Every sales manager has lived this moment: you're in a Monday standup, screen-sharing your Salesforce dashboard, and someone asks "wait, is this data current?" You click Refresh. The numbers change. Confidence evaporates.
Salesforce dashboards are powerful for visualising CRM data, but their auto-refresh behaviour has real limits — particularly when the data you actually care about lives outside Salesforce itself. This guide covers exactly how Salesforce dashboard refreshes work, where the ceiling is, and what teams use instead when they need live, scheduled, or database-driven data.
How Salesforce Dashboard Refresh Actually Works
Salesforce Classic and Lightning Experience handle dashboard refresh differently, and it's worth being precise about this.
In Salesforce Classic:
You can schedule a dashboard to refresh on a set cadence — hourly, daily, or weekly. The schedule triggers a background job that re-runs the underlying reports and stores a snapshot. When someone opens the dashboard, they see that cached snapshot, not a live query.
In Lightning Experience:
The manual Refresh button re-runs the underlying reports on demand. Scheduled refresh is also available but operates the same way: it runs the reports and caches the result. You're always looking at data as of the last refresh, not real-time data.
Key constraint: Salesforce dashboards reflect data stored in Salesforce. If your true source of truth is a PostgreSQL database, a data warehouse, or any external system, a Salesforce dashboard cannot pull from it directly.
Step-by-Step: Scheduling a Dashboard Refresh in Salesforce Classic
If you're on Classic and want to schedule a refresh:
Note that filters applied to a dashboard are not carried over when a scheduled refresh runs. The refresh re-runs the underlying reports with their default filter settings.
The Limitations You'll Hit Quickly
Once you try to build anything beyond basic CRM reporting, Salesforce dashboard refresh runs into real friction:
1. No sub-hourly refresh. The minimum scheduled interval is one hour. If you're tracking intraday sales activity, pipeline changes in real time, or monitoring for specific thresholds (like "alert me when conversion rate drops below 20%"), hourly is too slow.
2. Salesforce data only. Reports and dashboards are built on Salesforce objects: Leads, Opportunities, Accounts, Cases. If your product usage data is in Postgres, your billing data is in Stripe (synced to a database), or your operational metrics live in a data warehouse — none of that is visible to a Salesforce dashboard natively.
3. No conditional refresh. There's no way to say "only refresh and send this dashboard when revenue this week is below last week." Every refresh is unconditional.
4. Row limits on reports. Salesforce reports cap at 2,000 rows in the dashboard view. Complex aggregations across large datasets hit the limit and silently truncate results.
5. Cost. Salesforce licenses are expensive, and access to advanced analytics features (like CRM Analytics, formerly Einstein Analytics) adds significant per-seat cost.
When Your Data Lives Outside Salesforce
This is where most teams eventually land: the data they actually want to monitor is in a database, not in the CRM.
Consider a SaaS company tracking:
None of that lives in Salesforce natively. You can use Salesforce Connect or data import jobs to push some of it in, but that adds pipeline complexity, sync delays, and cost.
The alternative: build your dashboards directly on top of the database where the data already lives.
Building Auto-Refreshing Dashboards from Your Database
AI for Database is built specifically for this use case. You connect it to your database — PostgreSQL, MySQL, Supabase, BigQuery, and others — and build dashboards using plain-English queries instead of SQL.
Here's what that looks like in practice. You'd write something like:
"Show me daily signups grouped by acquisition channel for the past 30 days"
AI for Database translates that to SQL, runs it, and returns a chart. Then you set that query to refresh automatically — every hour, every 15 minutes, once a day — whatever cadence makes sense for your data.
For reference, the underlying SQL it generates would look something like this:
SELECT
acquisition_channel,
DATE(created_at) AS signup_date,
COUNT(*) AS signups
FROM users
WHERE created_at >= NOW() - INTERVAL '30 days'
GROUP BY acquisition_channel, DATE(created_at)
ORDER BY signup_date DESC;You don't write this. You write the plain-English question. The dashboard updates on schedule, and anyone on your team can open it and see current data without knowing anything about SQL.
Comparison: Salesforce Dashboards vs Database Dashboards
Feature | Salesforce Dashboard | AI for Database Dashboard
Data source | Salesforce objects only | Any connected database
Minimum refresh interval | 1 hour | Custom (minutes to daily)
Real-time data | No | Yes (on refresh)
Conditional alerts | No | Yes — action workflows
SQL knowledge needed | No | No
Row limit | 2,000 rows | None
Requires CRM license | Yes | No
The right tool depends on what data you're visualising. For pipeline and CRM data, Salesforce is fine. For product metrics, revenue analytics, operational data, or anything that lives in a database — you want a tool that connects directly.
Setting Up Conditional Alerts on Your Database Data
Beyond scheduled refresh, AI for Database supports action workflows — conditions you define that trigger notifications automatically.
Examples:
This is the pattern that replaces manual dashboard-watching. Instead of someone refreshing a Salesforce dashboard every morning and eyeballing whether things look right, you define the threshold once and get notified when it's crossed.
The workflow connects directly to your database — no triggers, no stored procedures, no DBA involvement.
The Bottom Line
Salesforce dashboard refresh is useful for what it is — a way to keep CRM-based dashboards reasonably current. But it hits hard limits when you need sub-hourly refresh, data from external databases, or conditional alerting.
If your metrics live in a database — and for most companies, the most important ones do — it's simpler to build dashboards that connect directly to that database, query in plain English, and refresh on a schedule you control.
Try AI for Database free at aifordatabase.com and have your first auto-refreshing database dashboard running in under 10 minutes.