Email and messaging
Slack automation consultant: alerts people act on, and approvals that happen in Slack
SBD builds Slack apps and automations for teams of 1 to 1,000 people: alerts that reach the person who can act, approvals handled with a button in Slack, and summaries posted where the team already works. Workflow automation builds start at $1,000, with the fixed price set after a $500 Automation Audit that is credited to the build.
- Tool
- slack.com
- Usually paired with
- HubSpot, Airtable, n8n, Zendesk
- Timeline
- Most builds live in 2 to 8 weeks
- Owned by
- You, with accounts, credentials and code in your name.
What we build on Slack
- Post alerts with context to the person who owns the record
- Thread updates so one deal or order is one conversation
- Let people act from Slack: approve, assign, reply
- Log what happens in Slack back to the record
What we sync
Workflows we are asked for
- HubSpot to Slack alerts with context
- Approval buttons in Slack
- Daily digests instead of a firehose
- Slack replies logged to the CRM
Full write-ups of Slack integrations
Tools Slack usually needs to talk to
-

Slack and HubSpot
The CRM and marketing platform most B2B teams of 10 to 1,000 run on.
-

Slack and Airtable
A relational database with a spreadsheet face, used as the operations base for teams who have outgrown sheets.
-

Slack and n8n
The automation platform we build most systems on: code where needed, visual where not, cloud or self-hosted.
-

