A common thing we hear in our calls with CFOs is they don't have capacity to change billing. CFOs often don't have their own engineering so they need to get help. They like seeing the pricing model flexibility, the connectivity with ERPs, and even the CPQ functionality but they don't know how long it'll take to migrate.

So they stay with the status quo. Another year of the old system, at the old system's error rate, with the old system's workarounds.

However, I've done the math and most billing migrations at Solvimon go live within 2 months of contract start. The decision does however take longer than the migration.

How the last 14 migrations went

I pulled out some numbers on our 14 most recent migrations, contract start to first production invoice run.

!!!! HEATMAP !!!!

Here's what I found:

• Most go-lives take under 2 months.

• Half were running production invoices within ~6 weeks.

• There's a small long-tail running to 6 months.

We're talking about fancy fintechs and companies doing tens of millions of billable events a month, so it isn't a cluster of easy cases.

The long-tail has a consistent cause, which I'll get to in a bit.

Why it's faster than even I expected

I've been doing billing for years, and I know migrations are tough - because you don't rebuild your pricing in a new system, you sort of have to find a new way and new primitives to express it in.

Tiers, volume rates, regional rates, usage dimensions, commitment drawdowns: these are rules, and rules migrate in many days because you configure them once and every customer on them comes across together.

Your billing system isn't a system for charging customers.  It's a system for remembering every promise you've ever made.

Lots of companies overestimate this part because they remember how long the pricing took to invent in the first place. Re-expressing is much faster than making a whole new decision.

Historical data isn't a special project either. Usage history goes through the same ingestion APIs as live events, with original timestamps. Customers and subscriptions come in through file imports or a one-time sync from the CRM or ERP where they already live. The tooling for this is the same tooling you'll use in production, which means testing the import is testing the integration.

And the switch itself is an anticlimax if you let it be one. Our standard plan runs about 6 weeks: configure meters, products and plans in the first 2, validate invoices in week 3, then prepare the cutover, aligned to your billing cycle. If you bill monthly on the 1st, you cut over on the 1st. The dramatic version of a billing migration, where everything moves in one heroic weekend, is a choice, and you don't have to choose it.

Week 3 deserves a sentence, because it's the part people skip in their imagination. You export draft invoices from the new system and compare them, line by line, against what the old system produced for the same period. Missing tiers, wrong aggregation, an override nobody documented, a tax difference. Finding these before cutover is the entire job. Everything else is scheduling.

The honest part

A minority of migrations run long, and ours have too. The cause is consistent enough that we now look for it on day one: pricing exceptions at volume. Hand-agreed rates that appear in no plan. Discounts someone applies manually each month. Contract terms that live in an email thread from 2022. Any company that has been selling for 5 years has some of these... the question is whether they still fit in one person's head.

It took us too long to understand this: that work exists whether you migrate or not. Those exceptions are already costing you time every billing cycle, already producing the invoices someone corrects by hand. The mess predates the migration. The migration surfaces it, and unlike the status quo, it lets you schedule the cleanup: move your standard book first, live within weeks, and bring the exceptions across in a second wave, deliberately.

If you want to know which group you're in before talking to any vendor, there's an afternoon-sized test. Take 20 of last month's invoices and try to derive each one from your usage data and your contracts alone. Nothing else, no tribal knowledge, no asking Sandra in finance. If you can reproduce all 20, your migration is short and you should stop being scared of it. If you can't, you've just found the actual project, and it was never the vendor switch.

We've watched companies spend longer deciding than migrating, more than once, and the deciding is the part with no output. Nobody's invoices got better during a stakeholder alignment phase.

Run the 20-invoice test this week. It costs you an afternoon and it converts the scariest question in the evaluation, "how long will this take?", into a number you produced yourself.