A quick 30-minute intro call to understand your objectives and whether we’re the right team for you.
We’ll be in touch soon to arrange your discovery call.
No retainer commitment. No junior analysts. No generic recommendations. Just the truth about your marketing operation — and what to do about it.
Fill in the application below. Our senior team reviews every submission personally.
If there’s a clear fit, you’ll hear back within 48 hours to book a 20-minute kick-off call.
We run a limited number of audits each month — not every application proceeds.
Your buyer takes weeks to decide and finishes in a quote, a call or a booking. Most attribution gives up long before that. Ours doesn’t.
We audit the full stack, GA4, server-side tagging, call tracking, post-purchase surveys, then rebuild attribution around the journey your customers actually take. Click to closed sale, one number.
That’s the real cost of renting demand: every signal about your buyer stays on their side of the wall. You can’t see what a customer is truly worth, so you can’t prove direct would be cheaper, so the commission line survives another year.
Going direct is what makes real measurement possible, and real measurement is what makes the case for going direct. We build that foundation: your tracking, your data, your numbers, owned end to end and audited so the platforms can’t mark their own homework.
A considered purchase takes many touches over weeks: the video that planted the name, the search three weeks later, the quote finished on a second visit. Last-click attribution hands all of that credit to the final step, then tells you to cut the thing that started it.
We model the whole path and measure to the outcome that pays: the qualified consultation, the completed quote, the booking. CAC, payback and LTV by route in, so you know where to scale, what to stop, and what each pound genuinely returned.
Our sale happens over the phone or in a consultation. Can you actually track that?
Yes, and if we can’t connect it, we won’t pretend otherwise. Call tracking, quote-flow tracking and CRM integration tie the offline outcome back to the journey that produced it. You stop optimising for the lead and start optimising for the consultation that shows up, the quote that completes, the booking that pays.
Our blended CAC looks fine. Why would we dig deeper?
Because blended CAC flatters the deck and breaks the business. It averages your cheapest repeat customer with your most expensive cold one and calls it health. We break it out by route in: what a customer costs direct, what one costs through the aggregator once you count the commission and the data you hand over, and which is actually growing.
The platforms all claim the same conversion. Who's right?
Usually none of them, which is the point. Meta, Google and the aggregator each mark their own homework, and together they’ll happily report 150% of your revenue. We build one source of truth outside the platforms, so the numbers you take to the board are yours, not theirs.
Which attribution model do you use?
The honest answer: the model matters less than what you measure to. First-click, last-click and data-driven all have a place, and we’ll use whichever fits the question being asked. What we won’t do is measure to impressions, reach or platform-reported ROAS. Every model we run resolves to the same commercial outcomes: CAC, payback, LTV, and the share of sales that come to you direct.
We report into a group. Can your numbers survive that scrutiny?
That’s the environment we built this for. Inside a top-10 global insurance group, our reporting goes upward to group level, which means every number has to be defensible, repeatable, and explainable to someone who wasn’t in the room. If your board or investors will interrogate the numbers, better that we build them expecting it.
How long until the data is trustworthy?
The audit takes days and usually finds the same things: broken tags, double-counting, a journey that goes dark at the phone call. Fixing the foundation takes weeks. From there, trust compounds, and the longer it runs, the harder the question it can answer, up to the one that matters: is direct paying?
