Needs more than a quick edit8 of 12 checks passed
The memo contains an unsupported factual claim about market expectations, requiring evidence re-check, and lacks a clear competitive moat, so a PM would need to redo some reasoning.
Output
Memo: Next Year's Strategy
To: CEO From: VP Product Re: Where we play next year, and why I don't think it's enterprise
You asked me to tell you if you're wrong. I think you are, at least for next year. Enterprise isn't a bad market. But our numbers say our winning position is somewhere else, and an enterprise push would starve that position of the engineering time and sales capacity it needs. Below is the case and what I propose instead.
---
1. Diagnosis
Cadence runs three different businesses, and only one of them is working well.
| Segment | ARR | Annual gross profit after support, per customer | CAC | Payback | NRR | Logo churn |
|---|---|---|---|---|---|---|
| SMB self-serve | $6.8M | $248 | $740 | 31 mo | 88% | 34% |
| Mid-market | $12.9M | $9,845 | $9,800 | 9 mo | 118% | 7% |
| Enterprise | $3.3M | $40,440 | $118,000 | 26 mo | 104% | 9% |
(Gross profit after support = ARR × gross margin − support cost. Total ARR is about $23M.)
Mid-market is the business. - It is 56% of ARR. - It pays back CAC in 9 months. - It grows on its own: 118% NRR on $12.9M means about $2.3M of net new ARR next year before we sign a single new logo. - We win 61% of the time against our main competitor. - Expansion comes naturally as customers add locations and seats.
Enterprise looks big per logo but is weak as a business today. - Gross margin is 64%, versus 81% in mid-market. - Each customer costs $21K a year to support. - CAC is $118K, payback takes 26 months, and the sales cycle is 9 months. - The hidden cost is engineering. 5 of the last 8 deals needed custom features, at about 14 engineer-weeks each, built by the core team. That is roughly 70 engineer-weeks diverted from the product that 890 mid-market customers pay for.
SMB is roughly break-even and pulling capacity away from everything else. - Each customer generates $248 a year after support. - With 34% annual churn, the average customer lives about 3 years. Lifetime contribution is roughly $730, against $740 of CAC. - SMB produces 62% of our support tickets. The cause is structural: we offer unlimited phone support on a $79/month plan, so support costs ($610) eat more than half of revenue ($1,100). - Trial conversion is 4%.
The core problem: we are trying to be everything to everyone. The cost shows up in two places. The roadmap gets hijacked by single enterprise deals, and the support budget gets consumed by SMB.
Why the enterprise plan makes this worse. Assume 10 new AEs close 20 enterprise logos next year, which is optimistic given a 9-month cycle and ramp time. - That adds about $1.9M of ARR. - It costs about $2.4M in CAC. - If 5 of every 8 deals still need custom work, it consumes about 175 engineer-weeks, or 3–4 engineer-years, on custom features alone. - That is before the work on SSO, audit logs, and custom workflows. - Most of that ARR would land late in the year and pay back in year three.
Meanwhile, the mid-market roadmap would stall in the segment where we already win.
---
2. Guiding Policy
Where we play: We will be the best scheduling and time-tracking platform for multi-location operators with hourly workforces, from roughly 10 to a few hundred locations. Mid-market is the core. Every product, sales, and support decision should first answer one question: does this help a mid-market customer win, stay, or add locations?
What we stop: 1. Custom feature builds for single deals. If a feature isn't on the roadmap for a segment, we don't build it to close one logo. 2. The 10-AE enterprise hiring plan. We redeploy that budget (see below). 3. Unlimited phone support on the $79 plan. That plan moves to in-app help, chat, and a knowledge base. Phone support becomes a paid tier. 4. Custom workflows as a product initiative. This is a services business disguised as a feature.
What we deliberately keep: - Our 34 enterprise customers, served well on the standard product. - Inbound enterprise deals that can close without custom work. - SSO and audit logs. Upper mid-market buyers increasingly expect these, so they serve the core rather than a new segment.
