Lead generation
Three documented engagements. In every one, the account was generating leads the business could not close.
- Documented lead generation engagements
- 3
- Markets
- 3
- Ad spend managed
- $20M+
- Years operating
- 3
More leads is the easiest thing to sell you and the least likely thing to help. The number that matters is how many of them become clients.
The mechanism
Why lead volume is the wrong target
The platform optimises toward the event you send it. Send "form submitted" and it will become extremely good at finding people who submit forms. That is a genuinely different population from people who become clients, and the gap between them is where most lead generation budget goes.
This is not a subtle effect. In the three engagements on this page, the share of leads that were actually qualified sat at 41%, 53% and 31% when we took the accounts over. More than half of what those businesses were paying for could never have become revenue, and the accounts were optimising to produce more of it.
The mechanism is straightforward. A broad campaign — "legal help", "health and wellness", national rather than local — generates volume cheaply, so cost per lead looks excellent, so it attracts budget. In one account, national campaigns consumed 41% of spend and produced 19% of booked jobs. The reporting said they were the most efficient thing in the account.
The second problem is that the conversion the business actually cares about happens somewhere the platform cannot see. A closed case, a booked job, a signed contract — those live in a CRM, or a phone call, or a diary, weeks after the click. Unless that event is fed back, the platform is permanently optimising toward a proxy.
And phone calls make it worse. For legal, healthcare and home services, calls are typically a third to two thirds of all conversions. An account without call tracking is not measuring most of its own results. One account here had call tracking covering 61% of its network — the other 39% was invisible.
Proof
The engagements.
2750
Qualified Leads / Month
from 850
224% More Qualified Leads and 55% Lower CPA for MediPath Health
We took MediPath from a slow landing page and loose lead qualification to a system that nearly doubled the lead-to-appointment rate — while nearly doubling the share of leads that were fully qualified in the first place.
14850
Monthly Leads
from 8,500
74.7% More Leads and 41% Lower Cost Per Lead for SummitGuard Services
We took SummitGuard's network from inconsistent lead quality and invisible tracking across locations to a fully localized, accountable system — growing the network from 28 to 41 locations while making it more efficient, not less.
355
Monthly Case Intake
from 149
138% More Case Intake and 56% Lower Cost Per Case for SterlingLex
We took SterlingLex from real inquiry volume that too rarely became a real case to a qualified, high-intent intake system — and did it while cutting monthly ad spend by roughly 24%, not increasing it.
In order
What we change, and in what order
We close the measurement loop first.
Offline conversion import so that qualified leads and closed business flow back into Google and Meta. Full call tracking with dynamic number insertion. Enhanced conversions for leads. Google unified the web and lead implementations at Marketing Live 2026. That removed most of the friction that made this a project rather than a configuration.
Then we change what the account optimises toward.
Not form fills. Qualified leads where the volume supports it, closed business where it does not. This is the single change that separates a lead generation account from an ecommerce account with a different conversion event.
Then we tighten intent.
High-intent keyword structures, aggressive negative lists, and qualification questions early enough in the form that unqualified enquiries self-select out before they reach an intake team. One account here dropped unqualified enquiries from 47% to 22% and reduced monthly spend from £55,000 to £42,000 while case intake more than doubled.
Then structure by what the business actually sells.
By service and by location rather than by platform convenience. Multi-location businesses cannot be managed as one account; the account that grew from 28 to 41 active locations did so on location-level reporting and weekly reviews, not on better creative.
Then landing pages, because they carry more weight here than in ecommerce.
A slow or vague page in a considered purchase does not lose you a sale, it loses you the enquiry entirely.
Honest scoping
Who this is for, and who it isn't
This works where a lead has a knowable value and someone can tell you whether it closed. Professional services, healthcare, home services, multi-location operators, B2B with a defined sales cycle. It works best where a CRM exists and can be connected, because most of the approach depends on that.
It does not work if nobody tracks what happens to a lead after it arrives. We cannot optimise toward closed business that is not recorded anywhere, and we will not pretend that optimising toward form fills is the same thing.
It does not work well below roughly $5,000 a month, because lead generation accounts generate fewer conversions than ecommerce ones and the platforms' models need volume to learn. Data-driven attribution has a minimum conversion threshold; smaller lead generation accounts frequently sit under it, which limits what automation can do.
And it does not work if intake is the bottleneck. If leads are already arriving faster than anyone can call them back, more leads is not the problem to solve, and we will say so.
Asked and answered
Common Questions
Because volume and revenue are different problems. The question worth asking is what share of last month's leads your intake team would describe as real, and whether anyone has measured it. In the accounts here that number ranged from 31% to 53% before we touched anything.
You tell us, through your CRM. Offline conversion import sends that back to the ad platforms so they optimise toward closed business rather than form fills. Without it we are guessing, and we would rather set that up in month one than optimise against a proxy for a year.
Usually yes, and that is the point. Cost per lead falls when you buy worse leads. What should fall is cost per qualified lead and cost per closed client. We report all three, and we lead with the ones that got worse.
Yes. For legal, healthcare and home services it is not optional — calls are frequently the majority of conversions, and an account without it is optimising on partial data.
Substantially. Location-level structure, tracking and reporting are the work, and national campaigns almost always look more efficient than they are because they take credit for demand that already existed.
Longer than in ecommerce, because your sales cycle sits between the click and the result. Expect measurement changes in month one, lead quality changes by month two or three, and closed-business changes a full sales cycle after that.
What ran
Delivered across these accounts.
- Google Ads ManagementRead more
- Meta Ads ManagementRead more
- Landing Page DesignRead more
- CRO & A/B TestingRead more
- Tracking & AttributionRead more
- Creative ServicesRead more
Taken from the services-delivered table of each case study above, so this lists what actually ran rather than what could.
Get in touch
Tell us what you’re running.
Channels, monthly spend, and the part that isn't working. That is enough for us to tell you whether there is anything here worth a call — and we would rather say so now than after a proposal.
Keep going
The rest of it.
- The other oneEcommerceBlended ROAS, repeat purchase rate and cost per acquisition, across apparel, jewellery, beauty, food and electronics.Read more
- EverythingAll servicesPaid advertising across every major platform, Answer Engine Optimisation, and the conversion and tracking work that makes all of it pay.Read more
- ProofThe workEvery engagement, with the numbers, the approach, and the parts that did not work. Ecommerce and lead generation, in the currencies our clients actually bill in.Read more