Automated client reporting pulls each client’s data from GA4, the ad platforms and the CRM, computes every figure in code, and drafts the commentary for an account manager to edit. The model writes about the numbers; it never produces them.
All Marketing & Creative Agencies →
Dashboards in Looker Studio solved the charts years ago. What still takes an account manager a day per client is the part the client actually reads: what changed, why, and what the agency will do about it. AI marketing reporting tools promise to write that part, and most of them do it by handing a model the data and asking for a summary.
That works until the model rounds a figure, compares the wrong periods or describes a trend that is not there, and the client spots it. One wrong number in a report damages trust in all the others. So we split the job. Code pulls the data, reconciles it across platforms and computes every metric and comparison. The model receives those figures as a fixed set and writes commentary that may cite only them; a check rejects any draft containing a number that was not supplied. The account manager adds the judgement and sends.
APIs on schedule, errors logged
KPIs and comparisons, in code
only the supplied figures
every figure traced to the set
edits, adds judgement, sends
any untraceable number
It can draft the commentary well, provided it is not also responsible for the numbers. We compute every figure in code and give the model a fixed set to write about, then check the draft for any number that was not supplied. An account manager edits and sends, because the client relationship and the recommendations are judgement calls the agency owns.
Those tools move data and draw charts, and we often build on them. What they do not do is define each client’s KPIs consistently across platforms, catch a broken data feed before the report goes out, or write the narrative. This adds those layers on top of the data you already collect.
The report is marked incomplete for that source and the account manager is told before anything is sent. Silent gaps are the worst failure in reporting, because a missing week of spend looks exactly like a cheap week. Every pull is logged with what was fetched and when.
What we built, what stage it is at, and a link to the live thing wherever there is one.
yourOS is an operating system with an AI at its heart. You talk to it like a person, out loud or typed. It makes your calls, plans your days and works on your projects with a crew of agents, on your own private server. Our own product, built from the first commit; the waitlist is open.
Website, backend and app assets for Sonora, a free AI sound-therapy app on iOS and Android published by Cognitive AI.
Brand, website, prototype, apps on five platforms, the AI backend and the algorithm-intelligence pipeline for Podswap, a social growth engine that runs without a human in the loop.
In closed beta: Roxy, a private AI assistant with permanent memory running on infrastructure the user owns. Brand, site, product, backend and conversational design.
Live at coherence.com: brand identity generated with AI agents we directed, the motion system and website for the group behind Sonora, AboutAI, WakeUp and MudShadows. Agent work remains in build.
Live at domainbooth.com for DomainBooth, a premium domain brokerage: a new animated mark, every page redesigned and rebuilt with designed mobile layouts, and a custom CMS for its sales record, featured names, press and enquiries.
We start with one client’s report and the KPI definitions behind it.
30-minute technical call · no deck