Share your client and goals. Get a campaign performance analysis report with scored metrics, benchmarks, and a prioritized action plan.
Start by sharing your client's name, ad platforms, and the time period you want reviewed. Juma connects to your Google Ads or Meta Ads account and pulls live campaign data automatically. The Flow runs a campaign performance analysis against industry benchmarks and your own historical baselines, then returns a performance scorecard, benchmark comparisons, and recommendations sorted by effort and impact.
Search terms audits reveal the gap between the keywords you bid on and the queries that actually trigger your ads. That gap is where budget leaks. This step pulls the full search terms report for the specified time period and audits it for irrelevant queries, missed exact-match opportunities, and structural issues.
The audit produces three lists:
All three feed directly into the optimization action plan in Step 4.
Period-over-period comparison shows what improved, what declined, and by how much - and flags whether the movement was caused by account adjustments or external factors. This step pulls data from two consecutive periods, compares all active campaigns, and checks the account change history to explain any metric that moved more than 15% in either direction.
For each campaign, the comparison covers:
Any movement the change history cannot explain gets flagged for manual review before the action plan.
The final step translates scored findings into a prioritized action plan with specific actions, expected impact, effort level, and a timeline for each item. Items are grouped into three tiers: quick wins for this week, strategic shifts for next cycle, and tests to validate before scaling.
Each action includes:
The output is structured so the team can act immediately without a separate planning meeting.
Campaign analysis is a recurring task: same client accounts, same goals, same targets, reviewed month after month. A Juma project stores the context that stays constant so every analysis starts from what's already known instead of re-asking the basics. If the client project already exists, add these items to it.
Add a short description to each knowledge item in the project's info field so Juma knows what each file contains and when to use it. For example:
This Flow reduces manual campaign analysis from 2–4 hours to a single conversation. Manual analysis requires exporting data from each platform, formatting it for comparison, looking up industry benchmarks, and writing recommendations - each step in a different tool. This Flow handles all of it in one place, with no context-switching between platforms.
If Google Ads or Meta Ads is connected to the project, Juma pulls the live campaign data automatically at the start of every session. You skip the export step entirely. If no integration is connected, paste performance data directly or upload a report file and the analysis runs the same way.
For teams managing multiple clients, the time savings compound. Each client project stores its own context, integrations, and historical baselines, so the setup work you do once pays off on every subsequent analysis without repeating setup questions.
The campaign performance analysis scores each campaign across five dimensions: efficiency (CPA, ROAS), volume (conversions, clicks), quality (conversion rate, bounce rate), trend (week-over-week direction), and budget health (pacing, allocation). Campaigns that fall below threshold on any dimension get a deeper review at the ad group and ad level.
Every scored finding feeds into a tiered recommendation set. Recommendations are sorted into three groups: quick wins to act on this week, strategic shifts to plan for next cycle, and tests to validate before scaling. Each recommendation traces back to the specific metric or campaign that surfaced it, so the connection between finding and action is always clear.
The analysis also runs a search terms audit as a follow-on step, flagging irrelevant queries burning budget and high-performing queries worth promoting to exact match. The scorecard and search terms audit together give a complete picture of where the account stands and what to do next.
Yes, this Flow works without a connected integration. Paste performance data directly into the conversation or upload a report file exported from either platform, and the campaign performance analysis runs the same way. The integration is optional - it removes the manual data step and keeps numbers current, but it is not a requirement.
The main difference is repeatability. With an integration connected at the project level, anyone on the team can re-run the analysis at any point without re-uploading or re-pasting data. The live connection means numbers are always current, with no risk of working from a stale export.
For teams running recurring monthly reviews, connecting the integration once and storing it at the project level is worth the five-minute setup. For a one-off analysis or a new client evaluation, pasting the data directly works fine and returns the same quality of output.
Include performance targets by channel (target CPA, target ROAS), which campaigns to focus on, and any standing constraints like budget caps or seasonal factors. Adding a past campaign performance report to the project knowledge base also lets Juma benchmark against the client's own track record, not just industry averages. The more context provided upfront, the more actionable the output.
At a minimum, include the client's name, the platforms to analyze, and the time period. That gives the Flow enough to run a valid campaign performance analysis and return a scored output. Performance targets and constraints take the recommendations from directionally correct to specifically executable.
Context stored at the project level carries over automatically. A Campaign Analysis Context document covering targets, focus campaigns, and standing constraints means Juma reads those parameters at the start of every session without being re-entered each time.
Yes. Set up a separate Juma project for each client and store the campaign context, integrations, and historical baselines there. Every conversation in that project starts with the right context already loaded, so Juma does not re-ask setup questions and the analysis starts immediately from the right frame.
This setup makes the analysis repeatable at scale. When a new team member runs the analysis for a client, they do not need to be briefed on the account history - it is all in the project. When you return the following month, the prior cycle's output is already there to compare against.
The project structure also keeps client data cleanly separated. There is no risk of context from one client bleeding into an analysis for another. Each project is its own contained workspace, and the integrations connected to it only surface data relevant to that client.
