What you can do here today, and what nothing here promises.
A plain description of the product and its commercial model, not a contract.
This is a template, pending review by counsel
AdBuyMCP is in design-partner phase and its legal documents have not been through a qualified solicitor. This page is a plain-English summary of how the product actually behaves and what it charges, written from the code rather than from a precedent, and it is published in that state deliberately: our own pre-launch checklist lists the terms of service, privacy policy, data processing agreement and subprocessor list as templates to be reviewed before they are relied on.
So read this as an accurate description of the product and its commercial model, and not as a contract. Nothing here is an offer capable of acceptance, and no agreement exists between us until one is signed. An engagement is governed by the design-partner agreement covering it, and where that agreement and this page differ, the agreement is the one that counts.
The company behind the product
AdBuyMCP is a product rather than a company, and a product brand with no company behind it is the exact shape of the thing a procurement team is trained to reject. So the counterparty is named, and it is on the public register.
AdBuyMCP is built and operated by Tenhaw LTD, a company registered in England and Wales, number 12735685, operating from London. The service is offered to businesses rather than to consumers.
These pages describe what the product does and what it charges. They do not create an agreement. An engagement is governed by the design-partner agreement signed for it, and that agreement covers the things this page deliberately leaves open: the liability position, the data processing terms, the invoicing and credit arrangement, and what happens at the end.
Questions about any of it go to hello@adbuymcp.com. The register entry is at Companies House, and it is the one fact on this page that nobody here can edit.
What works today, and what does not
The whole loop — brief to persona to plan to creative to approval to launch to delivery to measurement — runs today in a deterministic sandbox that needs no credentials and no card. That is a real product you can drive, not a video of one. What is not ready is ordinary customer spend. The live buying clients are written but not proven: most pass tests against mocked responses only, one connector has had read-only calls accepted by a live vendor, and several are stubs that make no vendor call at all. None has cleared our launch standard of a controlled paid activation reconciled to a provider invoice, so live spend sits behind a design-partner engagement rather than a signup form.
Drive the whole loop in the sandbox
Brief to persona to plan to creative to approval to launch to delivery to measurement, deterministic and repeatable, with no API keys, no credentials and no card. It is a working product rather than a video of one, and every impression, reach, CPM and delivery row it produces is sandbox output and labelled as such.
Drive the same loop from an MCP client
The MCP server exercises the same REST routes, the same tenant boundary, the same wallet gates and the same approval records as the browser. There is no second execution path with weaker rules. The server is self-hosted, single-tenant and bearer-token today; there is no public hosted connector.
Fund and run a live campaign self-serve
You cannot. Of the 25 connectors, 0 have cleared our own launch standard — a contracted supplier request accepted, a controlled paid activation, immutable creative identity, report rows reconciled to an invoice and a rehearsed failure — so live spend sits behind a design-partner engagement rather than behind a signup form.
Treat a sandbox number as a forecast of your campaign
You should not. Reach, frequency, delivery and lift figures produced in the sandbox are deterministic synthetic output for evaluating the software. They are labelled as modelled wherever they appear, they create no media obligation, and no completed measurement on this platform has ever supported a causal claim.
The counts come from the supply matrix, which grades every connector on the platform’s own six-level evidence ledger and publishes the blocker on each. Read it rail by rail.
How to read what is published here
Most of this site is an argument backed by data modules that read our own code. That makes it unusually checkable and it does not make it a warranty.
The numbers are sandbox output
Any plan, reach figure, CPM or delivery row shown on this site is deterministic sandbox output and is captioned as such. It demonstrates the method. It is not a forecast of your campaign and it creates no obligation on either side.
The assistant can be wrong
The on-page assistant answers from published material on this site. It is grounded rather than infallible, and anything load-bearing should be checked against the page it cites or asked of a person. What you type into it is sent to a model provider; the privacy page says exactly what happens to it.
