Grow — customer acquisition
What we usually find
Platform-reported conversions, analytics and the accounts disagree. Every scaling decision becomes a judgement call made on numbers nobody fully trusts.
Campaigns are organised the way the ad interface suggests rather than around margin, capacity or the products that actually make money.
More traffic sent to a page that converts poorly buys more of the same result at a higher price.
Several things change at once, results are read too early, and nothing is learned that survives to the next month.
Step one
Nothing else in this discipline is reliable until the numbers are. Tracking is rebuilt and verified end to end — conversion events, values, deduplication between platform and analytics, and server-side events where the platform supports them.
The output is simple: a cost per acquisition you can defend in a conversation about the P&L, and a clear view of which channel produced which customer.
Events defined against real commercial outcomes, then verified against what the business actually banks.
Where the platform supports it, events sent from the server so measurement is less dependent on the browser.
A model that reflects how customers really arrive, with its limits stated rather than hidden.
Channels
Search captures demand that already exists. Paid social creates it. They behave differently, they are measured differently, and running them as though they are interchangeable is why multi-channel budgets so often underperform a single well-run one.
Campaign architecture follows the commercial model: what the margin is, which products can carry acquisition cost, and where capacity constrains volume.
Search, Performance Max and shopping structured around margin and intent, with search-term hygiene as an ongoing discipline.
Audience and creative strategy built for acquisition, with a testing programme rather than occasional new creative.
Sequenced by where someone stopped, and capped so it stops being a tax on customers who were going to convert anyway.
Conversion
Traffic cost is only half the equation. A landing page that converts at twice the rate halves the cost per customer without touching the bid — which is usually the cheaper of the two levers.
Landing page work is treated as part of the acquisition system, not a separate project handed to someone else: structure, message match, form design, speed and the specific friction the data points to.
Operating rhythm
Reporting is against the target that was agreed, in the terms the business uses. Tests are run one variable at a time and read when they have enough data to mean something.
When performance drops, the first question is what changed — in the account, in the funnel, in the market or in the operation. Budget is the last lever, not the first.
What we report on
Cost per acquisition
Against a figure the business can sustain.
Contribution
What each channel returns after the cost of serving it.
Conversion rate
Measured per step, so it is clear where people stop.
Channel mix
Where incremental spend still produces incremental customers.
Capabilities
How to engage
For businesses starting paid acquisition, or restarting it properly after an unsuccessful attempt.
Priced on scope, billed monthly
Delivery: Monthly service. Setup normally completes within the first 2–3 weeks; management continues month to month.
For established businesses that need consistent acquisition across more than one channel.
Priced on scope, billed monthly
Delivery: Monthly service, continuing month to month once onboarding is complete.
For larger multi-channel programmes where measurement and attribution carry as much weight as the campaigns.
Priced on scope, billed monthly
Delivery: Monthly service with a scope agreed at onboarding.
Questions
No, and anyone who does is either guessing or excluding enough cases to make the promise meaningless. What is committed to is a defined target, honest measurement against it, and a clear recommendation when a channel is not going to reach it.
Something not covered here? Ask directly.
If you know what you want built, send the brief. If the problem is still taking shape, a short call is the faster route. Either way you speak directly to the person doing the work.
Or email directly: hello@perspektivagroup.com