What is Outreach gross margin trajectory through 2028?

Direct Answer

Outreach gross margin trajectory through 2028: 75-80% in FY26 → 73-78% in FY27 (slight compression from AI compute cost) → 76-81% in FY28 (compute optimization + scale benefits). The four pressure points: (1) AI compute cost from Smart Email Assist + Kaia (Anthropic + OpenAI inference fees scale with attach rate), (2) professional services dilution as enterprise tier scales, (3) cloud infrastructure (AWS) cost pressure, and (4) localization + data residency costs for international expansion.

The four levers + the FY28 outlook + comparable benchmarks.

The Numbers — Gross Margin Trajectory

• 2022 estimated: ~76-78% (pre-AI rollout, pure software margin)
• 2023 estimated: ~75-77% (AI investment starting)
• 2024-25 estimated: ~75-79% (Smart Email Assist + Kaia GA, compute cost ramping)
• 2026 estimated: 75-80% (AI attach growing, compute optimization ongoing)
• 2027 target: 73-78% (peak AI compute pressure, professional services dilution)
• 2028 target: 76-81% (compute optimization + scale benefits, AI margin recovers)

The 4 Pressure Points

• Pressure 1: AI compute cost — Smart Email Assist + Kaia depend on Anthropic + OpenAI inference. Per-1000-AI-emails cost ~$0.50-2.00; per-Kaia-call cost ~$0.30-1.20. Scales with attach rate, dilutes margin if not priced through.
• Pressure 2: Professional services dilution — Enterprise tier scaling brings $20-60M services revenue at 25-40% margin (vs 80%+ software margin). Each $10M services adds $1-2M gross profit but dilutes blended margin 2-3 points.
• Pressure 3: Cloud infrastructure (AWS) — Outreach runs on AWS; activity-graph storage + processing costs scale with customer base. Estimated $15-30M annual AWS bill by FY27.
• Pressure 4: Localization + data residency — EMEA + APAC + LATAM expansion requires regional AWS deployments (data residency for GDPR, India localization, Brazil LGPD). Adds 1-2% to infra cost.

The 4 Levers To Defend Margin

• Lever 1: AI compute optimization — fine-tune smaller models (Haiku-class) for 70-80% of Smart Email Assist requests; reserve Sonnet/Opus-class only for complex personalization. Reduces per-email cost 40-60%.
• Lever 2: Consumption pricing pass-through — Smart Email Assist consumption pricing ($5-15/user/mo or per-1000-emails) directly passes AI compute cost to customer. Margin-neutral if priced correctly.
• Lever 3: Bundled enterprise pricing — Enterprise tier $190-230/user/mo includes AI add-ons at marginal cost; customer wallet expansion offsets compute cost growth.
• Lever 4: Vertical solutions premium — FinServ + Healthcare + Industrial verticals at 20-30% premium pricing offset compute + localization costs.

Comparable SaaS Gross Margin Benchmarks (FY26)

• HubSpot: ~85% (mature multi-product SaaS)
• Salesforce: ~75% (services-heavy, hardware-touching)
• Datadog: ~80% (compute-intensive but well-optimized)
• Snowflake: ~67% (compute-pass-through model)
• Asana / Monday: ~88-90% (pure software, no AI compute drag)
• MongoDB: ~71% (compute-intensive)
• Outreach FY26 estimated: 75-80% (in line with peers)

Why FY27 Is The Peak Pressure Year

• Smart Email Assist attach climbing toward 50-60% target (per q1736) → AI compute cost scales fastest
• Enterprise tier services revenue ramping → margin dilution acute
• International expansion + data residency → infra cost adds
• BUT compute optimization investments not yet fully realized → benefit lag
• Result: gross margin compresses 2-3 points before recovery in FY28

Why FY28 Recovers

• Fine-tuned smaller-model deployment fully shipped → AI cost per request drops 40-60%
• Smart Email Assist consumption pricing fully optimized → margin-neutral at scale
• Cloud infrastructure tier optimization (Reserved Instances, Savings Plans, Spot) → 15-25% AWS cost reduction
• Vertical solutions premium pricing rolled out → margin lift 1-2 points
• Combined: gross margin recovers to 76-81% by FY28

