Allocator means M&A technical due diligence, portfolio construction, deciding which programs to fund and which to kill
Your CEO asks you to present next quarter's technology roadmap to the board. You have $4M in Budget, twelve competing proposals from your engineering leads, two acquisition targets the PE sponsors want evaluated, and a mandate to grow EBITDA by 15%. Every dollar you put into one program is a dollar you can't put into another. You're not just an Operator anymore - you're an Allocator.
An Allocator is an Operator who makes Capital Allocation decisions across a Portfolio of programs, acquisitions, and investments - using M&A Technical Due Diligence, Portfolio Construction, and NPV math to decide what gets funded, what gets killed, and what gets acquired.
An Allocator is the role you step into when your job shifts from executing a single program to choosing which programs exist at all.
You already know the three core skills:
The Allocator does all three simultaneously, across the full scope of a P&L. You're not picking one project to run. You're deciding which projects should exist, which should be acquired from outside, and which should be killed to free capital for better uses.
An Operator turns Budget into Revenue. An Allocator decides which Operators get Budget in the first place.
Consider two engineering leaders with identical teams:
Leader B creates 10x more Value Creation despite worse Execution. The Capital Allocation decision dominated the Execution decision.
If you run a P&L, you are already an Allocator whether you use the word or not. Every headcount you approve, every Build, Buy, or Hire decision you make is a Capital Allocation decision with real opportunity cost. The gap between good Operators and great ones is almost always Allocation quality, not Execution quality.
For PE-Backed companies, this is how your sponsors already think. PE operators are evaluated on Portfolio Alpha - Returns above what the market delivers. When you allocate across your programs with the same rigor they allocate across their Portfolio of companies, you speak their language and earn Operating autonomy.
The Allocator role operates on a cycle with four phases:
List every active program, Capital Investment, and Cost Center. For each, estimate:
This is your current Portfolio. Most Operators have never written this list down.
Compute NPV and IRR for each program. Compare against your Hurdle Rate. Then add the Portfolio Construction lens: two programs with identical NPV are not equal if one depends on the same outcome as your existing bets and the other is independent. Shared failure modes create Tail Risk that individual NPV calculations miss.
Before committing Budget to internal builds, ask: can we acquire this capability faster? For every major internal proposal, run a Build, Buy, or Hire analysis:
Fund the top of your ranked list until Budget runs out. Everything below the line gets killed or deferred. This is where most Operators fail - they spread Budget across too many programs instead of concentrating capital where Risk-Adjusted Return is highest.
The decision rule is forward-looking: if a program's Expected Return from this point forward no longer beats the Hurdle Rate, kill it. Money already spent is irrelevant to the marginal decision of what to do next. The only question is whether the next dollar invested earns more here or in your best unfunded alternative.
You're operating as an Allocator any time you face these situations:
Annual or quarterly planning - When you have a fixed Budget and more proposals than you can fund. This is the canonical Capital Allocation moment. Use NPV ranking, Portfolio Construction to ensure independent risk profiles across bets, and kill anything below the Hurdle Rate.
M&A evaluation - When your PE sponsors or CEO bring you an acquisition target. Run M&A Technical Due Diligence on their technology and Knowledge Capital. Compare the acquisition's expected NPV against what you could build internally with the same capital.
Program triage - When a running program misses milestones or its market shifts. Re-evaluate its NPV with updated assumptions. Compare against your next-best unfunded proposal. If the unfunded proposal now has higher Risk-Adjusted Return, kill the running program. The opportunity cost of continuing a losing bet is the Expected Return on the winning bet you're not making.
Headcount allocation - Every Hiring Targets decision is a Capital Allocation decision. A senior engineer at $250K total annual cost is a Capital Investment. Where that person sits in your Portfolio of programs determines whether you generate $2M or $200K in NPV from that spend.
You do NOT need the full Allocator framework for small tactical decisions within an already-funded program. If you're choosing between two database schemas, that's Execution, not Allocation. The Allocator lens applies when you're choosing between programs, not within them.
You're the Group CTO of a PE-Backed Holding Company with $3M in quarterly technology Budget. Four programs compete for funding:
Your Hurdle Rate is 20%. Total requested: $4.1M. You have $3M.
