Diagnose account performance
Ask about spend, ROAS, CPA, pacing, conversion tracking, top campaigns, wasted budget, and week-over-week movement across selected accounts, then inspect the answer as charts, tables, or KPI blocks.
AI marketing analyst
Copilot in Benly
Find winning patterns in your ad accounts, compare against competitors, and use pre-built prompts to turn Copilot into a 24/7 growth analyst.
Summary
The workflow at a glance
Copilot is Benly's AI marketing analyst for teams that want the answer and the evidence in one place. It starts from selected Meta, TikTok, or Google Ads accounts, workspace context, AdSpy competitor intelligence, saved prompts, custom instructions, and lightweight file context. Instead of returning a generic marketing paragraph, Copilot streams an analysis, calls the right platform or competitor tools, and turns the answer into reusable blocks such as KPI cards, tables, charts, ad galleries, forecasts, creative analysis, and quick actions.
- 01Select Meta, TikTok, or Google Ads accounts and ask questions across performance, creative intelligence, competitors, budget, forecasting, and visualization.
- 02Start from segment prompt cards or the full prompt library, then save the prompts that turn into repeatable team workflows.
- 03Move between your own account performance and AdSpy competitor context without rebuilding the brief from scratch.
- 04Get streamed answers as structured outputs: KPIs, tables, charts, ad galleries, creative analysis, forecasts, quick actions, and saved thread context.
01 / Workflow
What Copilot does
Copilot turns Benly's connected marketing data into a working analyst surface. The page starts with account selection, segment pills, suggested prompt cards, and an input that supports prompt recipes, saved prompts, custom instructions, DeepThink, and lightweight file attachment. Once a question starts, Copilot keeps the selected account and workspace context attached to the conversation, streams the response, calls the right platform or AdSpy tools, and renders the result as structured blocks instead of a wall of generic text.
- Use segment pills for Performance, Creative Intel, Competitors, Budget Watch, Forecast, Visualization, and Forecasting.
- Open the prompt library to browse recipe categories, inspect prompt details, bookmark favorites, or reuse saved prompts.
- Keep analysis tied to selected account context, workspace instructions, thread history, and competitor intelligence.
- Turn the answer into usable output with KPI cards, charts, tables, ad galleries, creative analysis blocks, forecasts, and quick actions.
02 / Workflow
From prompt to evidence
The important difference is that Copilot is not only a text box. Teams can start from guided prompts, refine the question, attach context, and let Copilot decide which data sources or competitor tools should be used. That makes the workflow more reliable for recurring analysis: a prompt can become a weekly account audit, a creative fatigue review, a budget watch routine, or a competitor comparison that the team reuses instead of rewriting every time.
- Prompt recipes cover audits, spend and ROI, creative analysis, audience breakdowns, trends, visualization, and competitive intelligence.
- Saved prompts let a team preserve the analysis questions that work instead of depending on one operator's memory.
- Custom instructions keep thresholds, preferences, custom metrics, and business memory close to the analysis.
Use cases
Where teams use it
Each use case starts from a real marketing question and ends with reusable context the team can bring into the next review, test, or brief.
Find creative winners
Analyze top ads, creative fatigue, video hook metrics, copy patterns, landing page performance, and the differences between strong and weak creatives so the next test starts from evidence.
Compare against competitors
Use AdSpy tools inside Copilot to review competitor ad galleries, platform distribution, hooks, formats, creative strategy, and the gaps between market behavior and your own account.
Turn prompts into repeatable analysis
Start from prompt cards or recipe templates, save prompts that work, and build a repeatable growth-analysis workflow the team can run during reviews, planning, and reporting.
03 / Workflow
How teams use it
Teams use Copilot when they need to move from a business question to an evidence-backed answer quickly. A growth lead can ask why ROAS moved, a media buyer can audit wasted spend, a creative strategist can compare hooks and formats, and a founder can review competitor activity without waiting for a manual report. The same surface can support quick checks and deeper investigations because the output can shift from short explanations to tables, visualizations, creative examples, or forecast-style blocks.
