Vertical Fit, Measured
Same offer, same copy: education returned fourteen times the interested replies of law firms. The list is the strategy.
- Timeline
- 12 months · 5 campaigns
- Role
- Built and ran outbound for four brands on one system, with an Outlyst team
- Status
- Client engagement
- Primary stack
- Smartlead · Apollo · Google Sheets
Headline metrics
- What it tackles
- A UK signal-hardware brand sells a fix for a fault no database records: bad mobile reception decided by wall thickness and building age, in a market as wide as every commercial building in the country.
- What it delivers
- A vertical-testing programme that held one offer constant across four building types and mapped exactly where the product sells: 14 interested from 5,867 education prospects against 1 from 2,905 in legal, a ranked map the brand keeps long after the campaigns expire.
The problem in full
The client sells mobile signal boosting hardware in the United Kingdom, the equipment that fixes reception inside buildings with thick walls and poor coverage. It is a physical product for a physical fault, which creates a targeting problem no contact database can solve. Wall thickness, construction date and site layout decide whether a building has the problem, and none of those is a field anywhere. Headcount, funding stage, technology stack, everything a database is genuinely good at, are close to irrelevant here.
The addressable market made it worse, not better. Mobile signal affects nearly every commercial building in the country, so the list could in principle have been every company in the UK. A market that wide is not a plan. And the person who feels the fault is not the person who signs for the fix: whoever stands in the dead spot suffers it, but the facility manager or head of IT is who can authorise an engineer through the door. The question was never whether the email was persuasive. It was which buildings actually have the problem.
This brand was one of four businesses run by the same UK founder, and over twelve months all four ran on one shared outbound system. That context is what makes this engagement valuable: with the system, the operator and the mechanics held constant across four products, whatever moves the numbers has to be the market itself. This engagement is where that shows up cleanest.
System design
I designed it as an experiment rather than a broadcast, and ran it with an Outlyst team. The building type was the variable, chosen explicitly, and everything downstream was a consequence of it. The starting point was the installed base: the buildings the system had already been fitted into, colleges, hotels, offices, coworking spaces, are the only ground truth anyone has about where the fault lives. Each campaign named one building type, sized it before committing, because a vertical with a few hundred UK sites gives you an opinion rather than an answer, and fixed the readout in advance: interested replies over prospects, compared against the vertical before it.
Lists came from Apollo with three filters in a fixed order: geography, then industry, then title. UK only, because installation is on-site engineering work and a prospect abroad cannot be served whatever they reply. The industry filter was set once per campaign and never touched again, which is what makes one campaign comparable to the next. Titles prioritised the person who owns the building over the person who owns a software budget: Facility Manager and Head of Facilities first, Head of IT and IT Director second. Every list was deduplicated to unique people, so every figure here counts people, not sends.
The offer carried the campaign, and the offer was not the product. The first line asked whether the mobile signal at their site had been tested recently, a question about a place the reader knows better than anyone. The credential was an installation they would recognise, and the ask was a complimentary site visit by an engineer: concrete, physical, and the prospect keeps the answer whether or not they ever buy. Sequences stayed short with wording variants for deliverability, on dedicated warmed domains built for this brand alone, which is how a list of several thousand UK facility managers could be worked in one campaign without deliverability falling away.
Every reply landed by webhook in a single Google Sheets correspondence tracker, and a person read and tagged all of them: interested, opportunity, not interested, wrong person. Positive replies went to the company while the conversation was still warm, so an engineer visit could be booked against a live thread rather than a weekly export. Then the number that mattered: interested replies per thousand prospects, this vertical against the last. That number, not anyone's opinion, chose where the next few thousand names came from.
The calls that made it work
Sell the visit, not the hardware
Nobody accepts a meeting about signal boosters. Plenty of people accept a free engineer walking their building with a meter and telling them what their signal actually is. The ask was made free, physical and small, and it converted because it gives before it takes. It is the piece of this engagement I would carry to any hardware brand tomorrow.
The title had to own the building
The buyer here is not defined by seniority or budget but by authority over a physical site. Facility Manager and Head of Facilities outrank a CTO for this product, because they can authorise an engineer through the door and already know which floors people complain about. Pointing the list at building owners rather than software budgets is what made the replies actionable.
One building type per campaign
Copy and list rebuilt each time, but only one variable allowed to move: the vertical. A broad facilities and IT campaign established the baseline at 8.6% across 8,228 prospects, then coworking, legal and education each got their own clean reading. Four campaigns produced four comparable numbers, which is the entire reason the map is trustworthy.
Treat a settled vertical as an asset
Legal returned 1 interested reply from 2,905 prospects on the same offer, in the same month, as education's 14 from 5,867. That is not a campaign to fix, it is a question answered: 2,905 prospects is a complete reading of the vertical, not a sample. Closing legal and concentrating future spend on thick-fabric buildings is worth more than any copy revision could ever be.
The results in full
Across the four vertical campaigns, 17,621 unique UK decision makers were contacted and 1,015 of them replied, a 5.8% reply rate in a category where almost nobody was prospecting by email; the full engagement, including one small late probe, reached 18,540 people and 38 hand-tagged positive replies. The interest concentrated exactly where the physics says it should: education returned 14 interested from 5,867 prospects against legal's 1 from 2,905, fourteen times the interested replies from a list twice the size, and roughly seven times the rate. Old school buildings have terrible mobile signal, and the people who run them cared, immediately and in writing.
The companies that replied positively read like the map drew itself: Millfield, Sandwell College, New College, Halifax Academy, Woodard Academy, Arts University and Lycee International from education, alongside Gail's Bakery, Tilda, A G Barr and International Beverage, food and drink production being the other place thick fabric meets someone paid to care, plus workspace operators like Plexal, Trampery and WorkPad. A reply rate belongs to a campaign and expires with it. The ranked vertical map belongs to the company, survives every change of copy, sender and year, and points every future pound of prospecting at ground that has already answered.
Continue reading
The Market Decides
Same product, same email: 0.9% in the US and 2.8% in MENA. Market selection, measured.
Creator Content, Gulf-Wide
Arabic creator content sold across the Gulf, where offering something beat asking for something.
The Trust Purchase
LinkedIn ghostwriting sold to founders, opened with a free audit of their own profile.