---
3. Coherent Actions for Next Year
These actions reinforce each other. Product makes mid-market expansion easier. Go-to-market (GTM) captures that expansion. Fixing SMB frees money and support capacity to fund both.
Product (mid-market roadmap, Q1–Q3) - Multi-location reporting. This is the top request, and it directly drives location expansion, our main source of NRR. - Payroll integrations. We will prioritize the top 3–4 providers by customer count. Integrations increase switching costs and reduce churn. - Shift templates. This reduces setup time for each new location, so expansion becomes cheap for the customer and for us. - SSO and audit logs. These are table stakes for larger mid-market accounts. We will scope them tightly and avoid gold-plating. - Engineering capacity rule: zero engineer-weeks go to one-off custom builds.
Go-to-market - Redirect the enterprise AE budget. Fund roughly 4–5 mid-market AEs and 3–4 account managers whose job is location and seat expansion inside existing accounts. A 9-month payback and 118% NRR make this the most efficient growth dollar we have. - Build a competitive playbook. Codify why we win 61% against our main competitor, and target that competitor's mid-market base. - Handle enterprise inbound on the standard product. One senior AE and the existing team qualify enterprise leads. A deal requiring custom work is a no unless the feature is already on the roadmap.
SMB economics - Restructure support. Move the $79 plan to digital-first support and introduce a higher tier that includes phone support. The target is cutting SMB support cost per customer from $610 to under $350. - Overhaul onboarding and trials. A 4% conversion rate signals weak activation. Fixing onboarding should also reduce tickets. - Find out whether SMB feeds mid-market. We need to learn how many mid-market customers started on SMB. If SMB is a real feeder, we keep it as a low-cost funnel. If not, we run it purely for margin.
Enterprise - Improve margins on the existing base. Standardize their configurations and reduce bespoke support. - Track demand without chasing it. We log every enterprise request, so that if we return to enterprise later, we can build a standard product rather than custom one-offs.
---
4. How We'll Know It's Working
Leading indicators (quarterly) - Multi-location reporting and payroll integrations shipped by end of Q2. - Engineer-weeks spent on custom deal work: 0. - SMB support tickets per customer down 30% by Q3. - Mid-market win rate against our main competitor: 61% → 65%.
Outcomes (year-end) - Mid-market NRR ≥ 120%, up from 118%. - Mid-market new-logo ARR grows at least 25%, with payback held at ≤ 12 months. - SMB support cost per customer under $350, and SMB contribution per customer roughly doubled. - Enterprise gross margin above 70%, up from 64%, with no logo churn among the top 10.
When we would revisit enterprise: I'm not saying never. I'd reopen the question in Q3 if either of these happens: - Mid-market customers are outgrowing us into enterprise-scale deployments, meaning enterprise becomes the natural next step for our core. - We close at least 3 enterprise deals on the standard product.
Either signal would mean enterprise has become an extension of our strength rather than a separate bet. At that point, a focused enterprise motion built on a product we actually have would make sense.
---
Bottom line: The exec team can rally behind this: win multi-location mid-market, stop subsidizing segments that don't pay. It builds on where we already win, funds itself by fixing SMB economics, and keeps enterprise open as a future option rather than a costly bet. I'd welcome the chance to pressure-test this with you before the offsite.
Claims the judge couldn’t find in the brief
- Upper mid-market buyers increasingly expect SSO and audit logs.
- Integrations increase switching costs and reduce churn.
Grades and run details
Decision model 88 · LLM judge 62
Decision model checks
- passUses the supplied evidence correctly25%
- passAddresses the actual decision100%
- passRespects explicit constraints49%
- passIdentifies material uncertainty98%
- partialAvoids unsupported claims51%
- passProduces the required deliverable83%
- passReads the segment economics90%
- passTells the CEO plainly100%
- passDiagnosis before prescription100%
- partialProposes tests that could fail72%
- partialNames an advantage that's hard to copy78%
- passSays what it gives up100%
Run
- Run
- #1
- Time to output
- 44 s
- Submitted
- 5 Oct 2026