For people who find companies for AI data buyers. Technographic data tells you which software a company runs. In sales it hints at budget and fit. In an AI data deal it is closer to an inventory, because buyers take operating records out of named systems: Mode's published list of the sources it buys includes Slack, Jira, Salesforce, HubSpot, Zendesk and QuickBooks. This guide covers what technographic data is, what a technographic record can and cannot tell you, how to judge a provider, and how to use system signals to qualify a target.
Want the systems a company names itself? Check it against buyer rules, free.
Not all company data is worth the same. Our research shows AI labs pay most for the records of real work, held in the systems below. We identify which of these systems each company runs, and it is one of the most important factors in the AI Data Value Score.
Every company comes with the systems we identified, shown as tags. Company level only: we show which systems a company runs, never what is inside them.
Technographic data is a record of the technology a company uses: its software, platforms and infrastructure, company by company.
A typical technographic record pairs a company with a product, the product's category, a first-seen and a last-seen date, and sometimes a confidence level and the evidence behind the detection. Sales teams buy it to find companies that run a competitor's product or a complementary one. Software vendors use it to size a market, and IT service firms use it to find companies whose systems they know how to support.
For anyone sourcing companies for AI data deals, the question changes. You are not asking what a company might buy next. You are asking what records its work leaves behind, and in which systems. A company that has run its support desk, CRM and project tracking in the same tools for eight years holds a connected history. A company that works out of inboxes and a shared drive holds something thinner, whatever its headcount.
It helps to think of a company's technology in four layers. They differ in two ways that matter to you: how visible they are from outside, and how much they say about the records a buyer would take.
Content management, analytics, tag managers, ecommerce, chat widgets and hosting. The easiest layer to detect and the least useful here: it describes the website, not the work.
Highly visibleHelp centers, client portals, logins, booking, payments and document upload. Partly visible, and a good sign that customer records, tickets and orders exist behind them.
Partly visibleChat, email, documents, work tracking, CRM, support, accounting, payroll, design and code. Where the records buyers want actually live, and almost entirely hidden from outside.
Mostly hiddenCloud hosting and the email and office suite a company works in. Useful mainly as a clue to how the team works day to day, not to the records it keeps.
Partly visibleBuyers do not buy "data" in general. They take agreed copies of records from specific systems, over agreed date ranges.
Mode's published list of the sources it buys names Microsoft, Gmail, Outlook, Google Drive, Slack, Jira, Confluence, Salesforce, HubSpot, Zendesk, Notion, Monday.com, Dropbox, Zoom, AutoCAD, Figma, NetSuite, Paychex, QuickBooks and ServiceTitan, and notes that many more are accepted. Mode also states that it agrees on the systems and date ranges before any export (as published, checked 8 October 2026). micro1 lists mature operations, documented processes and modern software tools among the things it looks for.
The press has noticed the same thing from the other side. Forbes (16 April 2026), Fast Company and Gizmodo covered startups selling their old Slack and email. For a broker, that makes a company's systems the most concrete clue to what it could offer, and to how hard the export would be.
| System type | Examples | Records it holds | Why a buyer cares |
|---|---|---|---|
| Chat and collaboration | Slack, Microsoft Teams | Threads where problems are raised, argued over and solved | Shows how people coordinate real work, in their own words |
| Email and calendar | Gmail, Outlook | Business correspondence, meeting invites and follow-ups | Often the longest continuous record a company has |
| Documents and knowledge | Google Drive, Confluence, Notion, Dropbox | SOPs, playbooks, templates and internal wikis | Written procedures are among the records buyers name first |
| Work tracking | Jira, Monday.com | Issues and projects from opening to close, with comments | A task from start to finish, with every decision in between |
| CRM and sales | Salesforce, HubSpot | Deal histories, account notes and pipeline stages | The commercial reasoning behind deals won and lost |
| Customer support | Zendesk | Tickets, replies, resolutions and macros | A question, the work to answer it and the outcome, linked |
| Finance and payroll | QuickBooks, Xero, NetSuite, Paychex | Invoices, reconciliations, month-end closes and approvals | How money decisions are recorded, checked and corrected |
| Calls and meetings | Zoom | Recordings and transcripts | Spoken decisions that never reach a document |
| Design, field work and code | AutoCAD, Figma, ServiceTitan, GitHub | Drawings, design files, job records and repositories with history | Specialist work with a visible trail. Troveo cites about $5,000 per code repository in the shut-down startup market. |
Technographic providers do not see inside a company. Each record is an inference from outside signals, and every provider keeps its own approach to itself. What you can judge is the record: what it claims, how fresh it is, and whether it covers the internal systems AI data buyers name.