Nothing here is an offer
Published fees, channel floors and engagement descriptions are information, not an offer capable of acceptance. They are accurate as at the date shown at the head of this page, and they can change; a change is not a breach of anything.
What this site does with your own data — a cookieless counter that runs for everyone, a consent-gated analytics category, the booking calendar and the assistant — is set out on the privacy page, which carries the same template notice as this one.
10% of media, and a wallet you fund before you spend.
Two tiers, one separate charge for asset production, and no undisclosed margin, rebate or arbitrage anywhere in the path. The fee accrues one ledger entry per delivery row, carrying the date, the campaign, the channel, the spend and the fee, so it is visible against the media it was charged on rather than summarised at the bottom of a statement.
- Standard
- 10% of mediaplatform fee: None
- Up to and including £25,000 of trailing 30-day media spendOne fee ledger entry accrues per delivery row, carrying the date, the campaign, the channel, the spend and the fee. The fee is visible against the media it was charged on rather than summarised at the bottom of a statement.
- Scale
- 5% of mediaplatform fee: £749 per month
- Strictly above £25,000 of trailing 30-day media spendThe tier is evaluated on trailing 30-day spend, so it switches on when your spend does rather than at a contract renewal. The monthly fee is charged for any month in which the scale tier was active on any day, and it appears as its own statement line.
Finished assets are a distinct hard charge, priced at an 80% margin on raw provider cost — retail is cost divided by one minus the margin, so 5× cost, floored at 50p — debited from the wallet at submit and automatically refunded if the provider fails. It never touches the media fee. Blending asset costs into a media percentage is how an ad-tech margin becomes invisible. Keeping them apart costs a simpler headline and buys a statement where every line can be checked against something.
Prepaid, and the reason is not our cash flow
The wallet is the whole of the payment model stated on this page, and that is deliberate rather than an omission. Whether media spend flows through us as principal or client-direct as agent is an open business decision, and invoicing, credit and any late-payment position all follow from it. Publishing terms that depend on a decision nobody has taken would be inventing a term of business. What follows is what the code enforces today.
- £100 minimum top-up
- The smallest credit the buyer-facing top-up contract accepts. Top-ups, spend, asset charges, asset refunds and the platform fee are five distinct entry kinds in one ledger, so a statement reads as a sequence of events rather than a total.
- A launch reserves the fee as well as the media
- The funding check at launch covers the media and the maximum platform fee on it, not just the media, so a campaign cannot go live with enough money for the buy and not enough for the fee. If the wallet is short, the refusal names the exact amount: "wallet is £X short of the launch commitment", rather than a generic decline you have to work backwards from.
- Media spend fails closed
- A launch that the wallet cannot fund does not launch. The check is enforced by default and callers may only opt into warn-only behaviour for sandbox previews, which is the right way round: a preview that warns is useful, a live buy that warns is a bill.
- Asset charges are a hard gate with no override
- A real provider cost is incurred the moment an asset job is submitted, so there is no warn-only path for assets at all. If the wallet cannot cover it, it does not submit.
- Funds stay held through the pause boundary
- On a live line, a reservation is not released until spend has been reconciled to the provider's own invoice or account export, and the final match is rechecked inside the same transaction as the release. Missing, mismatched or stale evidence keeps the money held.
- Sandbox credit is not money, and cannot become money
- Without a payment key, a sandbox top-up credits the wallet with no payment and no network call at all. Live execution refuses a wallet containing sandbox-minted credit outright, so a sandbox balance can never be spent on real media.
- GBP only, and VAT is not modelled
- The entire money surface is sterling. VAT is not modelled anywhere in the platform and remains an open business decision rather than a configured rate.
Funds held in a wallet are a claim on Tenhaw LTDrather than client money held in a segregated account. Reservations against live campaigns are held until spend is reconciled to the provider’s own invoice, which limits how far ahead of delivery the balance runs, and it does not make the balance bankruptcy-remote. Fund what you are about to spend rather than a quarter in advance, and raise it in the design-partner contract if it matters to your finance team.
