Investment firm
£40M+ in 18 months.
A property investment firm launched with more than £5bn of founder credibility, eleven external agencies and no ownership of its own performance data. Over an eighteen-month engagement we rebuilt the growth engine as a single operating model, brought it in-house and took the firm past £40M in revenue.
The situation
Credibility on paper, eleven agencies in practice.
The firm launched with more standing than most established players. Its founders had secured over £5bn of investment across their careers and built the proposition on what had already worked, aimed at high-net-worth investors across the UK, the Gulf and Asia. Go-to-market was outsourced: eleven external agencies running bottom-of-funnel lead generation, all competing for the same narrow audience with the same products.
Three months in, performance sat below plan. Campaign data belonged to third parties with no agreement in place to share it, and the structure was linear. Leads passed forward to sales and nothing passed back.
Nothing about the output read as one firm either. Eleven partners meant eleven interpretations of the brand, from tone to typography, bidding against one another in the same auctions for the same investors.
The diagnosis
Eleven agencies, twenty searches.
Search analysis exposed the imbalance. Category-level demand ran past 3,500 searches a month while the product-level terms the agencies bid against drew around twenty, sampled clean through a VPN so personal history could not skew the read. The firm was paying eleven partners to fight over the smallest pool of demand in its market, leaving the wider funnel unclaimed. The problem was structural, which meant the fix had to be built rather than briefed.
The market was asking category questions. The spend was answering product ones.
The operating model
Plan, build, action, and a loop that learns.
We rebuilt marketing as one system with three functions. Directives set quarterly by the executive team gave planning a single voice. A central implementation function built every campaign against those directives. The sales division acted on what implementation produced.
Around all three ran the learning loop: weekly reports, monthly objective reviews and quarterly resets, fed by quantitative platform data on one side and qualitative signal from the sales floor on the other. Budget followed evidence. When an ad set breached its cost thresholds it was switched off and the next tested variable took its place.
Four stations on one loop. Each quarter starts better informed than the last.
| Directive | Hypothesis | Build | Threshold | Review |
|---|---|---|---|---|
| Open the category funnel | Category demand converts cheaper at guide stage | Three landing pages, two lead magnets, one calculator | CPL under £12 | Weekly |
| Lift MQL to SQL rate | Trust and transparency drive commitment | Messaging rewritten on the six drivers, nurture sequence live | SQL rate rising month on month | Monthly |
| Replicate what converts | Lookalikes hold CPA at scale | 1% lookalike built from the top converting sequence | CPA inside stage threshold | Weekly |
One page, owned by the executive team. Every campaign in market traced back to a line on it.
Phased targeting
Three campaign layers, one funnel.
Awareness campaigns collected audience signal and let the platform algorithms learn. Traffic campaigns segmented and retargeted on what awareness surfaced. Conversion campaigns closed once the data justified the message, and lookalike audiences were replicated from the best-converting sequences. CPC and CPA thresholds at every stage of the funnel kept the testing disciplined, opening new segments without setting fire to the budget.
Each layer earns the next. Thresholds keep the spend honest.
The funnel lab
Landing pages and lead magnets, tested like variables.
The funnel was treated as a laboratory. A series of landing pages went live against every stage, each paired with a different lead magnet: investor guides, downloadable briefings and a tailored return report calculated from the prospect’s own numbers. The hypothesis held. Top-of-funnel capture produced far cheaper leads, and the tailored report became the workhorse of the mid funnel.
- 1A category promise in the headline, matched to what the market searches for.
- 2The lead magnet does the converting. Guides, briefings and the calculator were rotated and tested.
- 3The trust strip answers the top decision drivers before the form asks for anything.
- 4Two fields. Every extra field taxed conversion, so nothing extra survived.
The landing page anatomy. Layouts and copy are representative; the structure and the numbers are the client’s.
Category education for the 3,500 monthly searchers the old model ignored.
Downloadable analysis that kept captured leads warm between decisions.
A calculated report built on the prospect’s own numbers. The workhorse of conversion.
The audience system
From marketing-qualified to sales-qualified.
Converting leads into conversations needed the audience understood at depth. Questionnaires, surveys, forum analysis and review mining built a database of decision drivers, and they fell under six themes. Core messaging was rewritten to answer each one, then marketing and sales carried the same answers so a lead heard one firm from first click to signed commitment.
Prove the record before asking for the meeting.
Show workings, fees and downside in plain terms.
Short forms, fast follow-up, no wasted steps.
Evidence over adjectives at every stage.
Teach the category and own its questions.
Named experts, an audited process, real assets.
The six decision drivers, each answered in messaging before sales ever picked up the phone.
The work in market
MQL capture, run as a system.
Creative was built centrally against the directives and released in phases, so each layer of the funnel carried its own message. Awareness taught the category. Traffic answered the trust and transparency drivers. Conversion put the prospect’s own numbers to work.
Representative units from the three campaign layers. Formats and structure match what ran; identity and imagery are withheld.
Data and reporting
From a dump of lead counts to one readable spine.
Reporting under the old model was a dump of lead data, useless for decisions. The rebuilt system asked set questions every week, moved objectives every month and repointed directives every quarter, so the executive team could read performance in minutes. The clearest way to show the difference is side by side, from the platform reporting itself.
| Metric | Reading |
|---|---|
| Leads | 21 in the period |
| Cost per lead | £53.90 |
| Spend | £1,131.94 |
| Metric | Reading |
|---|---|
| Leads | 508 from one campaign |
| Cost per lead | £7.81 |
| Spend | £3,969.62 |
The agency-era account against one month of the rebuilt system.
Brought in-house
A function the firm owned.
Over the course of the year the model replaced the agency sprawl with an internal marketing department. Data ownership returned to the firm and reporting made performance visible for the first time; some agencies’ reluctance to share their numbers was itself a finding. Blended cost per lead under the old structure ran above £50. Guide-led capture on the new system generated enquiries from £7.81.
The outcome
£40M in revenue, and the structure to keep earning it.
The firm recorded a £5M month and passed £40M in revenue within the year, and underneath both sat an internal function running on directives, evidence and feedback long after the engagement closed. That structure is the part that lasts.
| Measure | Before | After |
|---|---|---|
| Cost per lead | Above £50 blended | £7.81 best performing |
| External agencies | Eleven | One internal function |
| Record month | £5M in sales | |
| Revenue | £40M+ inside the year | |
| Data ownership | Third parties | The firm |
Where does your brand sit against the business?
Every engagement starts with the Brand Alignment Diagnostic, a fixed first step that scores your brand against the business and shows what to fix first.