40M+
Companies in the Atlas
A map of companies Zylabs has modeled and the suppliers serving them. Companies shown include a US telecom operator with 1,906 suppliers graded by presence and share, a global bank with 456 suppliers across 1,536 dated placements, a US retailer with 334 suppliers on the public record and 50 more it cannot see, a health services group, a streaming service, a life insurer, an Australian grocer, an Indian conglomerate and a US payments network. Each line is a company-to-company relationship the Atlas holds, with the date it was read.
More than 40 million companies in the Atlas, reading now
The world’s largest living map of companies.
A database tells you what a company is. The Atlas keeps a current picture of what is happening inside it and what it depends on: who serves it, what it is building, who decides, what it runs, what it spends, and how each of those joins the companies around it. Name a company and we read it back to you.
Trusted and used by
Sparkle
Sparkle1M+
Suppliers on the graph.
60+
Countries and markets covered
Our perspective
What is Zylabs?
Zylabs keeps a living map of what is happening inside companies and how they depend on one another. Where a database stops at what a company is, the Atlas says what it is doing, who serves it and what it depends on, with a method and a date on every line. We never read a company on its own: every company connected to it is read too, because each one leaves evidence about it.
- How is this different from a database?
- A database describes a company by what it is, using a sector code, a headcount band and an address, while the Atlas records what the company is doing and what it depends on.
- Where does the picture come from?
- We read more than 60 channels on a schedule, split between what a company and its people write themselves and what the company is compelled to file or a third party is compelled to record. The model then derives what no channel states by triangulating across them, which is where a supplier’s trend, a team’s size and a wallet band come from. Every channel carries a stated blind spot that is read before the channel is called.
- How current is it?
- Every line carries the date it was last read, and a supplier placement or a seat is re-read on a cadence of days to weeks while a disclosed figure moves when the company next reports.
- Can I check it against an account I know?
- Checking a record is what the demo is for. Name an account you know cold, we read it back to you line by line, and you tell us which lines we got wrong.
“How did you get that information?”
Consulting partners, IT services sellers, security founders and procurement heads have all asked us this inside the first ten minutes of a first meeting, usually in stronger language. The answer is a six-stage method that we publish in full, and a source discipline we publish the shape of without naming a single source.
“How the hell did you get that? I have run this account for four years and I did not know half of it.”
“We knew we had a lot of vendors. We did not know we had that many, and we certainly could not have named them.”
“The program on slide four is one nobody outside the building was supposed to know about. Where does that come from?”
“Every research vendor shows me a stack. You showed me what they are walking away from and who is being paid to move it.”
“Three of the suppliers on that page I had never heard of, and I sign the contracts.”
“We have never published a client list. How did you get this information?”
The Atlas methodology
One company becomes one record, in six stages.
A company controls what it chooses to publish about itself. It controls neither the record it is required to leave elsewhere nor the account of the work written by the people who did it. The six stages below are how that becomes one graded, dated record, and how that record joins every other record on the map.
This is the research method a strategy firm runs by hand, on one account, over six weeks. We built it to run on a hundred thousand accounts, on a schedule.
Stage 01 · Resolve46 aliases collapsed to one entity
01 · RESOLVE
What is this company, actually?
We separate the legal entities, trading names, subsidiaries, acquisitions and the units that buy on their own budgets before anything else runs. A picture that fails here cannot recover later, because a group read as a single account produces a wallet estimate wrong by an order of magnitude and a stakeholder map in which nothing has an owner.
46 aliases collapsed to one entity on the retailer record
02 · READ
What has this company been compelled to leave behind, and what have its people written about the work?
A company controls what it chooses to publish about itself, and it controls neither the record it is required to leave elsewhere nor the account of the work written by the people who did it. We read both. Primary sources are what the company and its people write themselves, which covers job specifications, engineering writing, conference talks, tender documents and practitioner accounts. Secondary sources are what the company is compelled to file and what third parties are compelled to record about it. Each channel runs on its own cadence and carries a stated blind spot that is read before the channel is called.
60+ channels, and 12.6 million records screened to build one supplier map
03 · TRIANGULATE
What happens when two channels disagree?
The same question is put to channels that fail in different ways, because a statutory filing, a job specification and an account written by the person who did the work are each wrong about different things. Where they point the same way the reading is stronger. Where they conflict, the record carries both readings with their dates rather than resolving them into a single value that neither supports. The record also states which classes of evidence were silent on the account, because a channel that cannot see a kind of company returns nothing and nothing reads exactly like a clean account.
Both conflicting readings published side by side, with their dates, rather than resolved
04 · DERIVE
What is true that no source states?
The model computes what no channel reports. Supplier presence and trend come from evidence density over time. A wallet band comes from top-down and bottom-up benchmarks reconciled against each other rather than averaged. Team size comes from record counts. Program phase comes from the language of the documents that name it. The gap between what a company says and what its hiring does falls out of the two read side by side. This stage is what separates a record from a search result, and it is the stage whose machinery we do not publish.
50+ of the 123 catalogue fields held by a deep record at the median, more than 10 of them carried by evidence
05 · GRADE
How much weight does this line carry?
Every line carries the method that produced it and the date it was read, shown wherever the line is shown. Precedence between the four methods is fixed, so a stronger method is never overwritten by a weaker one however recent the weaker reading is, and a modeled value is labeled modeled everywhere it appears.
14M+ graded facts held in the Atlas
06 · CONNECT
Who does this company depend on, and who depends on it?
We never read a company on its own. Every supplier, client, subsidiary and partner connected to it is read as well, because each of them leaves evidence about the company that it never publishes itself. A supplier's clients and a client's suppliers are the same evidence read in two directions, which is why a services firm that has never published a client list still has one here, dated, with the delivery location attached.
