I've led several moves from on-premises products to SaaS, across WiFi analytics, network access control, threat intelligence and smart grid platforms. Each time, the engineering was the easier part. The hard parts were pricing, operations, and bringing existing customers along.
It's a business model change first
Perpetual licences become subscriptions; revenue becomes recurring but smaller up front; sales compensation has to change or the field will keep selling the old model. Get finance and sales aligned before engineering writes the first migration script.
Design for multi-tenancy, not lift-and-shift
Hosting the old architecture in the cloud gives you all the costs of SaaS with none of the benefits. Tenant isolation, configuration over customisation and a single version for everyone are what make SaaS economics work.
Operations becomes product
Uptime, upgrades, monitoring and support are now part of the product experience. Invest in observability and on-call before launch, not after the first outage.
Bring customers with you
Long-time on-prem customers worry about data residency, control and change. Offer a clear migration path, early-adopter incentives and honest answers on compliance. A few lighthouse customers who move first will persuade the rest better than any brochure.
Use SaaS to build an ecosystem
Once you're in the cloud, APIs, SDKs and marketplaces become possible. Some of my most valuable platform growth came after migration, when partners and developers could build on top of us.
The cloud doesn't make a product better. Shipping faster, learning from usage and building an ecosystem do; the cloud just makes them possible.
Treat the migration as a new product launch with existing customers, and it becomes a growth engine rather than a cost centre.
