Comparison
Switchboard vs The DIY stack
For teams considering gluing together a self-hosted inbox, a standalone ESP, a bot builder, and Zapier — here is the honest math on what you give up for the flexibility you gain.
The short version
Is Switchboard right for you?
The DIY stack is real and legitimate — a self-hosted open-source helpdesk, an open-source ESP, a chatbot builder, and Zapier / n8n for glue. You keep full data ownership, pay mostly infra and engineering time.
Switchboard bundles the same underlying capabilities into one managed platform. You lose some customization flexibility; you gain one vendor, one bill, one security posture, and one on-call rotation.
The decision comes down to whether your team has a full-time DevOps / security engineer to keep the DIY stack alive. If yes, DIY can work. If no, the glue is where the work actually lives — and it is never done.
Pricing
What you actually pay
DIY figures assume self-hosted open-source tooling (~$200/mo infra for a small team) plus the engineering cost to wire it together. Labor at $100/hr fully loaded, 20% of one engineer = $8k/mo. Switchboard is flat and includes all of it.
| Team size / scenario | Switchboard | The DIY stack |
|---|---|---|
| Small team, getting started | $49/mo | $200 infra + $8k eng time ≈ $8,200/mo TCO |
| Growing team, 3 integrated tools + Zapier glue | $199/mo | $500 infra + $12k eng time ≈ $12,500/mo TCO |
| Team scale, HA deployment + compliance | $499/mo (plus BYOC if self-host required) | $1,500 infra + $20k eng time ≈ $21,500/mo TCO |
Feature matrix
Side-by-side capabilities
| Capability | Switchboard | The DIY stack |
|---|---|---|
Unified inbox (omnichannel) Each tool has its own inbox; unification lives in Zapier | ||
AI agents built-in Needs its own bot builder; routing via Zapier is fragile | ||
Marketing automation (email drip + campaigns) Separate ESP product | ||
Lead scoring Separate product or custom scripts | ||
Shared contact record (support + marketing) Contact records live in N different databases | ||
Flat platform pricing (no per-seat) | ||
No per-resolution AI billing | ||
Self-hosted option | ||
White-label | ||
Native mobile apps Varies per tool; rarely unified | ||
SLA policies | ||
CSAT surveys | ||
Website chat widget | ||
Instagram / Facebook / WhatsApp / SMS / Email Possible but each channel is a separate integration | ||
Public API + webhooks | ||
Help center / knowledge base | ||
Roles + permissions | ||
Audit log Per-tool logs, not consolidated | ||
Per-tenant rate limits Would require custom implementation per tool | ||
HIPAA-ready deployment |
Where The DIY stack wins
A DIY stack wins on ceiling, not floor. If your use case has edges that no SaaS product models — a custom data model, a compliance regime that mandates specific deployment controls, a pricing mechanic that vendors will not accommodate — you need to own the code. Open-source tools like self-hosted inboxes and ESPs are genuinely capable. You keep every byte of customer data on your infrastructure. You pay no per-seat fees. You can extend anything. If you have a platform team that already runs production infrastructure, DIY is legitimate and sometimes the right answer.
Where Switchboard wins
Switchboard consolidates the inbox, the AI agents, the marketing automation, the KB, and the operator multi-tenancy into one system with one contact record. No Zapier glue, no 'our ESP knows a lead exists but our helpdesk does not', no on-call rotation for a stack nobody documented. One bill, one security review, one SSO integration, one audit log. For teams where the support platform is not the product, the opportunity cost of running DIY — measured in weeks of engineering that could have gone into the actual product — almost always exceeds the SaaS fee.
Decision
Which one should you pick?
If you have a full-time DevOps engineer and you treat your support stack as a platform investment → pick DIY — you'll make it work and own the upside
If you need deep custom data model or a deployment regime no SaaS supports → pick DIY — this is what self-hosted OSS is for
If your team is the product, not the support stack → pick Switchboard — engineering time is too expensive for this glue work
If you want self-hosted but not the maintenance burden → pick Switchboard BYOC — we manage the stack on your cloud
If you are 2-10 people and DIY sounds like fun → pick Switchboard — fun until it is an outage at 2am
FAQ
Common questions
Can we self-host Switchboard?+
Yes. Switchboard offers self-hosted and BYOC (bring-your-own-cloud) deployment. BYOC means the stack runs on your AWS / GCP / Azure account, under your data governance, but Switchboard manages upgrades and operations. Best of both worlds for regulated teams.
What open-source is Switchboard built on?+
Switchboard is a managed platform — customers do not see the underlying components, and we keep the list inside the product. We commit to portability of your data (exports in standard formats) and reasonable notice on any substrate changes.
What if we outgrow Switchboard?+
Your data exports — contacts, conversations, tags, custom attributes, attachments — in standard formats. No lock-in through proprietary data shapes. If you ever decide to go DIY later, the door is open.
Is DIY actually cheaper long-term?+
Honestly, rarely. The infra is cheap; the engineering time is not. Most teams underestimate the glue work — Zapier stops working, auth tokens expire, the bot builder deprecates a webhook format, the ESP changes its API. A 20% fractional DevOps budget usually exceeds the SaaS fee by a factor of ten.
Can we migrate from a DIY stack to Switchboard?+
Yes. Contact records, conversation history, tags, KB articles, and canned responses all import via standard CSV / JSON / API. Typical DIY-to-Switchboard migration is 1-2 weeks including validation.
Ready to move off per-seat, per-resolution billing?
Flat pricing. Every channel included. No seat math.