Clean, AI-ready marketing data shouldn't be painful.
broch is the platform and the engineer, for one fee: every channel in one loch you own, a dashboard custom-built for the way you sell, and AI that answers from your own numbers. Live within 7 days.
Otter & Ash is a demo of one: a made-up furniture brand with two years of data. Your work email gets you in; read-only, so click anything.
Any of these sound familiar?
Use broch and get back to doing great marketing. We handle the data and the insights. Pick one and see the screen that retires it.
Every number lives somewhere different
Four logins, three exports and a morning lost to the Monday spreadsheet, every week. Here everything is in one place: every channel on one screen, every morning, from the same numbers.
Everything in one place. Every channel, one screen, the same numbers everywhere.
Reports that load slowly and stop at one platform
Looker Studio (Data Studio, as was) is free for Google channels. Off Google you rent a connector per source, every chart queries the platforms live, and nothing is stored. Then somebody has to edit it: blended data sources, calculated fields. Anyone ever enjoyed that? Here every chart reads your loch, stored and joined, and the edits are ours to make.
This month against last, from the loch. Nothing queried live.
Connectors that are expensive and add up quickly
A fee per source, every month, for a pipe that stops the day you stop paying. Here every source is read on a grant you own and can revoke, and the connectors are in the price. You can see every channel feeding your numbers, when it last ran, and how far back it re-reads so late conversions still land.
Every channel, its last run, its lookback and its freshness, in your broch.
The £70,000 ($134,000) data hire
Those are the median salaries a data engineer is advertised at in the UK and the US, before you've recruited or managed one. With broch, a data engineer comes as part of the service, working hand in hand with you to keep your data clean and easy to read: ask for a change from any page, and we tell you when the data is off, so every decision is made on numbers you can trust.
UK median, £70,000: ITJobsWatch, six months to August 2026. US median total pay, $134,000: Glassdoor.
Data tech is confusing, and gets expensive
A warehouse, a BI tool, a semantic layer, SQL, and a bill for each. Here you type a question in plain English and every answer shows the data it used.
Plain English in, the working shown.
A custom dashboard, built for your business.
This isn't SaaS you set up yourself. An engineer builds it around the way you sell, then keeps building.
Custom-built, live in 7 days.
Every channel in one loch you own, dashboards built for the way you sell, and nothing in your existing workflow touched. Up to ten sources.
Need a change? Give us 24 hours.
A new graph, a metric pulled through, a page laid out differently. Ask from any page and it is live tomorrow, whatever the size.
Save your team 10 hours a week.
The reporting nobody has time for runs on its own. A brainwave reads every channel and writes up what moved, what is wrong and what to do next.
Fair questions.
What you're buying, what it runs on, and one for the nerds.
Is this software, or a service?
Both, for one fee. The platform is the dashboard, the loch and the AI. The service is a data engineer on demand, inside your team, who customises all of it to the way you sell: new pages, new metrics, new sources, as often as you need them. One set fee a month covers the software and the engineer. No seats, no per-user pricing, nothing to configure yourself.
Why not Triple Whale, or a platform like it?
Triple Whale is a good product, and it works well if you are happy to pay for its connections, run the data flows yourself and build on top of it. Its plans are priced by your store's revenue, from a free tier up through Foundation and Automate, and the two things that would make it yours sit outside them: syncing the data to a warehouse you own is an add-on, and a person building on it for you (Moby Concierge) is another, scoped per customer, with dedicated coverage kept for Enterprise. So the choice is: run it yourself, put somebody in-house on rolling it out across the business, or pay for the concierge on top. Their pricing page ↗ sets it out. broch is the other shape: the connections, the storage, the engineer and the custom build are one fee, the data lives in a project that transfers to you, and you tell us what you want to see.
What's under the hood?
A data loch that does the warehouse's job, on the architecture Snowflake, Databricks and Google converged on. Your files sit in Cloudflare R2, a bucket you own, in open formats any tool can read. DuckDB queries them. Postgres runs the app and everything transactional. The connectors are managed tooling we run for you, in the price.
Can we get nerdy? Rows or columns?
Both, each doing the job it is good at. Your shop till, your CRM and our own app are OLTP, online transaction processing: row-oriented databases like Postgres, which store each order's fields together, so writing one order or fetching one customer is instant.
Analysis is OLAP, online analytical processing: a column-oriented engine, DuckDB, which stores each column together, so a question like revenue by channel over two years reads only the columns it needs instead of every row in the store. On analytical queries that is routinely a hundred times faster than a row store, which is why every serious analytics database is built this way and why the dashboard never makes you wait. Row stores for running the business, column stores for asking it questions: that split is best practice, and it is the one we ship.
Why not build it ourselves?
You can. The bill of materials is connectors, a warehouse, the definitions, the dashboards, the upkeep, and the month it takes from the person with the least of it. Most teams get part of the way and stop. Here all six come with an engineer, for one fee, in 7 days. Hold us against every other way to spend the money.
Why not Snowflake?
You can set Snowflake up. Then who runs it? A warehouse needs a data team, or at the very least a full-time data engineer at £70,000 a year, to model the data, keep the pipelines alive and build the dashboards on top, before the warehouse bill itself. That is the difference with broch: we set all of it up, it is hands-off, and you tell us what you want to see. Software and a service in one fee, which is how it ends up the far cheaper option. You'll never need to buy a warehouse.
Can we see it before we talk to you?
Yes. Otter & Ash is open to anyone with a work email: enter it, we send a 6-digit code, and you're in. No password, no call, read-only, so click anything. Open the demo ↗
No demo to book. You've just had it. Let's get you started.
A free call where you tell us about your business and we listen. Then we build it around the way you sell, and it's live within 7 days.