Step 1: Apply the Hurdle Rate filter. Program C has an IRR of 18%, below the 20% Hurdle Rate. Kill it immediately. Remaining pool: A ($1.2M), B ($1.5M), D ($600K). Total: $3.3M - still $300K over Budget.
Step 2: Rank by Risk-Adjusted Return. Rank by IRR: A (42%) > D (38%) > B (31%). Fund from the top: A ($1.2M) + D ($600K) = $1.8M spent, $1.2M remaining. Program B costs $1.5M - it doesn't fit.
Step 3: Portfolio Construction check. Programs A and D both reduce Operating costs. They share a common failure mode: if teams resist adopting the new tools, both programs underperform together. This creates Tail Risk - a single adverse outcome wipes out two-thirds of your Portfolio's expected NPV. Program B opens a new Revenue line with an independent outcome. Even though B has lower IRR, Portfolio Construction tells you that concentrating all capital on bets with the same failure mode is dangerous.
Step 4: Make the Allocation decision. Fund A ($1.2M) at full scope - highest IRR, fund it completely. Fund D ($600K) - high IRR, small cost. Remaining: $1.2M for B, but B needs $1.5M. Negotiate B down to a $1.2M reduced-scope first phase that captures 70% of the NPV ($2.66M). Total allocated: $3M. Expected Portfolio NPV: $4.5M + $2.1M + $2.66M = $9.26M.
Step 5: Document the kill. Program C is officially killed. Write the one-paragraph rationale: IRR of 18% is below the 20% Hurdle Rate, and the capital produces more value in Programs A, B, and D. This is the Allocator's hardest job - communicating the kill to the team that proposed C.
Insight: The Allocator's value wasn't in finding the best single program (Program A wins on IRR - that's obvious). The value was in the composition: killing C despite political pressure, negotiating B's scope to fit Budget, and recognizing that A and D share a failure mode that creates Tail Risk. A pure IRR ranking would have missed the Portfolio Construction dimension entirely.
Your company processes invoices manually - 50,000 per month at $2.10 Cost Per Unit (labor plus overhead). Two options to automate:
Step 1: Compute annual savings for each option.
Current annual cost: 50,000 x $2.10 x 12 = $1,260,000.
Build: 50,000 x $0.30 x 12 = $180,000/yr. Annual savings: $1,080,000.
Acquire: 50,000 x $0.25 x 12 = $150,000/yr. Annual savings: $1,110,000.
Step 2: Compute NPV for Build, showing Discount Factors.
The build takes 18 months. Savings begin Year 2 (first full year after the build ships). We model 5 years of savings, Years 2 through 6. Discount Factor at 20%: Year N = 1 / (1.20)^N.
| Year | Savings | Discount Factor | Present Value |
|---|---|---|---|
| 2 | $1,080K | 1/1.44 = 0.694 | $750K |
| 3 | $1,080K | 1/1.728 = 0.579 | $625K |
| 4 | $1,080K | 1/2.074 = 0.482 | $521K |
| 5 | $1,080K | 1/2.488 = 0.402 | $434K |
| 6 | $1,080K | 1/2.986 = 0.335 | $362K |
| Total | $2,692K |
Net NPV: $2,692K - $1,800K = $892K before risk adjustment.
Failure assumption (stated explicitly): If the build fails, we estimate losing half the $1.8M Implementation Cost: $900K. Why 50%? Early-phase deliverables - architecture design, vendor evaluation, partial automation of the simplest formats - retain value and can be repurposed. Code written for integrations that never reach production is the lost portion. Adjust this percentage in your own scenarios based on how modular the build plan is.
Risk-adjusted NPV: 0.70 x $892K + 0.30 x (-$900K) = $624K - $270K = $354K.
Step 3: Compute NPV for Acquire, with Knowledge Asset valuation.
The target's technology is production-ready. Savings begin Year 1.
| Year | Savings | Discount Factor | Present Value |
|---|---|---|---|
| 1 | $1,110K | 1/1.20 = 0.833 | $925K |
| 2 | $1,110K | 1/1.44 = 0.694 | $770K |
| 3 | $1,110K | 1/1.728 = 0.579 | $643K |
| 4 | $1,110K | 1/2.074 = 0.482 | $535K |
| 5 | $1,110K | 1/2.488 = 0.402 | $446K |
| Total | $3,319K |
Total acquisition cost: $3,200K + $400K = $3,600K.