| Question to ask | Why it matters for an AI data deal |
|---|---|
| Is it an internal system or a website tool? | Analytics tags and chat widgets describe the website. Buyers pay for the records in internal systems: ticketing, CRM, finance, code and documents. |
| When was it last confirmed? | Companies switch tools. A record from two years ago can point at a system the company has already left, with its history exported elsewhere. |
| Does the company use it, or sell it? | Software companies list the tools their product connects to. Those are integrations, not systems the company runs for its own work. |
| Can you explain it to someone else? | A record you cannot explain to a buyer, or to the company itself, is hard to act on and easy to dismiss. |
| Does it cover the systems of your sector? | Clinical, legal and fleet systems are niche. Strong coverage of mainstream sales tools says little about a lab, a law firm or a carrier. |
Missing signals cost you good targets. Wrong signals are worse: they send introductions to companies that never ran the system at all. These four account for most of the wrong ones.
A software company lists fifty tools its product connects to, and every one shows up as a detection.
Separate tools a company runs from tools it connects to.
A consultancy that implements a CRM for clients is listed as running that CRM itself.
Read the context of each mention. "We implement" and "our team works in" are different facts.
A tag on a holding group's site is credited to every operating company in the group.
Tie each signal to the company that would sign. The operating company's own domain and careers page are the better source.
A CRM first detected in 2024 is read as a long history, because the company was founded in 2005.
History in a system starts when the system did. A migration can leave older records in an archive, or lose them.
Whether you are comparing affordable technographic data providers or enterprise contracts, the same tests apply. Price only matters once you know the data answers your question.
| Test | What to ask | Why it matters for AI data sourcing |
|---|---|---|
| Internal system coverage | Which categories do you detect: chat, work tracking, CRM, support, accounting, payroll? | Web-stack coverage is common. Internal systems are where buyers take records from. |
| Evidence per detection | Can I see where each detection came from, and when? | Separates a job ad from last month and a tag left on a page years ago. |
| First seen and last seen | Do you keep both dates for each product at each company? | Last seen says whether a system is still in use. First seen hints at how much history it may hold. |
| False positives | Do you separate tools a company uses from tools it integrates with or sells? | Integration pages are the largest single source of wrong signals at software companies. |
| Company matching | How do you map a domain to a company, and subsidiaries to parents? | The company that signs must be the one whose team runs the systems. |
| Small company coverage | What share of your records are companies under 200 people? | Buyer programs publish size rules between 6 and 30 employees, and small firms leave fewer signals. |
| Refresh | How often is a domain checked again, and can I request a fresh check? | Stacks change. A stale detection costs you an introduction and some credibility. |
| License and use | May I use results to qualify companies for a third party? | A broker's job is passing company names to buyer programs. Some licenses forbid exactly that. |
Technographic products are usually sold in one of three ways: an annual contract with seats and exports, a credit or lookup plan, or a bulk file delivered on a schedule. Technographic data providers for enterprise tend to lead with the first: broad market coverage inside a CRM, for a large sales team. That is worth paying for when the whole market is your territory.
A sourcing team for AI data deals rarely needs the whole market. It needs depth on a shortlist, a clean API, and the option to score a whole shortlist in batches rather than one company at a time. That is closer to a lookup plan than to a seat license, and it is why a broker searching for an affordable provider is usually asking the right question.
Our plans follow that shape. Basic $99 a month: 5,000 lookups. Pro $299 a month: 25,000 lookups. Scale $799 a month: 100,000 lookups. Each plan scores the domains you send, one at a time or in batches of up to 100 per call, and batch scoring is on every plan; the plans differ only by lookups. Company lists are a separate, one-time purchase: each covers one of 21 US sectors and holds the top 3,000+ companies of that sector by AI Data Value Score, as a CSV snapshot with no updates, from $199 a list. There is no free plan or trial; the demo is free and limited per day, so you can see a real result before you pay.
Technographics come second. A company that misses a published size or country rule does not become a fit because it runs Salesforce. Once the firmographic filter is passed, systems decide which targets deserve an introduction first.
Confirm size, country, language and years against each program's rules. As published, checked 7 October 2026: micro1 lists 30+ employees, US first and primarily English; Mode lists 20+ full-time US office employees, 10+ for accounting firms and 6+ for law firms, and several years of records the company owns. The firmographic side is covered on firmographic data providers.
List which categories the company is likely to run: chat, work tracking, CRM, support, finance, documents. Two or three connected categories are a stronger signal than one well-known logo, and they show how an export would be scoped.
Run check your company. It reads the company's public pages and shows the systems it says it uses, each with the quote it came from, such as a careers page line about working in Jira and Slack every day. In the same pass it compares team size, location and history with the published rules of micro1 and Mode. It takes about 20 seconds.
Look the domain up in the AI Data Value Score. The operational systems and customer systems factor groups appear next to the likely data assets, history and activity status, so you read systems in the context of the whole company.
