Multi-Product Portfolio Leadership¶
Meridian Logistics now runs three distinct product lines under Amara: Core Carrier Ops (Ravi's and Tom's pods — onboarding, integrations, pricing), Marketplace (a newer bet connecting carriers to spot-market loads, led by Sofia), and Internal Tools (Priya's pod, serving Meridian's own ops staff). Each has a different maturity, a different growth rate, and a different claim on the same finite engineering budget. Amara's job is no longer to have an opinion on any one roadmap — it's to decide how much of the company each product line deserves, and to defend that allocation when a Lead argues (correctly, from their seat) that their line deserves more.
Classify before you allocate¶
Every product line falls into one of four postures. Misclassifying one is the single most common portfolio mistake — treating a Growth product like a Core product starves it just as it's proving out, and treating a Core product like a Growth bet starves the revenue that funds everything else.
| Posture | Signal | Investment logic | Meridian example |
|---|---|---|---|
| Core | Established, funds the company, customers expect stability | Protect reliability, invest in efficiency, resist scope creep | Core Carrier Ops |
| Growth | Proving product-market fit, metrics trending but not proven | Fund aggressively, tolerate higher risk, short review cycles | Marketplace |
| Harvest | Mature, declining relevance, still has revenue | Minimize investment, extend life cheaply, plan the sunset | (none yet at Meridian) |
| Seed | Pre-PMF, small team, exploring | Cap the spend, set a kill/scale decision date, protect from Core's process overhead | Internal Tools' newest ops-forecasting feature |
The allocation model¶
Amara builds a simple portfolio table each quarter and takes it to the CEO, Lena, before finalizing headcount. The percentages are engineering-capacity share, not headcount, since pods vary in size.
| Product line | Posture | Current % of eng capacity | Revenue/strategic weight | Target % next quarter |
|---|---|---|---|---|
| Core Carrier Ops | Core | 55% | Funds 90% of company revenue | 45% |
| Marketplace | Growth | 30% | 0% of revenue, board's #1 growth bet | 40% |
| Internal Tools | Seed/Harvest mix | 15% | Cost-saving, no revenue | 15% |
Let's verify this allocation is internally consistent with headcount.
python3 -c "
eng_total = 40 # total product-eng headcount
current = {'core': 0.55, 'marketplace': 0.30, 'tools': 0.15}
target = {'core': 0.45, 'marketplace': 0.40, 'tools': 0.15}
for k in current:
cur_hc = round(eng_total * current[k])
tgt_hc = round(eng_total * target[k])
print(k, 'current_hc=', cur_hc, 'target_hc=', tgt_hc, 'delta=', tgt_hc - cur_hc)
print('current total', sum(round(eng_total*v) for v in current.values()))
print('target total', sum(round(eng_total*v) for v in target.values()))
"
Running this gives: core 22→18 (-4), marketplace 12→16 (+4), tools 6→6 (0), current and target both summing to 40. The move is a clean four-engineer swap from Core to Marketplace — small enough to execute without a reorg, large enough for Sofia to feel the difference in her pod.
The three questions that decide reallocation¶
Before moving capacity between lines, Amara forces every request through the same three questions, in this order:
- What decision does more capacity unlock that less capacity doesn't? Not "we'll go faster" — a specific milestone that flips from "next year" to "this quarter."
- What does the losing line stop doing, specifically? If nobody can name the thing that stops, the "loss" wasn't real capacity, it was slack that shouldn't have existed — a finding in itself.
- What's the review date to re-litigate this? Every reallocation is time-boxed. Nothing is permanent without a second decision.
Worked example: Sofia's case for more capacity, Ravi's case against¶
Sofia comes to Amara with data: Marketplace's match rate (loads matched to carriers within 24 hours) is at 61%, and her model says two more engineers gets it to 78% within a quarter — the threshold the board set for green-lighting Marketplace as a standalone product line. Ravi pushes back: Core Carrier Ops has a customer-committed SLA migration due the same quarter, and losing two engineers risks missing it, which risks actual contracted revenue.
Amara's process, not instinct:
- She asks Ravi the loss question directly: what specifically slips? Ravi names it — the SLA migration moves from "on track" to "50/50," with a named enterprise customer whose contract renews on that timeline.
- She asks Sofia: is 78% match rate the actual board threshold, or her own stretch goal? Sofia checks — it's the actual threshold, documented in the board deck from two quarters ago.
- The reversibility test breaks the tie: Marketplace's board threshold has a hard external decision date (the next board meeting, six weeks out). The SLA migration has a customer relationship that can absorb a two-week slip with an early warning call. Marketplace can't absorb slipping past the board meeting.
- Decision: two engineers move for six weeks only, with an explicit return date, and Ravi gets a customer-communications assist from Amara to manage the SLA slip risk. Both Leads leave knowing the criteria, not just the verdict.
Portfolio review cadence¶
| Cadence | Who's in the room | What's reviewed |
|---|---|---|
| Weekly | Amara + whichever Lead has an active fire | Tactical only, not allocation |
| Monthly | All five Leads | Cross-line dependencies, early-warning metrics per line |
| Quarterly | All five Leads + CEO Lena | Full portfolio table, reallocation decisions, posture reclassification |
How It Actually Works¶
Portfolio leadership runs into a specific measurement trap: products at different maturity stages need different success metrics (a new bet is judged on learning velocity, a mature product on efficiency and margin), but a single company-wide dashboard almost always defaults to one uniform metric set, which systematically punishes early-stage bets for not yet looking like the mature product and rewards mature products for incremental optimization that's actually lower-leverage than it appears. Left unmanaged, this metric mismatch causes companies to starve their next bet in favor of the current cash cow every single funding cycle, not because leadership doesn't believe in the new bet intellectually, but because the standard dashboard makes the cash cow look like the obviously better investment quarter over quarter. The mechanism that protects a portfolio's future bets is deliberately different, stage-appropriate success criteria negotiated and defended before the funding review, not during it — because defending a nonstandard metric live, in the room where the mature product's numbers are already up on the same slide, is a losing argument almost every time.
Exercise¶
Build a portfolio table for the product lines or major initiatives you oversee (or, if you're not yet managing a portfolio, use two projects your own team runs as a stand-in). Classify each by posture, estimate current capacity share, and write the three-question test output for one reallocation you're currently being asked to make or are considering making yourself.