Net NPV before Knowledge Asset: $3,319K - $3,600K = -$281K.
Knowledge Asset valuation (cost-to-replicate method): The target's most valuable Knowledge Capital is its proprietary format mappings for 200 invoice vendors, accumulated over 3+ years of production use. To replicate this knowledge independently - separate from building the processing system itself - you would need to hire 3 specialized engineers and invest approximately 2 years of their time building and validating mappings against production data. At $250K total annual cost per engineer: 3 x $250K x 2 years = $1,500K. This is a floor estimate: it covers engineering labor but not the edge cases and vendor-specific exceptions that only surface after years of production exposure.
Adjusted NPV: -$281K + $1,500K = $1,219K.
Failure assumption: We estimate 10% probability of integration failure, resulting in total loss of the $3,600K invested. This rate is lower than the Build's 30% because the core technology is production-proven - the remaining Execution Risk is organizational integration. If integration fails, the acquired technology cannot be deployed and no savings are realized.
Risk-adjusted NPV: 0.90 x $1,219K + 0.10 x (-$3,600K) = $1,097K - $360K = $737K.
Step 4: Compare and apply the Portfolio lens.
Build risk-adjusted NPV: $354K. Acquire risk-adjusted NPV: $737K. Acquire wins by $383K.
Three factors drive the gap:
The Knowledge Asset valuation is the single most debatable number in this analysis. If you estimate the cost to replicate at $1,000K instead of $1,500K, the Acquire risk-adjusted NPV drops to $558K - still ahead of Build, but the margin shrinks from $383K to $204K. Show this Sensitivity Analysis in any real presentation.
Insight: Operators instinctively want to build. Allocators show the Discount Factor math year by year and discover that timing delays cost more than intuition suggests - the 18-month delay alone accounts for $627K of the NPV gap, more than half the difference. The Knowledge Asset valuation is where reasonable people will challenge you. The Allocator's discipline is making the methodology explicit (cost to replicate, with stated inputs) so others can challenge the assumptions instead of dismissing the conclusion.
Show your Discount Factors. The largest NPV errors come from sloppy timing assumptions - treating delayed Cash Flow as if it arrives on day one. A table with one row per year, showing the Discount Factor and resulting present value, is the difference between analysis that persuades and analysis that gets challenged in the first five minutes.
Knowledge Asset valuation is where every acquisition argument gets won or lost. Always state the methodology (cost to replicate, comparable transactions, or Revenue multiple) and the specific inputs behind your number. A Knowledge Asset valued at 'roughly $1.5M' with no method is an invitation for your CFO to halve it.
Killing a program is a forward-looking marginal decision. Compare the Expected Return from this point forward against the Hurdle Rate and your best unfunded alternative. Money already invested is irrelevant - the only question is whether the next dollar earns more here or somewhere else in your Portfolio.
Spreading Budget thin instead of concentrating capital. Operators who become Allocators often fund eight programs at half-Budget instead of four programs at full Budget. Underfunded programs have higher Execution Risk and longer Payback Periods. Better to fully fund fewer programs and kill the rest.
Ignoring the Shadow Price of engineering time. An internal build is never free - it consumes engineering capacity that could produce NPV elsewhere. That capacity has a Shadow Price equal to the Expected Return of the next-best program those engineers could work on. Allocators account for this; Operators often don't.
You manage a $5M annual technology Budget across a Multi-Brand Portfolio of three PE portfolio companies. Brand A proposes a $2M CRM migration (NPV $3.1M, IRR 24%). Brand B proposes a $2.5M warehouse automation (NPV $5.2M, IRR 35%). Brand C proposes a $1.5M customer analytics platform (NPV $2.8M, IRR 29%). Your Hurdle Rate is 15%. How do you allocate?
Hint: Rank by IRR, fund from the top, and check whether you can fit a third program. Consider whether any two programs depend on the same outcome - if so, that shared failure mode matters for Portfolio Construction.
