Mid-Week Claude

Any account manager. Any client. Any question. Answered.

Mondays are the obvious use case. The rest of the week is where Scout earns its keep - ad-hoc questions about any client, with live data, the right ad account, and the client’s history loaded by default. No re-briefing. No copy-paste prompts. Just Claude, with your clients.

Claude · Driftline
How's Driftline's CPP this week?
scout·get_metrics("noobiie", "meta", 7d)
$18.40 - down 12% wk/wk on $4.2k spend. The new UGC angle is carrying it; static creatives are flat at $26.
What's the top creative on Meta?
scout·get_creative_images("noobiie", top=3)
The "first-purchase morning" UGC - 2.4× the CTR of the next best, 41% of spend. Promoted fact says you flagged this angle two weeks ago.
Did the email re-engagement flow convert?
scout·get_metrics("noobiie", "klaviyo", 14d)
Yes - 3.1% click, $42 RPR over 14 days. Founder voice override says don't recommend a discount on top of it.
What an account manager gets asked between Mondays

Six clients. Three Slack pings. Two ‘quick question’ emails. By 11am Tuesday.

‘Why did CPP go up last week?’ ‘Which flow performed best?’ ‘Can you check that the new collection campaign is actually pulling weight?’

Today, the account manager opens five tabs per question. With Scout, they open Claude.

End to end

Live data. The right account. The client’s context.

An account manager asks Claude about a client. Scout switches context to that client, pulls fresh numbers from the relevant platforms, references promoted facts and voice overrides, and Claude answers in your agency’s voice - with the specific numbers the question demands.

Need a chart? Scout renders one inline. Need the answer in an email? Claude drafts it. Need to dig deeper? Ask follow-ups in the same thread.

Why did Driftline’s CPP jump last week?
CPP went from $14.20 → $18.40 (+30%). Two drivers:
Cold ROAS softened to 1.9× - paused two underperforming creatives Tuesday.
Welcome flow held at $8.40/recipient - kept revenue stable.
SCOUT TOOLS CALLED
set_active_client·Driftline
get_client_context·memory + voice
get_metrics·Meta · 14d
get_metrics·Klaviyo · 14d
Across the book

Scout knows which client owns which Klaviyo.

Three questions about three different clients, in three minutes, in the same chat. Scout switches client context per turn. No re-auth. No ‘which account did you mean?’ Just answers.

Learn more about multi-account →
Team continuity

What one account manager learns, the next inherits.

An account manager has a mid-week conversation that surfaces a new insight about a client. Scout captures it as an observation. The next account manager asking about the same client sees it. Knowledge accumulates instead of evaporating.

Learn more about agency memory →
FAQ

Common questions about mid-week Claude.

Do I have to re-brief Claude every time?

No. Scout loads the client’s memory, voice and live data by default the moment you scope to a client - so you ask the question and get an answer that already knows the account.

Can I ask about several clients in one conversation?

Yes. Scout switches client context per turn, with no re-auth and no “which account did you mean?” - three clients in three minutes in the same thread.

Is the data live or cached?

Live. Scout pulls fresh numbers from the relevant platforms when you ask, so a mid-week read reflects what’s actually happening today.

Can it turn the answer into something I can send?

Yes - ask Claude to draft the client email, render a chart inline, or fold the insight into the next Monday report. It’s the same conversation.

Its ability to work with AI tools makes it infinitely accessible to everybody across departments - and gives clients a clear view of what's happening in their business, and more importantly what to do with it.

KF
Kelvin, Founder & Principal, Compound GP

Ask Claude anything about any client this week.