Budget and ROI: what a good investment looks like

Context: what the market spends (so we calibrate, not copy)

• Marketing budgets overall sit at 7.7% of company revenue (Gartner CMO Spend Survey, 2025), with events the largest offline channel — 19.3% of non-digital spend — implying events ≈ 7–8% of total marketing budget for the average firm.
• Professional-services firms skew far heavier: nearly half allocate >30% of marketing program budget to events, and sub-$500M-revenue companies dedicate proportionally more than larger ones (Forrester Marketing Budgets Survey, 2025). This is INTO's reference class.
• High-performing organisations put ~63% of event budget into events they control — hosted dinners, workshops, roadshows — and 58–59% of marketers are shifting to more, smaller hosted events (Forrester State of B2B Events, 2024 + 2025). The market is moving toward exactly the modes this playbook defaults to.

Calibration for INTO, not a rule: events earn a large share of MarCom spend only while they clear the investment test below. The Forrester 63%-owned split is the sanity check on portfolio shape: majority owned formats, minority sponsored/attended.

Fully-loaded costing (the honest number)

Count everything or the ROI math lies:

1. Direct: sponsorship/tickets, venue + F&B, build/production/AV, design/print/swag, tools, paid promo.
2. Travel: flights, hotels, ground, per diems. Team of 3, 3 days at a North American conference ≈ $10–30K all-in is the published planning band for a mid-size team (futureforth/socialtables component costs — see memo-02).
3. People time: person-days × loaded day rate — for a consultancy this is also opportunity cost (days not billed or not selling).
4. Contingency: 10%.

Rules of thumb from the exhibit industry (directional): booth events total ≈ 3× the floor-space cost (the long-standing EXHIBITOR rule); non-booth sponsorships: plan 1.5–3× the rights fee all-in once travel, hospitality, and activation are loaded (synthesis — flagged in memo-02; the legacy IEG survey found sponsors spending ~$2.20 activation per $1 of fees). An owned executive dinner runs $5–20K fully loaded ($3.5–7K for ~10 guests standalone; $10–18K in conference cities) — vendor-published ranges.

Expected value: the pre-commit model

For a high-ACV, low-deal-count business, cost-per-lead is meaningless. Model expected value:

EV(pipeline) = expected qualified conversations
             × conversation → opportunity rate
             × average first-engagement value
EV(revenue)  = EV(pipeline) × win rate

Run conservative / base / stretch in the pack worksheet, using our CRM history for the rates; until history exists, use the playbook defaults and label them Inference (SOP §12). Two published anchors for the rates (vendor data — ceilings, not planning numbers): event-sourced leads convert to opportunity at rates far above other channels (HockeyStack 2025 panel data; Vendelux 2026), and dinner guests convert to qualified opportunity at 30–50%.

Worked example (illustrative): a $15K dinner with 12 ICP execs. Base case: 12 conversations × 30% opp rate = 3.6 opps; at $150K average engagement ≈ $540K expected pipeline (36× spend); × 25% win rate ≈ $135K expected revenue (9× before delivery costs). The same $15K on a sponsorship yielding 3 unbooked conversations at a 10% opp rate returns EV below cost. The dominating variable is ICP conversations held — which is why pre-booked meetings gate everything.

The investment test (all three must pass on the base case)

Fail any test → downgrade the mode and re-run: sponsor→attend-and-hunt, booth→meeting space + dinner, big city→home city. The cheapest de-risking is a $1–2K scout pass the year before sponsoring anything.

Payback framing for a 6–16 week (and longer) cycle

• Judge events on 90-day pipeline created (leading indicator), but cohort-track attendees for 12–18 months — enterprise-ACV deals mature slowly, and event cohorts keep converting past T+180 (memo-05).
• One-line payback statement in every budget sheet: "If we win N engagement(s) of ~$X, this event pays back Y×." If N > 2 for payback, the event is priced wrong for our motion.
• Tag every attendee/contact with the event_id campaign in the CRM at capture and never overwrite the original-source field — the T+90/T+180 readouts are only as honest as this hygiene (chapter 11).

The brand exception (capped)

Some events are legitimately reputation plays — a stage that positions INTO's "production over pilots" POV, an analyst room. Rules:

1. ROO-scored (pre-declared objectives, achievement-scored after), never a retroactive story.
2. ≤15–20% of the annual event budget may be justified on ROO alone (synthesis — flagged), and each such event still carries a named-account contact target.
3. A brand sponsorship with no access mechanism (no stage, no list, no side-event rights) is a logo purchase — decline.
4. Named by the CMO in the brief as the exception; second year requires re-justification with evidence.

Kill criteria and sunk-cost discipline

• Pre-commit: every scorecard names kill criteria; the budget sheet records the last-refund date. Kill criterion trips inside the window → withdraw.
• T-3w gate: pre-booked meetings < 50% of quota → downgrade (fewer travellers, cancel activation spend, keep the booked meetings).
• Post-event, the two-strike rule: an event that misses its pipeline floor (< 3× at T+90 with no compensating influence at 12 months) twice consecutively is cut from the portfolio. Walking away is discipline, not failure.
• Sunk fees never justify incremental spend. The sponsorship already paid is gone; only the marginal EV of travel + activation decides whether anyone boards a plane.

Approval mechanics (binding)

• Within the campaign envelope: CMO approves the budget sheet.
• Over the envelope: Class C — CMO + Jay sign-off before any commitment (TEAM.md §6.4, SOP §8). "Commitment" includes verbal yeses to organisers — the Event Manager takes sponsorship pitches away, evaluates, recommends (role SKILL.md); nobody approves on the call.
• Budget actuals land in the pack at T+7; variance >15% gets a written why in the post-event report.