All three programs clear the 15% Hurdle Rate. Rank by IRR: B (35%) > C (29%) > A (24%). Fund B at $2.5M. Fund C at $1.5M. Total: $4M spent, $1M remaining. A needs $2M - does not fit. Can A's scope be reduced to a $1M first phase that captures approximately 40% of the NPV ($1.24M, still above Hurdle Rate)? If yes, fund the first phase. If not, hold the $1M in reserve for mid-year reallocation. Portfolio Construction check: B (warehouse, operational) and C (analytics, Revenue-side) have independent outcomes - no shared failure mode. Total expected NPV: $5.2M + $2.8M + $1.24M = $9.24M on $5M invested.
An acquisition target has $800K ARR, growing 30% year-over-year. The asking price is $4M (5x ARR). Your M&A Technical Due Diligence reveals strong Knowledge Capital (proprietary algorithms, 3 senior engineers), but the codebase has $600K in estimated deferred maintenance that becomes Implementation Cost after the deal closes. The alternative is building equivalent capability internally for $2.2M over 24 months with a 60% on-time probability. Your Hurdle Rate is 20%. Which do you choose?
Hint: Project the target's Revenue forward 5 years at the stated growth rate and discount each year with explicit Discount Factors at 20%. For the Build path, 24 months of delay means Revenue starts Year 3 at the earliest. State your failure cost assumption: what fraction of the $2.2M is lost if the project does not ship? For the Knowledge Asset, use a cost-to-replicate method and show your inputs.
Acquire path: Project Revenue at 30% growth, discounted at 20%.
| Year | Revenue | Discount Factor (1/1.2^N) | Present Value |
|---|---|---|---|
| 1 | $800K | 0.833 | $667K |
| 2 | $1,040K | 0.694 | $722K |
| 3 | $1,352K | 0.579 | $783K |
| 4 | $1,758K | 0.482 | $847K |
| 5 | $2,285K | 0.402 | $919K |
| Total | $3,938K |
Costs: $4,000K acquisition + $600K deferred maintenance = $4,600K. Net NPV: -$662K.
Knowledge Asset (cost-to-replicate): The 3 senior engineers and proprietary algorithms represent accumulated production knowledge that the $2.2M build estimate does not fully replicate - the build gives you equivalent technology, but not the team's domain expertise or the edge cases discovered in production. Estimate: 3 engineers x $250K/yr x 2 years of domain-specific knowledge accumulation = $1,500K. Adjusted NPV: -$662K + $1,500K = $838K.
Build path: Revenue starts Year 3 (24-month build). Assuming equivalent Revenue trajectory delayed 24 months:
| Year | Revenue | Discount Factor | Present Value |
|---|---|---|---|
| 3 | $800K | 0.579 | $463K |
| 4 | $1,040K | 0.482 | $501K |
| 5 | $1,352K | 0.402 | $543K |
| Total | $1,507K |
Net: $1,507K - $2,200K = -$693K.
Failure assumption (stated explicitly): 40% probability of failure (1 minus the 60% on-time estimate). On failure, we lose half the Implementation Cost: $1,100K. Why 50%? Architecture work, market research, and early prototyping retain some value; custom implementation that never ships does not. Adjust this ratio based on how reusable the early deliverables are.
Risk-adjusted: 0.60 x (-$693K) + 0.40 x (-$1,100K) = -$416K - $440K = -$856K.
Verdict: Acquire at $838K vs Build at -$856K. Even applying a 10% integration failure rate to the Acquire path (0.90 x $838K + 0.10 x (-$4,600K) = $754K - $460K = $294K), the acquisition outperforms Build by over $1.1M in risk-adjusted terms. The 24-month delay is the primary driver: the Build path captures only 3 years of Revenue in a 5-year Time Horizon while the Acquire path captures all 5. The Knowledge Asset valuation is debatable - at $1,000K instead of $1,500K, the Acquire NPV drops to $294K before risk adjustment. Still positive, still ahead of Build.
The Allocator applies Capital Allocation math, Portfolio Construction thinking, and M&A Technical Due Diligence simultaneously across a full P&L. This is the skill that separates running one program well from deciding which programs should exist.
Disclaimer: This content is for educational and informational purposes only and does not constitute financial, investment, tax, or legal advice. It is not a recommendation to buy, sell, or hold any security or financial product. You should consult a qualified financial advisor, tax professional, or attorney before making financial decisions. Past performance is not indicative of future results. The author is not a registered investment advisor, broker-dealer, or financial planner.