboards report different expansion MRR for the same month?

Because the boundary rules differ, and each moves dollars between expansion and a neighbouring line. I've reconciled this argument more than once, and it's usually one of these:

The rate formula splits too. Chargebee and Geckoboard both publish the expansion MRR rate as the change in expansion MRR over the prior month's, which measures a flow against itself and divides by zero in any month that had none. Chargebee prints the standard version too: expansion MRR over MRR at the period start.

Related terms

Each of these connects to a different edge of the expansion line:

• Contraction MRR is the mirror movement, covering revenue that existing customers stop paying without leaving.

• MRR movements holds the full set of categories expansion sits in, and settles which one a mixed change belongs to.

• Net revenue retention is where expansion gets netted against losses to produce a single ratio.

• Annual contract value tends to predict expansion share, since bigger contracts expand more than small ones do.

• Consumption-based pricing generates expansion with no sales touch and no plan change, which is why usage-priced books report it differently.

• Hybrid pricing gives a book several expansion levers at once, so the sources breakdown matters more when pricing combines seats, usage, and credits.

FAQ

Is expansion MRR the same as upsell?

No. Upsell is one source of expansion MRR, the move to a higher-priced tier. Expansion also picks up cross-sells, add-ons, seat growth, rising consumption, and price uplifts. A team reporting upsell as its whole expansion figure understates every self-serve source.

Does reactivation count as expansion MRR?

It depends whose definition you follow, which is the problem. Chargebee's glossary lists reactivation as one of four types of expansion MRR. ChartMogul treats it as a separate movement, at 3.8% of net-new MRR against 34.7% from expansion. Keep them apart, because a returning customer says something different than a growing one.

Can expansion MRR be negative?

No. Expansion MRR aggregates increases only, so the floor is zero in a month where no existing customer paid more. Decreases post to contraction or churn. A dashboard showing negative expansion is netting two movements into one field.