The Ad Spend
The ad operations intelligence layer.
Version control for advertising.
The platforms were never going to build this. AI made it possible for someone else to.
The platforms erase it.
Google deletes change history after 90 days. Meta caps and buries it. LinkedIn stores none. The decisions behind hundreds of billions in spend — gone.
The data was designed to be incompatible.
Three platforms, three schemas, zero unification. Not by accident — by design. Walled gardens don't benefit from cross-platform intelligence.
Nothing was built to reason.
Dashboards render charts. You can't ask a chart why your CPA spiked Tuesday from Monday's targeting change. That required a human analyst — until now.
This couldn't have been built two years ago. The foundational AI infrastructure didn't exist at scale. Now that it's been commoditized, the value shifts to the application layer — and the proprietary dataset underneath it. Not by improving dashboards, but by replacing them.
We built the layer the platforms never would.
$680 billion in annual ad spend. No system of record for the decisions that drive it.
The B2B team that just lost their best media buyer
Two years of optimization decisions — erased by the platforms. No changelog. No institutional memory. The new hire starts from zero, and so does the budget.
$47K average cost to replace a knowledge worker — and the institutional memory walks out with them
(SHRM, 2024)The agency managing 20 accounts with 3 people
Their clients ask 'why did performance change?' and the honest answer is: we're guessing.
Only 20% of CMOs can accurately measure marketing ROI — and they're relying on the agency to close the gap
(Gartner 2025)The SMB running its own ads for the first time
The platform says 4x ROAS. The finance team says revenue is flat. Nobody can reconcile the two — because no system connects the ad decisions to actual business outcomes. They're spending, but they can't prove it's working.
61% of small businesses say they can't measure the ROI of their ad spend
(Clutch, 2024)We didn't research this problem. We lived it — across dozens of clients and millions in managed spend.
Built an agency to 40+ clients including Canva and Siemens — zero outside capital, 18 months. The same data gap showed up in every single account. We built internal tools to fix it. Clients started asking for access.
Enterprise
Sr. Manager, Marketing at RingCentral. Smart strategists spending half their time trying to justify metrics they couldn't trace — at the mercy of platforms that won't show what changed.
Agency
Founded The Matchbox. Grew to ~25 people, 40+ Fortune 500 clients, millions in managed spend. Canva and Siemens as clients within 18 months. Zero outside capital. Zero shortcuts.
The Pattern
Every client — same structural problem. Marketers pumping strategic decisions into platforms that operate as black boxes, then expected to explain why the numbers moved. Every tool on the market was still subject to the same opacity underneath.
Product
Built internal tools to solve it for our own team. Clients started asking for access. The Ad Spend was born. 137 companies onboarded, $68M+ in ad data analyzed — in a matter of weeks.
3× founder. Built The Matchbox to Fortune 500 clients managing millions in ad spend — zero outside capital. Saw the same data gap across dozens of client accounts before building the infrastructure to fix it. BA Economics, University of San Francisco.
Spent 6 years building change-detection systems at Meta (Buck2), Vercel (Turborepo), and Google (Bazel) — the same algorithms now powering The Ad Spend's intelligence layer. Maintains Stripe's leading Rust API library (async-stripe) — 1.9M+ downloads, $3K/mo in Stripe sponsorships. First Class Honours BSc Computer Science, Heriot-Watt.
Founding engineer at Josa — built to acquisition in 8 months, top 10% of YC applicants. Built an LLM automation system that generated $1M in revenue in 4 months. Processed $300M in political data with AI for UK transparency initiatives. BSc Computer Science (AI/ML), Heriot-Watt.
The system of record — live, learning, and compounding.
Full change history
Every change logged, timestamped, and linked to performance impact. The infrastructure the platforms never built. Git for your ad accounts.
Causal anomaly detection
160+ algorithms running continuously. When your CPA spikes, the system traces it to the specific targeting change that caused it — not a correlation, a cause.