What Could Break The FY28 Recovery

• AI compute cost rises faster than optimization (Anthropic + OpenAI raise rates)
• Smart Email Assist attach plateaus → consumption pricing doesn't materialize → AI cost stuck on Outreach P&L
• International expansion costs higher than projected (data residency requirements expand)
• Enterprise tier services revenue ratio rises faster than expected → blended margin dilution accelerates
• Cloud infrastructure providers (AWS) introduce AI-specific surcharges

A Markdown Table — Gross Margin Driver Sensitivity FY27

A Mermaid Diagram — Gross Margin Trajectory

Bottom Line

Outreach gross margin trajectory through 2028 is a "trough then recovery" pattern: 75-80% in FY26 → 73-78% in FY27 (peak AI compute + services pressure) → 76-81% in FY28 (compute optimization + scale benefits). Honest call: peer-level margin profile (Datadog, MongoDB territory), acceptable for IPO at $1.5B+ valuation.

The risk is FY27 trough deeper than expected — if margin drops below 73%, IPO story shifts to "growth + margin recovery in FY29" which compresses valuation. The compute optimization investments in 2026-27 are the gate. (See also: q1729, q1737, q1746)

Tags

Outreach, gross-margin, fy28-outlook, cogs, ai-compute-cost, professional-services, cloud-infrastructure, unit-economics, ipo-prep, rule-of-40

FAQ

Why is FY27 the peak pressure year for Outreach's gross margin? FY27 stacks several costs at once: Smart Email Assist attach is climbing toward its 50-60% target so AI compute scales fastest, enterprise services revenue ramps and dilutes margin, and international data residency adds infra cost.

Meanwhile the compute optimization investments have not yet fully paid off, so there is a benefit lag. The result is a 2-3 point compression to a 73-78% trough before recovery in FY28.

What does the AI compute for Smart Email Assist and Kaia actually cost? Smart Email Assist and Kaia run on Anthropic and OpenAI inference, costing roughly $0.50-2.00 per 1,000 AI emails and $0.30-1.20 per Kaia call. These costs scale with attach rate and dilute margin by 2-4 points if not priced through.

The main lever is fine-tuning Haiku-class models for 70-80% of requests and reserving Sonnet- or Opus-class models for complex personalization, cutting per-email cost 40-60%.

How much does professional services drag down the blended margin? Enterprise tier scaling brings $20-60M of services revenue at only 25-40% margin, versus 80%+ on software. Each $10M of services adds $1-2M of gross profit but dilutes the blended margin by 2-3 points. This pressure is most acute in FY27 as the enterprise tier ramps.

How does Outreach's FY26 gross margin compare to other SaaS companies? At an estimated 75-80% in FY26, Outreach is in line with its peers. HubSpot runs about 85%, Datadog about 80%, Salesforce about 75%, and Snowflake about 67%, while pure-software players like Asana and Monday hit 88-90% without AI compute drag.

Outreach lands mid-pack because of its AI inference and services mix.

What could prevent the projected FY28 recovery to 76-81%? The recovery breaks if AI compute rises faster than optimization because Anthropic or OpenAI raise rates, or if Smart Email Assist attach plateaus so consumption pricing never offsets the cost. International data residency costs running higher than projected, services revenue ratio climbing faster than expected, and AWS introducing AI-specific surcharges are the other named risks.

Any of these could leave margin stuck below 75% and pressure the IPO valuation.

Sources

• Https://www.outreach.io/about
• Https://www.outreach.io/products/smart-email-assist
• Https://www.bvp.com/atlas/state-of-the-cloud-2026
• Https://www.iconiqcapital.com/insights/state-of-saas
• Https://openviewpartners.com/saas-benchmarks/
• Https://www.crunchbase.com/organization/outreach-corp
• Https://www.gartner.com/en/sales/research