04 / Workflow
Why it matters
Most AI chat interfaces make marketers restate the context every time. That is a problem for performance teams because the context is the work: selected accounts, attribution windows, current spend, creative fatigue, competitor pressure, saved benchmarks, and the decision the team is trying to make. Copilot keeps the selected accounts, prompts, saved instructions, AdSpy competitor tools, and rendered analysis inside one workflow, which makes the output easier to trust and faster to reuse in the next decision.
FAQ
Frequently asked questions
Practical answers for teams comparing workflows, validating product fit, or deciding how this feature should sit inside their operating rhythm.
Is Copilot just another AI chatbot?
No. Copilot is designed as an analysis workflow for performance marketing teams, not as a blank general-purpose chatbot. The difference is context and output. Copilot can work from selected ad accounts, workspace instructions, saved prompts, thread history, AdSpy competitor intelligence, and lightweight file context. It also renders answers as structured Benly blocks such as KPI cards, charts, tables, ad galleries, creative analysis, forecasts, and quick actions. The goal is to help a team decide what happened, why it happened, and what to do next, without asking the user to rebuild the business context every time.
What data can Copilot analyze?
Copilot is built around selected Meta, TikTok, and Google Ads account context, plus workspace-scoped AdSpy competitor intelligence. That means a question can combine owned-account performance with market context: spend, ROAS, CPA, pacing, top campaigns, creative fatigue, hooks, formats, competitor ad galleries, and platform distribution. It can also use saved prompts, custom instructions, thread history, and lightweight attached files to keep the analysis closer to the team's real operating context. For public-facing claims, Benly keeps this conservative: Copilot is not presented as a replacement for every possible data warehouse, but as a connected marketing analyst surface inside Benly.
Can Copilot help me find winning creative patterns?
Yes. Copilot can help turn creative review into a more evidence-based workflow. A team can ask about top creatives, fatigue signals, video hook metrics, copy patterns, landing page performance, and the differences between strong and weak ads. When competitor context matters, Copilot can use AdSpy tools to compare hooks, formats, offers, and creative strategy across the market. The value is not simply labeling ads as good or bad; it is making the pattern visible enough that a strategist can decide what to refresh, what to scale, what to stop, and what to brief next.
Does Copilot include pre-built prompts?
Yes. The Copilot interface includes segment prompt cards, a full recipe library, prompt detail views, saved prompts, and usage tracking for saved prompts. This matters because most teams do not want to start every analysis from a blank text area. Recipes can help users start from common marketing questions such as account audits, spend and ROI checks, creative analysis, audience breakdowns, trends, visualization, and competitive intelligence. Saved prompts let the team keep the questions that produce useful analysis, so a strong workflow can become a repeatable operating habit instead of a one-off chat.
How does Copilot present answers?
Copilot streams answers and can render structured Benly blocks instead of only returning plain text. Depending on the question, an answer can become KPI cards, tables, charts, ad galleries, creative analysis, forecasts, quick actions, or other analysis views. This is important because marketing teams need outputs they can use in reviews and decisions. A paragraph can explain the finding, but a table can show the campaigns, a chart can show the movement, an ad gallery can show the creative pattern, and a forecast can make the next budget conversation easier.
Can Copilot compare my account with competitors?
Yes, when the question needs market context, Copilot can use AdSpy competitor intelligence alongside account analysis. A team can ask what competitors are launching, which hooks or formats appear repeatedly, how competitor creative strategy differs from their own, or where the market is leaning by platform and creative angle. The comparison is useful because performance data alone can explain what happened inside an account, but competitor context can explain what the team may be missing, where the creative bar is moving, and which hypotheses deserve to be tested.
How should a team use Copilot every week?
A practical weekly workflow is to save a small set of prompts for recurring analysis: one account health check, one wasted spend review, one creative fatigue review, one competitor scan, and one next-action summary. The team can run those prompts against the relevant selected accounts, inspect the structured outputs, and use the same thread context to move from diagnosis to a decision. Over time, saved prompts and custom instructions make the workflow more consistent because the team is not relying on one person to remember the exact question, threshold, or analysis format.