"Should we build or buy?" is the single most consequential comp-ops decision you'll make in any given 3-year window. The wrong answer costs 12-24 months of rework — a premature custom build that never scales, an oversized SPM that configures itself into a mess, a spreadsheet that outgrows its team. And "the right answer" depends on seven specific factors about your team and plan, not on what's in the industry analyst report.
This decision tree walks you through those seven factors and scores each of four architecture patterns against them. The recommendation is the architecture that scores highest — but the full score breakdown matters too, because a close second may be the right answer if you're about to hit a tipping point that the tool can see but you haven't factored in yet.
The four architectures — and when each wins
SPM Platform (Xactly, CaptivateIQ, Varicent, Performio)
Best when: 30–500 reps, moderate-to-complex plan, modest technical team, scale growing. Buy means you inherit battle-tested calculation logic, audit trails, and rep-facing statements. Falcon's estimate: $50K–$500K/year in license + implementation. The typical right answer for mid-size enterprise SaaS.
Custom Build (full-stack, internal team)
Best when: you have a dedicated 2+ person engineering team, extremely unique plan logic, multi-system integration requirements, and multi-year roadmap. Custom wins on flexibility and control. Cost: $1M+ over 2-3 years fully loaded. Common at hyperscaler companies with unusual plans, not common elsewhere.
Spreadsheet + Data Warehouse
Best when: under 30 reps, simple plan, tight budget, ad-hoc analysis needs. Excel/Google Sheets on top of a clean data warehouse query layer. Cost: ~$0 incremental if you already have the DW. Works well for 2-4 years; becomes unmaintainable above ~30 reps or at any meaningful plan complexity.
Hybrid (SPM + custom overlays)
Best when: your plan has 2-3 dimensions the SPM can't handle well but the rest is standard. SPM for the 80%, custom layer for the 20% that matters. Cost: SPM license + 1-2 engineer-quarters/year for overlays. Often the right answer for large enterprise but rarely the first instinct.
A premature custom build that gets abandoned 18 months in typically costs $800K–$1.5M in sunk engineering time (Falcon's estimate based on observed builds) plus another $500K in SPM migration when you eventually switch. An oversized SPM bought "to be safe" costs 3x the license fee of a right-sized solution plus the configuration overhead of fighting a product that wants more complexity than you have. The wrong choice isn't just suboptimal; it's actively expensive and sets comp ops back 2-3 years.
Architecture Decision Tree
7 questions. Answer about your situation. We score all four architectures.
ℹ️ How this tool works +
The question it answers: Given my team size, plan complexity, budget, tech team, growth, integration needs, and timeline — which of the four comp architectures fits best, and what's the confidence level of that recommendation?
What to answer:
- 7 multi-choice questions covering: team size, plan complexity, tech team size, budget, growth trajectory, integration depth, launch timing.
- Each answer adds weighted points to the 4 architectures. The architecture with the highest score is recommended.
What you'll get back:
- Recommended architecture with a tagline explanation.
- Confidence band: Clear Winner (>30% ahead) / Close Call (10-30% ahead) / Too Close to Call (<10% ahead — revisit specific factors).
- Per-architecture score bar so you can see the second-place option too.
- Pros and cautions specific to the recommended architecture.
Answer honestly based on today's reality, not where you wish you were. Scoring aspirationally produces recommendations that don't match what you can actually execute.
Benchmarks, ranges, and default values in this tool reflect Falcon's practitioner experience across consulting engagements. They are directional starting points, not substitutes for market survey data. For binding compensation decisions, validate key figures against Radford, Mercer, Carta, or WorldatWork survey data for your specific geography, industry, and company stage.
How to act on the recommendation
Clear Winner (>30% ahead)
Your inputs strongly favor one architecture. Proceed with confidence, but still validate the specific vendor choice (for SPM) or engineering roadmap (for custom) separately. The architecture decision is confirmed; the execution plan is the next question.
Close Call (10–30% ahead)
The recommendation is defensible but the second-place architecture is also viable. Look at the specific factor that tipped the decision — if it's close to flipping (e.g., you're at 28 reps and the threshold was 30), pilot both for a quarter before committing. Or pick the architecture that's more reversible (usually: start with SPM, stay flexible).
Too Close to Call (<10% ahead)
Your situation genuinely fits multiple architectures — which means the architecture decision is less important than execution quality. Pick based on team preference / hiring feasibility rather than optimization, and focus energy on the ops rigor that determines whether any architecture works.
Custom build → SPM is painful but common. SPM → custom build is almost never done (the configured logic is hard to extract). Spreadsheet → SPM is smooth. SPM → spreadsheet is only done when killing comp ops entirely. Favor the architecture that leaves you the most options if your situation changes — usually SPM for growing teams, hybrid for complex teams.
Making the architecture decision?
We help SalesOps and IT teams validate architecture choices before major commitments. Talk to us before the multi-year budget gets approved.
Book a 20-minute consultation →FAQ
Complementary. The Build vs Buy Diagnostic is deeper on the "should we build" decision specifically. This Architecture Decision Tree is broader — it recommends across 4 archetypes including spreadsheet and hybrid, which Build vs Buy doesn't cover. Use this first for the big-picture direction; use Build vs Buy for the custom-vs-SPM specifics once you've narrowed.
Answer the questions based on your current state and requirements (not what you wish you had). If the recommendation differs from your current architecture significantly, the switch may be justified — but budget for migration cost honestly. Running the Plan Effectiveness Scorecard alongside this tool helps you evaluate whether switching solves your actual problems or not.
Annually, at the start of your fiscal planning cycle. More often if you're approaching a threshold (30 reps, 100 reps, major product launch, M&A). The tipping points cluster — crossing one often means the architecture needs to scale with you.
Unlikely but possible if your plan is genuinely simple and your tech team is very small. Re-check your answers: honest budget, honest complexity, honest tech maturity. If it still lands on spreadsheet, operate on a 12-month expiry — at 40 reps you'll cross into SPM territory quickly, and planning the transition is better than being forced into it.