6 things we do not promise
A terms page usually establishes this by omission and a disclaimer. Setting it out as a list costs us the reader who wanted to assume otherwise, and that reader would have been disappointed later and at greater expense. Every item is published elsewhere on this site as well, so none of it is a small-print surprise.
- A service level agreement or uptime commitment
- There is none. No uptime target, no response-time commitment, no on-call rota, no escalation path and no status page. A shared support inbox is listed as a pre-launch item that has not been done. These are unbuilt rather than unmentioned, and they are not implied by anything on this site.
- An account team or a named launch specialist
- AdBuyMCP is software, not a media agency. Nobody here writes your strategy deck. What you get instead is a published fee, no undisclosed rebate, and the fee on the statement beside the media it was charged on.
- Vendor acceptance, clearance or broadcast timing
- On the workflow rails — cinema and addressable TV — no self-serve API exists. The platform generates the booking brief, drives the clearance step and holds the insertion-order state machine. It does not make the booking and it does not report the delivery, and it cannot guarantee that a sales house accepts a request, that clearance passes, that inventory is available or when a spot runs.
- Clearance, from the creative pre-check
- The creative pre-check enforces channel format limits as schemas and flags risk in copy. It is an assistive tool. It is not Clearcast, it is not CAA clearance, and it does not make your advertising compliant with the CAP and BCAP codes. That responsibility stays with you.
- Verified media quality, viewability, fraud protection or brand safety
- No verification partner is integrated. Placement reporting, supply-path transparency and independent verification are unbuilt capabilities rather than features held back, and nothing on this platform should be read as a media-quality assurance.
- A measured outcome, an incrementality figure or a case study
- There are none, so none are published. The geo-lift engine refuses a test the spend cannot power and returns the sample days the design would need instead of a number, and no completed test has reached a causal grade. A published customer outcome does not exist yet.
The full account of what is and is not held — certifications, audits, registrations and the open launch gates — is on the security page, which publishes the same list a due-diligence questionnaire would ask for.
5 things that stay with you
Short, and all but one of them are things the platform cannot discharge for you however carefully it is built: it records what it did and names what it did not check. The one it does enforce in code is the approval boundary, and that is the one worth reading twice — a process designed around it does not work, because a consequential request has to match an approval a named human granted for those exact bytes.
- Accurate briefs, and the rights in what you upload
- You provide the brief, the creative and the campaign instructions, and you hold the rights in everything you put into the platform. Where AI produces an asset, the platform records that it did; whether that output is fit for paid advertising in your market is a question for your own counsel rather than an assumption to bury in a default.
- Compliance with advertising law and the codes
- Your advertising has to comply with applicable law and with the CAP and BCAP codes. The pre-check helps and does not clear.
- A lawful basis for your own data
- Where a campaign uses your first-party data, you warrant that you have a lawful basis and any consents required for its use. The compiled plan carries a lawful-basis manifest naming each segment's provider and basis, and it attaches conditions you still owe — a legitimate interests assessment on file, a vendor consent-framework attestation verified. It is evidence, not a compliance guarantee.
- Keeping approval where it belongs
- The agent proposes; a named human authorises money. Agents cannot hold the exact-digest approval permission, cannot resolve human approvals and cannot enable spend on a staged live line. Do not build a process that routes around that, because the platform will fail closed rather than accommodate it.
- Your credentials and your workspace membership
- Keep account credentials secure and organisation membership current. Vendor credentials you connect are encrypted at rest and never returned by an API read, and removing a connection deletes them.
Three clauses, and the one that is deliberately not written here
Ownership and governing law are stated because they are stable and unsurprising. The liability position is not, because a cap is the most consequential clause in any agreement and the least suited to being announced on a marketing page before a solicitor has read it.
What you own, and what we do
We retain all rights in the software and in the content of this site. You retain all rights in your briefs, personas, creatives and campaign data, and grant a licence to process them solely to provide the service. Aggregated, non-identifying usage statistics may be used to improve the product. Your data is not used to train AI models.
