Building a target list from technographic data

A sourcing brief describes the data an AI lab wants. A target list describes the companies most likely to have it and be willing to license it. Getting from one to the other is mostly a filtering exercise, and technographic data, the record of which tools a company uses, is the layer that does the most work.

This post walks through how we turn a brief into a list our reps can call, layer by layer. It is written for data-acquisition and research teams who want to understand what happens between sending a brief and seeing companies under contract.

Every list starts with the brief a lab sends us: sector, headcount, tool stack, years of history, regions and whether it wants operating or winding-down companies. Our post on how to write a good sourcing brief covers what makes those inputs useful.

The first step is translating each requirement into something a dataset can filter on. "Mid-size fintech operations teams in North America" becomes a set of industry codes, a headcount band and a list of countries. "At least four years of engineering history" becomes a founding-year cutoff plus evidence that engineering activity began early. Ambiguity in the brief shows up here, and it is cheaper to resolve with the lab at this stage than after calls have started.

Company data handles the broad cuts. Industry, headcount, location, founding year and ownership type are usually enough to take an unbounded universe down to a manageable starting set.

Headcount needs care. A lab asking for companies with 50 to 300 people usually means 50 to 300 people during the years of history it wants, not today. A company that was 200 people for five years and is now 30 after layoffs may be a better fit than one that just grew to 200. Where historical headcount is available, it matters more than the current figure.

Firmographics also flag companies that are clearly out of scope, such as subsidiaries whose records live in a parent's systems, or businesses whose operations happen mostly offline.

This is where the list gets specific. Technographic data records which tools a company uses or has used, inferred from sources such as job postings that name tools, public integration and status pages, and DNS and email configuration. For company data licensing, the tools matter because they decide what the dataset will contain.

A lab looking for long-horizon engineering work might require evidence of an issue tracker, a code host and a team chat tool. A lab building finance-workflow environments might require an accounting system, a spend management tool and a shared document store. The useful questions are:

• Which systems from the brief does the company show evidence of using?
• How many of them, and do they cover the workflow the lab cares about end to end?
• How long has each tool been in use, and did any switch partway through?
• Are the tools ones whose exports preserve threads, links and history?

Coverage across several systems matters more than any single tool. Connection depth, the same project traceable from chat to ticket to pull request to email, is what makes a dataset valuable, and a company running five connected tools usually has more of it than one running a single tool at higher volume. Our earlier post on targeting by tool stack goes deeper on one common combination.

A company can match on sector, size and tools and still be the wrong target. Two checks remove most of the false positives.

The first is history. Technographic signals can show when a tool appeared, which helps estimate how many years of records exist. A company that adopted its current stack eight months ago will not satisfy a brief asking for four years of connected history.

The second is status. Labs often specify operating companies, winding-down companies or both. Signals such as hiring freezes, leadership departures, office closures or public shutdown notices help separate the groups. Our post on intent signals explains why timing matters, since tools cancelled during a wind-down can take the data with them.

What remains is ranked, not just listed. Companies with the strongest fit across tools, history and headcount go into the first tier, and companies with gaps go lower. Tiering lets reps spend their time where a successful call is most likely to produce data the lab wants.

Each company then needs a person to call. For an operating company, that is usually a founder, CEO, COO or CFO. For a winding-down company, it may be a director, an officer, or a liquidator or assignee with authority over company assets. Mapping those people, and the best channel to reach them, is part of building the list. Our post on reaching the person who can sign covers that step.

Imagine a brief for 50 to 250 person SaaS companies in the US and EU, with at least four years of history across an issue tracker, a code host and team chat, open to both operating and winding-down companies. Firmographics narrow the universe, technographics confirm the stack, history and status checks remove poor fits, and tiering puts the strongest candidates first.

From there, our onshore reps across the US, Canada, the EU and LATAM call the people who can sign, using cold calling, email and LinkedIn. Companies that agree go through the same seller process: scope and exclusions annexed to the contract, a read-only evaluation, in-house anonymization with consistent tokens across every tool, and seller review of an anonymized sample. Data is delivered to the lab on a rolling basis as companies complete.

Technographic data does not tell you whether a company will say yes. It does make sure every call is to a company that, if it says yes, has the data the brief asked for.

If your team has a clear picture of the sectors, tools and history it needs, send us a sourcing brief and we will show you how it translates into a target list.