Sunsetting a Product That Still Has Users
A software company has a legacy on-prem product generating $1.2M annual revenue from 140 customers. It consumes 30% of engineering maintenance capacity and has a security vulnerability class that is increasingly expensive to patch. A cloud replacement exists; 90 of the 140 customers have migrated. The remaining 50 include 12 that account for $700K of the revenue and run the product in regulated environments with strict change-control. Support costs per remaining customer are 4x the average. Sales occasionally still sells the legacy product to avoid losing a deal. Leadership is debating a hard sunset date.
1. Evaluate whether a hard sunset is the right move. What factors should drive the decision?
Read the worked answer
This is a portfolio decision, not a sentiment one. The drivers:
- Concentration risk. 12 customers = $700K of $1.2M. A hard sunset that loses them is a ~58% revenue hit, not a gradual decline. Their regulated, change-controlled environments mean migration is slow and may be contractually constrained.
- Cost trajectory. 30% of engineering for a shrinking base, with a worsening security liability, is a real and rising cost. The status quo is not free.
- Leakage. Sales still selling legacy undermines any migration plan and signals the sunset isn't credible internally.
- Migration momentum. 90 of 140 have moved; the remaining 50 are the hardest cases, so the migration curve will flatten. Extrapolating past progress understates the difficulty.
A hard sunset is defensible only if the concentrated, regulated customers have a viable path and a negotiated timeline. Otherwise it converts a manageable cost problem into a revenue cliff and a contractual fight. The right answer is likely a segmented sunset: a firm end-of-life for the long tail, a negotiated extended-support program (priced to reflect the 4x cost) for the regulated 12.
2. Recommend a plan for the remaining 50 customers, including how to handle the 12 regulated accounts and the sales leakage.
Read the worked answer
Segment and price the tail; negotiate the head.
- Long tail (38 customers, ~$500K). Publish a firm end-of-life date with migration support and incentives. This is where a hard sunset works - the cost per customer is high and the migration path is standard.
- Regulated 12 (~$700K). Offer a paid extended-support tier priced to cover the true cost (they are 4x average) and to fund the security patching. This buys time without subsidizing them from the rest of the business. In parallel, run joint migration planning around their change-control cycles. Set a contractual end date you can defend.
- Stop the leakage. Remove legacy from the sales catalog; require an exception approval for any new legacy sale, owned by product or finance, not sales. Otherwise the tail never ends.
- Protect the security liability. Whatever the timeline, the vulnerability class must be addressed or the extended-support tier must explicitly price the risk. Leaving it unpatched is not an option.
The principle: a sunset is a portfolio decision executed as a customer-segmentation plan. Uniform dates are simple but ignore that the cost and the revenue are concentrated in different places.