Liability, and where it is actually set
Nothing excludes liability for fraud, or for death or personal injury caused by negligence, and nothing on this page attempts to. Beyond those two, the limits and exclusions belong in the design-partner agreement rather than here, and they have not been settled with counsel. A cap published on a marketing page and later contradicted by a contract is worse than no cap published at all.
Governing law
England and Wales, with the exclusive jurisdiction of its courts. That is the position for a London company selling to UK advertisers, and it is one of the few things on this page that counsel is unlikely to change.
The agreement you sign is the one that governs.
Changes to this page are posted here, and the fees on it move only when the platform’s own billing configuration does, because both are read from the same place. None of that is a substitute for the design-partner agreement, which is where the liability position, the data processing terms and the commercial arrangement are actually set. If something here matters to a decision you are about to make, ask for it in writing.
What buyers ask before they sign anything
Seven questions, including the two most terms pages answer with a clause: what the payment terms actually are, and what happens to a prepaid balance if the company fails.
Can I buy media through AdBuyMCP today?
Not self-serve. The sandbox is complete, free and needs no credentials, and it runs the whole loop end to end. Live spend runs inside a paid design-partner engagement under a separate written agreement, because of the 25 connectors registered on the platform, 0 have cleared its own launch standard of a controlled paid activation reconciled to a provider invoice. Nothing on this site is an offer capable of acceptance.
What does it cost?
10% of media spend, disclosed per delivery row with no spread and no undisclosed margin, dropping to 5% plus £749 a month once trailing 30-day spend passes £25,000. AI asset production is a separate published charge and is never blended into the media fee. The wallet is prepaid with a £100 minimum top-up, and below about £10,000 a month in total media this is the wrong product, which the pricing page says at length.
What are the payment terms?
Prepaid, and nothing beyond that is set. The platform enforces a wallet: a launch reserves the media and the maximum fee on it, a launch the wallet cannot fund does not launch, and an asset charge is a hard gate with no warn-only path because a provider cost is incurred at submit. Invoicing, credit and any late-payment position are not stated here, because the underlying question — whether media spend flows through us as principal or client-direct as agent — is an open business decision, and publishing credit terms that depend on it would be inventing a term of business.
Do you offer a service level agreement?
No. There is no uptime commitment, no response-time commitment, no on-call rota, no escalation path, no status page and no named account team. A shared support inbox is a pre-launch item that has not been done. If a support SLA is a procurement requirement, we do not clear it today.
Is the creative pre-check the same as clearance?
No. It enforces each channel's format limits as schemas — copy that exceeds a limit is rejected at the boundary rather than truncated on its way to a provider — and it flags risk in copy. It is not Clearcast, it is not CAA clearance and it does not make your advertising compliant with the CAP and BCAP codes. On the workflow rails, vendor acceptance, clearance outcomes, inventory availability and broadcast timings are outside our control and are not guaranteed.
What happens to money in the wallet if the company fails?
It is a claim on Tenhaw LTD rather than client money held in a segregated account, and that deserves a straight answer rather than a clause. Reservations against live campaigns are held until spend is reconciled to the provider's own invoice, which limits how far ahead of delivery the balance runs, but it does not make the balance bankruptcy-remote. Fund what you are about to spend rather than a quarter in advance, and raise it in the design-partner contract if it matters to your finance team.
Are these terms final?
No, and the notice at the top says so. They are a template written from the product's actual behaviour and have not been through a qualified solicitor. An engagement is governed by the design-partner agreement covering it rather than by this page, and where the two differ the agreement is the one that counts.
Read the terms first, then bring a brief.
A design-partner engagement is the process of walking one rail through the platform’s own go-live standard with you, on your budget and your brief, under an agreement written for it. The working session is where that conversation starts, and it costs nothing and commits nothing.
45 minutes. Bring a real brief and we compile it live.