Capabilities
The layer underneath the whole system.
Every stage of the engine — map, reach, publish, capture, measure — runs on four systems the operator wires together and keeps running. This is what they are and how they connect.
Every stage of the engine — map, reach, publish, capture, measure — runs on four systems the operator wires together and keeps running. This is what they are and how they connect.
Data moves in one direction. Research produces the sector map, the verified list, and the content. Outbound and the site turn that into inquiries. The CRM records every one of them with its sector, buyer role, and source query attached. Analytics closes the loop: what was read, what was replied to, what it cost, and therefore what to research next. Each of the four is ordinary on its own. Wired together in this order, they are the engine.
Research produces the subsector map, the verified company list, the buyer roles, and the first drafts. Lists go to outbound. Content goes to the site.
Outbound and the site produce inquiries. The CRM records every one with sector, subsector, buyer role, and source query attached, and routes it the same day.
Every page read, sequence reply, and form submission is captured against the same properties, so the site and the CRM tell one story.
Cost per qualified inquiry by sector and channel, on your own numbers. The answer decides what to research and publish next.
HubSpot is the default. Pipedrive or Close when the sales team wants less system and more pipeline view. Zoho when it is already in the building. The tool matters less than the architecture inside it, and a default install has the wrong architecture for a technical product: it scores page views, it assumes a quarterly deal, and it does not know what the engineer was looking for on arrival.
We build the properties that carry sector, subsector, buyer role, and source query on every contact. We score on technical signals — the spec table compared, the standard page read, the application note downloaded — not on visits. Deal stages match an episodic capital purchase on a multi-year replacement cycle — three to seven years in test and measurement — with the long evaluation built in rather than forced through a template. Replies from outbound sequences and forms from the site land in the same record. An inquiry from an engineer reaches the right person the same day, and a weekly export puts the data in your hands regardless of who runs the seat.
Sector, subsector, buyer role, source query, and the standard referenced, on every contact and company. Reports segment by what the buyer wanted, not by lead source alone.
A spec table compared, a standard page read, an application note downloaded. Those are evaluation behaviours. Visits and email opens are not, and they do not score.
Stages that match an episodic capital purchase: inquiry, technical evaluation, budget window, procurement. Nothing gets marked lost because a template expected a quarterly close.
An inquiry from an engineer reaches the right person the same day, with the page that prompted it attached. Sales sees the question before the call.
A lead count says nothing about a technical product. An engineer who downloaded a brochure and an engineer with a budget window are the same row in most reports. We compute what the growth system costs per qualified inquiry, by sector and by channel, and we carry the arithmetic through to net: the cost to acquire the customer against what that customer is worth over an episodic purchase, including service and consumables where they exist.
The arithmetic is done on your numbers, in a sheet you keep. Margin, cycle length, and average order come from you; channel cost and inquiry volume come from the CRM and the analytics layer. CAC payback on a purchase that recurs every few years looks nothing like payback on a monthly subscription, and a budget built on the wrong model buys the wrong channels. The output is a number the CEO can sign against.
Not cost per lead. An inquiry counts when it carries a sector, a buyer role, and an evaluation signal. Everything else is traffic.
Payback on a purchase that recurs every few years is a different calculation from payback on a subscription. We build the one that matches how your product is bought.
Outbound, search, LinkedIn, and AI-answer citations, each with its share of qualified inquiries and its cost. The channel that looks cheapest per lead is rarely cheapest per customer.
One sheet: what each channel costs, what it returns, and how long the money is out. Built on your margin and your cycle, so the number holds up in a board meeting.
A default analytics install counts sessions. Sessions do not tell you that an engineer opened the datasheet, read the page for the standard, and downloaded the application note in one visit — which is what a shortlist forming looks like. We build GA4 through Google Tag Manager with events named for those actions, and we add server-side tagging where ad-blockers and privacy tools undercount, so the numbers the CRM scores on are the numbers that actually happened.
Search Console tells us which queries bring engineers, by application and by standard number. Microsoft Clarity shows how they read the page once they arrive. Dashboards live in Power BI or Looker Studio, in your account. A weekly report goes out on its own. A monthly review turns the numbers into one decision: what to publish next.
Datasheet opened. Standard page read. Application note downloaded. Named for what an evaluating engineer does, so a score in the CRM means what it says.
Engineering networks run blockers and privacy tools. Server-side tagging keeps the events that client-side scripts drop, so the count is closer to the truth and the CRM scores on it.
The queries that bring engineers, by application and by standard number. The list becomes the next pages to publish and the next properties to add.
One Power BI or Looker Studio view in your account: inquiries by sector and channel, evaluation signals, pipeline, cost. Weekly on its own; monthly with a decision attached.
n8n, Make, or Zapier where a trigger and an action are enough. Custom Claude-based agents where the work is judgement: mapping a sector into subsectors, verifying every company on a list against its live site, profiling the buyer roles, cutting a first draft from the intelligence base, flagging an anomaly in the numbers, writing the weekly report. The agents do the volume. The operator checks the output. Your marketing lead approves it in a shared Notion cockpit before anything ships.
We do not sell it. We show what it built: the sector maps on this site, and before that, 220+ technical articles and a 14,553-lead database for a technical product company whose entire growth stack was built solo. It runs under your team, on your sectors, with the same checks.
A sector becomes subsectors with demand signals, governing standards, and the queries engineers type. The map is the base every email, page, and post is cut from.
Every company on a list is checked against its own website before it becomes a prospect: makes and sells its own products or not, in the sector or not, still trading or not. Bad rows never reach outbound.
Application pages, standard pages, and sequence copy start as drafts cut from the map. Your engineers correct the substance; nobody writes from a blank page.
Enrichment and hygiene run on a schedule. The weekly report writes itself from the CRM and analytics. An anomaly — a deliverability dip, a form that stopped firing — raises a flag before the month is lost.
A system is defined as much by what it leaves out. Three things this practice does not offer.
We do not run paid media as a standalone service. Search and LinkedIn campaigns run inside the system where the map shows the query volume exists, sized to the inquiry cost they produce. Nobody here is paid on spend.
We do not run online retail. Carts, catalogues, and retail channels are a different discipline with different mechanics. We do not offer it, and we do not stretch the growth system to cover it.
We do not replace your engineers as the source of technical truth. Every page and every sequence gets its substance from them. The system’s job is to get that knowledge onto the page in the buyer’s vocabulary, with a queue short enough that they answer it.