Automated stakeholder intelligence
The CFO gets spend efficiency. The CMO gets pipeline impact. The media buyer gets creative performance. Generated automatically — replacing the analyst who used to build these manually.
Universal access
One Slack command. Key metrics across every platform, in seconds. No logins, no seat licenses. Anyone in the org — finance, sales, exec team — gets the same intelligence.
15-second install. Slack-native. No implementation. No seat licenses. Every connected account makes the system smarter.
The platform that gets smarter with every dollar spent — and every account connected.
We're building the operating system for paid media. Five layers — each one replaces the next role the team still does manually.
Every change across Google, Meta, and LinkedIn — logged, unified, and queryable. The system sees what changed, when, and connects it to what happened next. 160+ detection algorithms catch problems before they compound.
Layer 1 is LIVE. Layer 2 is in BETA.
Version control is the wedge. Each layer compounds on the data beneath it. No competitor can start at Layer 3 without building Layers 1 and 2 first — and they'd be starting with zero historical data.
$68M+ in ad spend. 137 companies. Compounding data.
Interrupt Media
B2B agency. The AI labor unit offset in action.
“This has fundamentally changed the way we run our business.”
Ben Lack, CEO, Interrupt Media
Free to paid. Each tier replaces the next role on the team.
Slack visibility, 5 alerts, 5 AI queries, weekly report, history browsing. The hook.
Full anomaly detection, unlimited alerts + AI chat, stakeholder intelligence, 365-day trend analysis. Replaces the junior analyst pulling weekly reports.
AI optimizer with human approval queue. Budget shifts, bid adjustments, one-click execution. Replaces the campaign manager doing manual optimizations across platforms.
+ ad spend feeRevenue attribution tied to specific ad decisions. CRM integration. Replaces the BI analyst building attribution models in spreadsheets.
+ ad spend fee$39/client/mo (min 2). Average agency runs 8 clients = $312/mo recurring at zero incremental acquisition cost.
1 agency = 6–12 accounts
Built-in expansion channel.
Three structural fractures. No competitor can bridge them — and the platforms won't.
A market of point solutions
no platformOne tool for attribution. Another for anomaly detection. Another for change history. Another for optimization. None of them share a data model. None of them compound on each other. The result is a stack of disconnected surfaces that each solve one problem in isolation — and create new ones in the gaps between them.
LLMs on dashboards
vs. causal intelligenceEvery competitor is pointing language models at the same surface-level metrics the platforms already show. They're summarizing dashboards — not reasoning about what caused the numbers to move. Causal intelligence requires the full decision history underneath: what changed, why it was changed, and what happened next. You can't get there from a dashboard summary.
The dataset nobody else is building
compounding dailyEvery connected account adds to a cross-account training set of strategic decisions — targeting choices, budget shifts, creative rotations — mapped to downstream impact across verticals, segments, and company sizes. We train across the strategic narrative of every account. No other company is capturing decision-level intelligence at this depth. Competitors would need to start from zero — and they'd need years of history they don't have.
Not a dashboard. Not a point solution. Not an LLM on top of metrics. One structurally neutral intelligence layer — training on a proprietary dataset of advertising decisions that compounds with every account and cannot be replicated by anyone who didn't start capturing from day one.
$2M to own the intelligence layer for a $680B market.
137 companies. $68M+ in ad spend ingested and analyzed. Optimizer in beta. Capital efficient by design.
~100 paying customers
HubSpot App Marketplace listing live
Agency channel activated
Optimizer GA. Approval queue + confidence scoring in market
$199/mo tier activated
Agency multiplier compounding
HubSpot CRM integration live
Fluid Touch Attribution in early beta
$499/mo tier in pilot
Cross-account intelligence dataset at scale — the compounding moat
2 engineers to ship Optimizer and begin Attribution. Every dollar of product spend directly unlocks the next revenue tier.
PLG distribution, agency channel, HubSpot Marketplace.
Three founders. Infrastructure. Lean by design.
The Ad Spend
The ad operations intelligence layer. Version control is how we get in.