Once a week Sam can review everything it can see and tell you what worked, what did not, and where your next pound is best spent.
It is the same kind of automation as a morning briefing, set up the same way, on a weekly rhythm instead of a daily one. The difference is what you get. A daily briefing tells you about yesterday. A weekly review has enough data behind it to tell you about a pattern, which is what makes a recommendation worth acting on.
- →At least one tool connected. The more Sam can see, the more the review is worth — ads spend next to store revenue tells you something neither does alone. Connect a tool to Sam.
- →A week of data behind you. Set it up now and the first review will be thin; the second will not.
Step 1 - Ask for a weekly review
Open Sam AI, or the Ads Specialist if the review is only about paid performance, and describe what you want in the Ask me anything.. box:
“Every Monday morning, review how last week went and tell me where to put my budget this week.”
Whoever you ask owns the automation and is where the review lands each week, so ask the one you actually check.
Sam replies with what it plans to cover based on what you have connected, and asks about anything it needs pinned down — usually the time, and sometimes what you count as success.
Step 2 - Tell Sam what "working" means to you
This is the step that separates a useful review from a wall of numbers. Sam compares against whatever you give it, so give it something:
- A target — a cost per sale you are trying to stay under, a weekly revenue number.
- A comparison — last week against the week before, or against the same week last month.
- A decision you are trying to make — "I have a fixed budget, tell me where to move it."
Without one of those, Sam reports what happened. With one, it tells you whether that was good and what to do next.
Then confirm the timing. Sam sets the schedule and tells you when the first review runs.
Step 3 - Read the review
It arrives in the chat with whoever owns it and is saved to the Workbench as a document, so the weeks stack up into something you can look back through.
What is in it depends on what you connected. Ads give you spend against results by campaign; a store gives you revenue, orders and which products moved; an email platform gives you what your list responded to. Sam puts them in the same document, which is the point — a week where ad costs rose and revenue rose faster is a different week from one where only the costs moved.
The recommendation sits at the end. Sam names what it would do next and why, in the terms you gave it in Step 2. Where it can see a clear opportunity it will say so directly rather than hedging.
Step 4 - Adjust it as you learn what you actually read
The first few reviews teach you what you skip. Say so in the chat — "drop the email section", "compare against the previous four weeks instead", "I care about profit not revenue" — and Sam updates the automation.
To check it is running, open the Settings tab, expand Automations under Specialist Settings, and select the history icon for Run History. Runs show as Success, Partial where something did not finish, or Failed in red with the error underneath.
Schedule it for a morning you can act on. A review that lands Monday at 8am gets used; one that lands Friday at 6pm gets read and forgotten.
- ✓Every week, Sam pulls the full period from each connector rather than a single day.
- ✓Against your target, Sam works out whether the week was actually good, not just what happened in it.
- ✓In the review, Sam names where your budget is working hardest and where it is not.
- ✓Every run, the review is saved to the Workbench so you build a record of the weeks.
Frequently Asked Questions
How is this different from a daily briefing?
Mechanically it is the same automation on a weekly schedule. Practically, a week of data is enough to separate a trend from a bad Tuesday, so the recommendations are worth more. Plenty of people run both.
Will Sam tell me exactly how much to move?
It depends on what it can see. With spend and results connected, Sam can be specific about which campaigns deserve more and which are wasting money. With less connected, expect direction rather than amounts.
Can it compare against previous weeks?
Yes, if you ask for that comparison. Sam compares against whatever parameters you give it, so say what you want measured against what.
Where does it arrive?
In the chat with whoever owns the automation, and as a document in the Workbench.
My review did not arrive.
Open Run History from the Settings tab. Failed shows the error underneath and Partial explains what did not finish. A connector that needs reconnecting is the usual cause.
Does it use credits?
Yes. A weekly review pulls a full period from every connector it touches, so it costs more per run than a daily briefing while running far less often. See how much usage you have left shows it per day.