The chart problem
You’ve seen the chart. Spend is up. ROAS is down. There’s a spike on Tuesday and a dip on Thursday. Now what? If you’re the person who manages the campaigns, you probably have context. You know the spike was because you launched a new creative. You know the dip was because you paused an underperforming ad set. The chart is a reminder of decisions you already made. But for everyone else who sees that chart — your manager, your client, your CEO — it’s just lines going up and down. They don’t have the context. They can’t tell the difference between a healthy fluctuation and a problem that needs attention. This is the fundamental limitation of chart-based reporting: charts show you what happened, but they don’t tell you what it means.What AI analysis adds
When ChartCastr delivers a scheduled Meta Ads report, it doesn’t just send a chart. It sends the chart with an AI-generated analysis that interprets the data in plain English. The AI examines the numbers, identifies what changed, and explains why it might matter. It’s not a template — it’s a dynamic analysis generated from your actual data every time a report runs.Before AI (chart only)
Your team sees a chart showing ROAS dropping from 2.4x to 1.8x over the past week. Someone Slacks: “What happened to ROAS?” Your ads manager has to stop what they’re doing to explain.After AI (chart + analysis)
The chart arrives with this summary:“ROAS declined from 2.4x to 1.8x this week. The primary driver was the Prospecting — US Broad campaign, where CPA increased 34% (17.15) without a corresponding increase in conversion value. This campaign historically sees CPA spikes during the first week of the month due to increased competition from advertisers refreshing monthly budgets. The Retargeting campaign maintained 3.1x ROAS, consistent with prior weeks.”Nobody needs to ask what happened. The analysis did the explaining. The ads manager doesn’t get interrupted. The team understands the data. Decisions can be made faster.
Why context makes the AI smarter
ChartCastr’s AI doesn’t work in a vacuum. When you have Streams enabled alongside your scheduled Pulses, real-time events are captured and stored as context. This means the AI has access to granular events — not just aggregate numbers — when writing its analysis:“Lead volume dropped to zero on Wednesday afternoon. This coincides with the ‘Summer Sale — Lead Form’ ad being flagged for policy review (captured via Streams at 2:15 PM). The ad was reinstated Thursday morning, and lead volume recovered to normal levels by Friday.”Without Streams providing that event context, the AI would only see the aggregate drop and have to guess at the cause. With it, the analysis is grounded in actual events from your account.
What the AI actually analyses
For Meta Ads reports, the AI examines:- Period-over-period changes in key metrics (spend, ROAS, CPA, CTR, impressions, conversions)
- Campaign-level variation — which campaigns drove aggregate changes up or down
- Anomalies — spikes, dips, or sudden changes that stand out from recent trends
- Stream context — real-time events (leads, ad rejections, processing) captured during the reporting period
- Trend direction — whether metrics are improving, declining, or stable over time
Who benefits from explained reports
| Role | What they get from AI analysis |
|---|---|
| Ads manager | Confirmation of expected patterns, early warning of unexpected ones |
| Marketing lead | Plain-English summary for leadership and cross-functional meetings |
| Agency client | Transparency into what’s happening without needing platform expertise |
| Finance | Context for spend changes beyond “marketing spent more this month” |
| CEO | One-paragraph summary of whether ads are working |
Getting started
- Connect Meta Ads as a source
- Create a Pulse — select metrics, set schedule, choose destination
- Enable AI summaries on the connection (on by default)
- Optionally add Streams for richer context in AI analysis