Make Review

Walk into the next ops review. Someone needs branching logic, an Iterator over a CSV, a Router firing three downstream paths. Zapier handles two and the third painfully. Make handles all three on the $9/month Core plan.

That''s the strategic bet. Make sits between Zapier and n8n in a way that matches a real org chart — ops people who think in flowcharts but won''t run Node.js workers. Twelve years in-market as Integromat, rebranded 2022, Celonis-owned since 2020. Three signals this isn''t the usual vendor risk.

The yellow flag is parent alignment. Celonis bought Integromat for process-mining synergy, and roadmap priority shifts when the parent has a different ICP. But the install base is too large to abandon, and operations-based pricing fits this audience. Pilot three branching workflows your team currently fakes in multi-step Zaps. If your ops manager stops asking for "just one more Zap," you''ve found your standard.

Make is built around operations, not seats. That single design choice shapes everything. A 50-person ops team pays for what scenarios consume, not who logs in. Compare Zapier''s task-plus-seat hybrid or Workato''s recipe-credit model. Make''s unit matches how automation scales — volume grows faster than headcount.

The canvas earns the claim. Iterator, Aggregator, and Router are first-class flow primitives, not workarounds. Fan-out, fan-in, batch processing, conditional branching — all modeled without code. Make Bridge gives the dev-to-prod scenario promotion mature ops teams need.

The strategic constraint is portability. Scenarios export as blueprint JSON, but that JSON only runs on Make. Compare n8n where workflows run on your own infrastructure indefinitely. AI Agents is the right next bet — flows that interpret natural-language goals — but the catch is it''s maturing while Zapier ships its own agent layer on a similar timeline. The moat is convenience plus integration depth, not impossibility-to-replicate.

Make doesn''t price per seat. That''s the whole TCO story, and it changes the budget.

Core: $9/month for 10,000 operations. Pro: $16/month. Teams: $29/month. Enterprise contact-sales. Free tier ships 1,000 ops/month indefinitely. 50 people on Teams × $29 × 12 = $17,400/year baseline, but seat count isn''t the real cost — operations consumption is. Compare Zapier''s task-plus-seat model where 50 people would clear $50K+ before task volume.

The catch is ops forecasting. An Iterator over 1,000 records makes 1,000 ops. A Router with 3 branches and 4 modules adds 12 ops per trigger. Operations math compounds in ways finance teams underestimate by 30-40% in six months. Build a 90-day baseline before signing annual, budget a 25% overage buffer. Compare Workato''s recipe-credit model where complexity hides in tier pricing — Make''s is honest, less predictable. Self-serve cancellation, no auto-renewal trap.

Week one with Make is drawing scenarios that look right and don''t run. The canvas is honest — modules turn red, the execution log shows which item broke, rerun-from-failure is one click. Zapier''s task history loses data shape; Make''s inspector is a real debugger.

The pairing that unlocks the platform is Iterator with Aggregator. Iterator splits an array; Aggregator collects it back. Once that lands you model real batch processing — fetch 50 Salesforce records, process each, recombine into a Slack digest — where Zapier needs three Zaps. Data Stores give key-value state.

The friction is the ops meter. Every Iterator pass is an op. Every Router branch is an op. Pro ships 10,000 ops for $16/month, but a daily Salesforce-to-Slack scenario over 200 records eats 6,000. The catch is you can''t forecast from the canvas alone. By year two you write scenarios with ops cost in mind.

The empty state when you first open Make. Complaint up front. New scenario screen feels 2018 — scattered helper text, a tutorial pane in the way, modules organized by category instead of frequency. Rougher than Zapier''s onboarding.

But on the canvas, the polish is real. Modules snap cleanly. Right-click gets you bundle inspection — every input, every output, every iteration of every run. The History panel retains executions 7 days on paid tiers with full replay. Run-once-from-here as a shortcut saves your week. Compare Pipedream where you fight the UI for your last execution.

Mobile parity is the honest yellow flag. The mobile app shows scenario status, alerts, and recent runs — useful at 9pm when something broke. But you can''t edit scenarios on mobile. A real product, not an afterthought. By month three you''ll have 40 scenarios, hit the ops meter once, found one connector bug.

Founded 2012 as Integromat, rebranded Make in 2022, acquired by Celonis in 2020. Twelve years in a category that ate most of its cohort — iPaaS startups from 2014-2018 are mostly gone. Make ships at scale alongside Zapier and Workato.

What worries me is the AI Agents bet. Make announced AI Agents in 2024. Zapier shipped Zapier Agents in parallel; n8n''s self-hosted AI gives engineering teams a more flexible substrate. Make''s bet may age poorly if the agent layer becomes the integration point.

The exit story is the yellow flag. Scenarios export as blueprint JSON, but that JSON doesn''t run anywhere else. Compare n8n where workflows run on your own infrastructure indefinitely. Make''s lock-in is integration depth plus the visual model your team has internalized — real, not catastrophic. For a Celonis-owned platform with 3,000+ integrations, survival outweighs exit concern.