Slack and Zendesk
The helpdesk most support teams of 10 to 1,000 run on.
-

Slack and Linear
Issue tracking for product teams, fast and opinionated.
-

Slack and Salesforce
The enterprise CRM, with a broad API that needs careful design.
What a Slack automation from SBD does
Slack is where most of the teams we work with already spend the day, which makes it the natural place for automations to report and to ask.
The risk is noise: a channel that receives every event from every system is a channel nobody reads. Our Slack work is about routing the right event to the right person, in a form they can act on without leaving Slack.
Typical builds are alerts from operational systems sent directly to the owner of the problem; approval requests with buttons for approving or rejecting a refund, a bill, a discount or an AI agent's proposed action, each recorded with who decided and when; daily or weekly summaries that replace a report nobody opened; and slash commands or shortcuts that create records in the CRM, helpdesk or project tool from a conversation.
We have built this. For a B2B sales team we automated lead capture, enrichment, AI qualification and CRM sync on n8n, with call summaries written back to the CRM and posted to Slack so the team sees them where they work. Our rental portfolio agent fleet, still in build, uses the same idea: routine actions approved in bulk, risky ones always sent to a person.
When a custom Slack app beats Workflow Builder and the app directory
A good fit
Slack's Workflow Builder and the apps in its directory cover forms, simple approvals and notifications from popular tools. If they do the job, use them; we will say so in the audit.
A custom Slack app is worth building when
- An approval in Slack must change something in another system, with a record of who approved it.
- Alerts must be routed by rules such as account owner, region or severity.
- An AI agent proposes actions and a person must approve them before anything happens.
- Several systems report into Slack and you want one app, one set of permissions and one place to change routing.
- You need to read message history at volume for summaries or search, which is limited for some app types.
Slack rate limits and the 2025 change to history reads
- 01
Slack rates its Web API methods in tiers, per method, per workspace, per minute: Tier 1 allows 1 or more calls a minute, Tier 2 20 or more, Tier 3 50 or more and Tier 4 100 or more. Some methods have special limits; chat.postMessage generally allows one message per second per channel, with a workspace-wide limit on top. When a limit is hit Slack returns 429 with a Retry-After header in seconds, and other methods are unaffected. An alerting system that posts a burst of events to one channel must queue them, or batch them into one message.
- 02
In May 2025 Slack changed limits for some apps.
For commercially distributed apps that are not approved for the Slack Marketplace, created after 29 May 2025 or newly installed after that date, conversations.history and conversations.replies moved to 1 request a minute with at most 15 objects per request. Slack states that internal, customer-built applications are not affected and keep 50 or more requests a minute with up to 1,000 objects. This is one reason we build Slack apps inside your own workspace, as internal apps you own.
Slack Events API, Socket Mode and app permissions
Slack apps receive events either over a public HTTPS endpoint or through Socket Mode.
Over HTTP, your app must respond with a 2xx within three seconds. Failed deliveries are retried three times: almost immediately, after one minute and after five minutes, each with an x-slack-retry-num header. If more than 95% of deliveries fail within 60 minutes, Slack temporarily disables the event subscription. Event deliveries are capped at 30,000 per workspace per app per 60 minutes, after which Slack sends an app_rate_limited event. Our handlers acknowledge at once, verify the request using the signing secret, skip retries they have already processed, and do the work from a queue.
Permissions are set by OAuth scopes when the app is installed. We request only the scopes each feature needs, use a bot token rather than a person's token so the app keeps working when people leave, and document every scope so your workspace admins can review it.
- Interactive buttons must be answered within Slack time limits, so approvals acknowledge first and update the message when the work completes.
- Configuration stores channel and user IDs, since channels get renamed.
- Messages that contain customer data are kept out of public channels by the routing rules themselves, so nobody has to remember.
How a Slack engagement with SBD works
- Step 1$500Automation AuditCredited to the build
- Step 2from$1,000Workflow automationFixed after the audit
- Step 3from$3,000Internal toolsFixed after the audit
- Step 4from$50Care planA month, after sixty days of hypercare
Every engagement starts with an Automation Audit at $500, credited in full to your build. For Slack we list every system that posts into your workspace today, which channels people read, and where decisions are being made by email or in someone's head. You get a ranked list of fixes with fixed prices and a blueprint for the first build.
Workflow automation builds start at $1,000, internal tools at $3,000 and multi-module systems at $5,000. We build Slack apps with n8n, Make, Supabase or custom code on Cloudflare or Vercel, installed in your workspace as an app you own. Work runs in two-week sprints, sixty days of hypercare follow go-live, and care plans start at $50 a month. See the pricing page.
Your Slack subscription is paid by you to Slack. We do not resell licences, and we do not publish your app to the Slack Marketplace unless you ask us to.
What you own when the build is finished
Everything below is in your name when the build ends.
- You own the Slack app configuration, its bot token and signing secret, the code and the logs, all created in your workspace and your hosting accounts.
- You receive a routing document that says which events go to which person or channel and why, a list of every OAuth scope the app holds with the reason for each, and a runbook for changing routes without breaking approvals.
- After go-live we review alert volumes with your team at the end of hypercare.
- Alerts that nobody acted on are either removed, batched into a summary, or sent to someone else.
- The aim is a workspace where an automated message is worth reading, because the ones that are not have been switched off.
- If you later move from Slack to Microsoft Teams, the routing document and the approval logic carry across; only the messaging layer changes.
Slack at enterprise scale
Larger organisations typically run Slack across many teams, sometimes several workspaces, with an IT or collaboration team that reviews every app before install. Operations, finance, sales and engineering all want automations posting into it, which is why routing, permissions and noise become the main problems.
App review and scopes
Each app requests only the scopes it needs, with a written list your Slack admins review before install. Internal apps are owned by your workspace.
Approval records
Every approval or rejection in Slack is written to a log with the user, time, item and outcome, so the record survives message edits or channel deletion.
Routing rules
Who receives which alert is held in a routing table, so it can follow changes of owner without a developer.
Noise control
We batch low-priority events into summaries, send urgent ones directly to an owner, and review alert volumes after hypercare with your team.
Data handling
Customer and financial data is kept out of broad channels, linked rather than pasted where possible, and retention follows your workspace policy.
Engagements we take on
Already on Slack, an alerts channel nobody reads
Every system posts to one channel and real problems are missed. We route each alert to its owner, batch the rest into a daily summary, and track whether alerts get acted on.
Already on Slack, approvals by email
Refunds, discounts and bills wait in inboxes. We move the approval into Slack with buttons, apply the decision in the source system, and keep a log of every decision.
Already on Slack, agents that need a supervisor
An AI agent drafts actions that someone must check. We send each proposal to Slack for approval, with routine ones grouped and risky ones always sent to a named person.
Read next
Sources
- Slack Developer Docs, Rate limits (tiers, chat.postMessage, Retry-After, Events API cap), checked 29 October 2026
- Slack Developer Docs, Changelog: rate limit changes for non-Marketplace apps (29 May 2025), checked 29 October 2026
- Slack Developer Docs, Events API (three second response, retries, disabling, Socket Mode), checked 29 October 2026
Standards on every build
Validation, retries, reconciliation, audit logs and alerts come as standard, with credentials and accounts in your name. Read the integration standards
What people ask about Slack
Our channels are already muted. Will this help?
That is the problem this fixes. Fewer messages, each with the value, owner and reason attached, to the person who owns it.
Start with the one Slack workflow your team still runs by hand
Bring it to a discovery call. You leave with a yes or no on fit, an outline of the build and what we would do first.
Book the call











