Audience: Executive leadership
Timeline: Within 1 week
Framing: This is a two-way door.
The decision
The core question: Should we sunset our legacy on-prem offering ($4.8M, 22% of revenue, declining 15% YoY) to focus the entire company on the cloud platform, or continue maintaining both?
My reversibility read: This is a one-way door, not two-way.
Your heuristic says two-way door, but I disagree given the specifics. Killing the on-prem product mid-contract for three strategic accounts with renewals in 14 months is a decision that likely cannot be undone — those relationships, once poisoned, don't come back easily. The $4.8M revenue is replaceable if cloud conversion works; the strategic account relationships may not be. You should apply one-way-door rigor: this decision needs a definitive answer, but you need to protect against the irreversible damage path. The two-way-door framing only applies if you consider a staged transition (e.g., 12-18 month sunset with committed support) rather than an abrupt kill date.
Key questions to answer before deciding
-
What are the exact contract end dates for the three on-prem-only strategic accounts, and what are the renewal terms/price sensitivity? If two renew in the next 6 months versus 14 months, your timeline compression changes entirely.
-
Can the on-prem team be reduced (not eliminated) — e.g., cut from 35% to 15% of engineering through a "security-maintenance-only" release train? This would preserve the strategic accounts while dramatically reducing the drain.
-
What is the actual NRR lift potential from migrating the on-prem base to cloud? If the three strategic accounts convert at 128% NRR, does the math justify a longer runway even at higher cost?
-
What does the competitive landscape look like — is on-prem demand collapsing across the market, or is this specific to our product's neglect? If competitors are still selling on-prem successfully, we may be underinvesting rather than the category dying.
-
What is the contractual obligation exposure if we sunset mid-term? Sales' warning about "poisoning" accounts suggests there may be minimum service terms or implicit renewal expectations that create real liability.
-
Is there a third path — spin out or sell the on-prem business to a services firm? This would offload the cost while preserving customer relationships (and potentially creating a referral pipeline).
Recommended frameworks
1. The Rigorous Decision Matrix ( Bezos-style) — Weight each option (sunset vs. maintain vs. hybrid) against: revenue impact, strategic account risk, engineering capacity unlock, and competitive positioning. Assign probability-weighted outcomes. For your case: sunset scores well on engineering unlock (+35% capacity) but poorly on strategic account risk (-high probability of losing 2 of 3 accounts). Maintain scores opposite. A hybrid (12-month maintenance mode) scores moderately on all four.
2. The Reverse-Bias Test — Ask: "If I were newly appointed CEO tomorrow and didn't have legacy emotional attachment to either product, what would I do?" For on-prem: declining 15% YoY, 35% engineering for 22% revenue, accelerating churn. The objective answer leans toward exit — but the strategic account concentration creates tail risk that changes the calculus.
3. Options Generation (not just binary) — You're framing this as sunset-or-keep, but the real answer is likely option three: shrink to maintenance mode. What would a 'minimum viable on-prem' look like? Single release train per year, only critical security patches, dedicated but lean support (not bespoke), with explicit migration incentives for the three strategic accounts.
Decision criteria
A good answer must pass these tests:
-
Does not lose the three strategic accounts in the next 18 months — Any path that triggers premature departure fails. Sales' warning is a real signal, not FUD.
-
Frees at least 15% of engineering capacity — The 35% allocation is unsustainable if cloud is the future. A valid path must unlock meaningful resources.
-
Has a clear 24-month financial projection — Either: (a) on-prem revenue declines to $0 with cloud conversion compensating, or (b) on-prem stabilizes at reduced cost with positive margin.
-
Provides a migration path for on-prem customers — Not just "we're shutting down" — there must be an explicit cloud offer with pricing/terms that make migration attractive (not punitive).
-
Executive team alignment — Given the timeline (1 week), the decision criteria must be explicit enough that the leadership team can actually decide, not punt to a committee.
Sources to consult
Research plan (since no pre-checked sources provided):
-
Contract analysis — Pull the exact renewal dates and contract language for the three on-prem strategic accounts. Understand minimum terms, early-termination clauses, and any contractual service-level commitments. Why: This determines your actual flexibility and liability.
-
Financial modeling — Build a 3-scenario model (sunset, maintain, hybrid) with: on-prem revenue trajectory (apply the 15% YoY decline), cloud conversion rates from on-prem base, engineering cost allocation, and margin impact. Why: You need hard numbers to present to exec leadership in one week.
-
Sales intelligence — Interview the account reps for the three strategic on-prem accounts. Get specific: What would they do if we announced sunset? What would it take to migrate them? Is there a competitive incumbent waiting if we exit? Why: Sales' warning is a hypothesis — validate or invalidate it with evidence.
-
Engineering capacity data — Quantify what the 35% engineering allocation actually covers. How much is net-new feature work vs. bug fixes vs. security patches vs. bespoke customer work? Why: You may find that 20% of that 35% is unnecessary if you shrink the scope.
-
Market research — Check if competitors (e.g., [relevant competitors based on your product category]) still sell on-prem. Are enterprise customers broadly abandoning on-prem, or is this a company-specific issue? Why: Determines if you're exiting a dying market or creating a self-fulfilling prophecy.
-
Customer success data — What has been the actual experience of on-prem customers who have churned? Is it product-related (we stopped innovating) or category-related (they moved to cloud regardless)? Why: Informs whether migration incentives would work.
Next steps
This week's concrete actions:
-
Monday: Request contract data for the three strategic on-prem accounts — exact renewal dates, terms, and any termination clauses. This is the single most important data point.
-
Tuesday: Commission the engineering lead to model a "maintenance mode" on-prem scenario — what does it look like to reduce from 35% to 12-15% of engineering with a single annual release train and only critical patches?
-
Wednesday: Conduct 30-minute interviews with the three strategic account reps. Script: "If we announced a 12-month sunset with a cloud migration offer, what would [account] likely do?"
-
Thursday: Build the 3-scenario financial model with your FP&A partner. Bring to exec meeting.
-
Friday: Exec leadership decision meeting.
My provisional recommendation: Do NOT do an abrupt sunset. The strategic account risk is too high and the timeline too compressed. Instead, propose a 12-18 month phased transition: (a) immediately reduce on-prem engineering to maintenance mode, (b) initiate migration conversations with the three strategic accounts with a compelling cloud offer, (c) set a public end-of-support date 18 months out, (d) measure conversion rate at 6 months and adjust. This unlocks engineering capacity, protects relationships, and gives you optionality.
When to escalate
Escalate immediately if any of these become true:
-
Any of the three strategic accounts signal they will not renew under any sunset scenario — This means the risk is not manageable; you need board-level or special-situation expertise.
-
Engineering cannot deliver a viable maintenance-mode scope — If they say "we can't go below 30% without compromising security compliance," the cost calculus changes fundamentally.
-
Legal flags significant early-termination liability — If sunsetting triggers penalty clauses that exceed the revenue benefit, you need counsel.
-
Cloud pipeline data deteriorates — If the 128% NRR or cloud growth slows, the math for "invest in cloud and exit on-prem" breaks down. You need to see Q3 pipeline before committing.
-
Two or more execs disagree strongly — Given the 1-week timeline, if you don't have near-consensus by Thursday, escalate to the CEO for a deciding vote. This is a one-way-door decision — it needs clear ownership and cannot be committee-ed to death.
Generated by YourBrief.io. Not financial, legal, or medical advice. Validate with primary research.