A system adopted last year holds a year of records. A company winding down or acquired may hold many years of them, but the decision may now sit with a new owner or an administrator. Every score shows the activity status: active, winding down or acquired, parked, or unreachable.
Introduce companies you know, with consent from someone authorized, through each program's own referral process. Never ask a company for samples or exports to "show the buyer". The buyer runs discovery under its own agreement with the company.
The company, its quotes and its result are invented to show how the signals fit together. Nothing here describes a real firm.
{
"domain": "lakeshore-residential.example",
"data_asset_score": 64,
"grade": "C",
"status": "active",
"history": { "first_seen_year": 2006 },
"likely_data_assets": [
{ "type": "customer_accounts", "confidence": "high" },
{ "type": "crm_records", "confidence": "medium" },
{ "type": "transactions", "confidence": "medium" }
],
"factor_groups": {
"operational_systems": "medium",
"customer_systems": "strong",
"knowledge_assets": "weak",
"history": "strong"
}
}
Fictional, shortened response shown only to explain the fields. Not a real result, a valuation or an offer. Full field list in the API documentation.
How a broker reads it: the size, location and history clear two programs' published rules, and the company names a CRM and an accounting system in its own words. A strong customer systems group next to likely portal and transaction records fits a buyer that takes operating records. Knowledge assets are weak, so written procedures may be thin or simply not public: a question for the company, not a reason to drop it. Next comes an introduction through a program's referral process, with the owner's consent, and nothing more. The buyer, not the broker, asks for a manifest and samples.
The AI Data Value Score rates any company from 0 to 100 for data value to AI buyers. It is built on our index of 120 million domains, 99.99% of the active internet, with domain history, and it looks at nine factor groups. Two of them are about systems.
We publish the names of the factor groups, never how they are measured or weighted. In an API response each group comes back as strong, medium, weak or none, next to the score, the grade, the data the company likely holds, its history and its activity status.
Use both if you have both. A technographic provider can name the product; the score tells you whether the whole company is worth an introduction. A strong systems group next to likely support tickets or CRM records is a reason to look closer. A weak or none result is not a no, because internal systems are often invisible from outside. Scores are estimates from public signals: not valuations, not offers, and not proof that a company wants to sell.
Last updated: 8 October 2026. All three plans use the same index and return the same fields, including the operational systems and customer systems groups. They differ only in lookups; batch scoring of your own domain lists is on every plan, and no plan includes company lists.
Enough to score every target a broker shortlists in a month, with history, status and the systems groups on each one, before any introduction is made. Batch scoring of your own domain lists, up to 100 domains per call.
Choose BasicFor brokers who score several shortlists a week. Batch scoring of your own domain lists, up to 100 domains per call.
Choose ProFor partnership and sourcing teams at data companies that score large target files. Batch scoring of your own domain lists, up to 100 domains per call.
Choose ScaleA business that sells records of the software and technology companies use: product names, categories, first-seen and last-seen dates, and sometimes a confidence level and the evidence behind each detection. Each provider builds its records from outside signals in its own way, so ask what each record claims and when it was last confirmed.
Rarely directly. Those systems run behind a login, so outside detection depends on indirect clues such as job ads, help pages or the company’s own words. The free check against buyer rules shows the systems a company says it uses on its public pages, each with the quote it came from.
Because they buy records from named systems. As published, checked 8 October 2026, Mode’s list of the sources it buys includes Slack, Jira, Salesforce, HubSpot, Zendesk, NetSuite and QuickBooks, and Mode agrees on the systems and date ranges before any export. Systems tell a broker which records could exist and how they would leave the company.
Pay for what the job needs: coverage of internal system categories, evidence and dates on each detection, and the right to use results to qualify companies for third parties. For a broker, a lookup-based plan on a shortlist is usually a better fit than a seat contract that covers the whole market.
Usually seats, CRM integrations, scheduled bulk delivery and broad market coverage. For AI data sourcing, the enterprise features that matter are an API, a bulk export of a segment and license terms that allow sharing a company’s name with a buyer program.
Operational systems and customer systems are two of its nine factor groups. The others are history, scale, knowledge assets, organization, industry value, expertise and activity status. Each group is reported as strong, medium, weak or none, next to the score, grade, likely data assets, history and status. The weights and rules that combine them are our own.
No. Internal systems are often invisible from outside, and every score is an estimate from public signals: not a valuation, not an offer and not proof that a company wants to sell. Confirm with the company’s own words first, then with the company itself, with its consent.
No. There is no free plan and no trial. The AI Data Value Score demo is free and limited per day, and the check against buyer rules is free. Plans start at $99 a month for 5,000 lookups, paid by PayPal or card, and the API key appears in your dashboard after payment.
Check what a company says it runs, with quotes, against published buyer rules. Score it for data value to AI buyers. Pick a plan when you are ready to do it at volume.