1M+ company-to-company relationships
The whole network, not one company
We never read a company on its own.
Every supplier, client, subsidiary and partner connected to a company is read as well, because each of them leaves evidence about the company that it never publishes itself. A supplier’s client list, a partner’s case study and a subsidiary’s filing all describe the account from the outside, and reading them together is what turns a thin profile into a complete picture.
A supplier’s clients and a client’s suppliers are the same evidence read in two directions. Switch the direction below.
A deep record holding 1,906 suppliers graded by presence and share. Every placement carries the year it was first seen and the year it was last confirmed, back to 2018.
What a record holds
Six blocks, every fact dated and graded.
01
Who serves it
Every supplier on the account, by unit and work area, graded on evidence recency, back eight years.
334 on the public record, 50 more it cannot see02What is in flight
Programs by function, with phase, delivery model, owning team and how visible each one is from outside.
10 programs, 7 in technology03Who decides
The executive layer and the seats beneath it, each verified in role, with the reporting line graded separately.
15 seats verified, 2 named unresolved04What it runs
The estate with a lifecycle state per component, and the named target wherever one exists.
18 components, 4 sunset or migrating05Which teams
The join of the other four: lead, size, open seats and the stack being hired against.
7 teams, sorted by open seats06What it spends
A wallet band by function, cross-checked two ways, with the benchmark and where it misapplies.
Technology $863M to $2.6BA large US retailer
CompanyDisclosed
$86.3BProvenanceMethod: Disclosed · confidence 0.95Corroborated across 3 independent channelsAs of: FY ended 30 Jan 2026Disclosed, not modeledrevenue
167,000 employees · 1,759 stores · capex $2.21B, up 15 percent
Who serves it8 years of record
334suppliers on the public record
Plus 50 it cannot see. 15 with evidence inside twelve months.
What is in flightDated
Analytics migrationMigrating · Jun 2026
Customer identity (CIAM)Building · Aug 2026
AI governanceEvaluating · Aug 2026
What it runsInferred
Teradata→ BigQuery
MicroStrategy, CognosSunset
Ping IdentityInvesting
Who decidesVerified in seat
EVP, Chief Digital and Information OfficerSince Sep 2022
VP, AI, Data and GovernanceNew seat · Jun 2025
MD, India engineering centerSince Sep 2020
What it spendsModeled
$0.9Bto $2.6B technology
Top-down and bottom-up agree within 17 percent. Acquisitions carry separate wallets.
Every line carries the method that produced it and the date it was read. Hover a figure.Open the full record →
The depth, shown
A relationship is a shape, not a logo on a slide.
Two reads from the Atlas, drawn exactly as the record holds them.
Each bar is how much evidence placed that supplier on the account in that year. The shape is the relationship: a bench that grew hard in 2019, thinned through the pandemic and rebuilt from 2023. The record holds this for all 1,906 suppliers.
A component recorded as "being replaced" with no target named carries half the information. Naming both states at once states the incumbent, the direction of travel, the fact that budget has cleared, and the delivery capacity the move will need. Owned by the store operations AI and innovation team, hiring against two open seats.
How this compares
Five things people put next to us, and what each one cannot answer.
Buyers arrive having already bought something adjacent, so the useful question is not whether the Atlas is better but which question each tool can actually answer. These five come up in most first meetings.
| What you may already own | What it answers well | What it cannot answer | |
|---|---|---|---|
| 01 | A contact and firmographic database | Who works at a company, its sector, its size band, its address and a revenue estimate, at very large scale. | What the company is doing right now, who serves it, or what it is moving away from. It describes a company by what it is, which is the one thing that changes least. |
| 02 | An intent platform | Which accounts showed activity against a topic somewhere on the open web. | What the activity was about, who inside the account caused it, or whether anyone with budget was involved. The signal also thins as buyers do their research inside assistants, where nobody is watching. |
| 03 | A technographics feed | Which products a company appears to have installed. | Which of them is being sunset, what is replacing it, which team is funded to do the work, or by when. A stack is a photograph, and the useful part is the direction of travel. |
| 04 | An analyst subscription or an expert network | What an informed human believes about a market or an account, with a name attached and a view you can challenge. | Run across a hundred thousand accounts, refresh on a schedule, or return an answer in less than a week. Accuracy is high and the cost per account makes it a tool for a handful of decisions a year. |
| 05 | A research agent pointed at the open web | What the web says about a company today, quickly and at almost no cost. | Reach the records a company is compelled to file but nobody indexes, hold a history it can compare against, grade its own confidence, or return the same answer twice. Reading is the cheap part, and knowing what to read for is not. |
Where we sit
Four things the Atlas does that none of them do.
01
We record what a company is doing, not what it is
A sector code and a headcount band describe a company at rest. The Atlas holds the estate caught mid-migration with both states named in one document, the program in flight with its phase and its owning team, and the seat created with no predecessor because the function did not exist last year.
02We compute the lines no source states
No channel reports that a supplier relationship is cooling, that a team is about forty people, or that a wallet sits between $863M and $2.6B. Those are derived from evidence read across classes, with the basis stated and the result labeled modeled wherever it appears.
03We read the map in both directions
The evidence that builds an account’s supply base is the same evidence that builds a supplier’s client list, so a services firm that has never published a client has one here, dated, with the delivery location attached. Neither party controls that record.
04We publish the method and what each instrument cannot see
Eight classes of evidence, each with a stated blind spot, a grade and a date on every line, the volume read for an account, and which classes were silent on it. A thin record labeled thin is useful, and a thin record presented as complete is not.
Where we are not the right answer
- If you need verified phone numbers and mailing addresses at scale, buy a contact database, because we do not produce them.
- If you need every registered company in a country down to the sole trader, buy a registry.
- If you need a named person to defend a number in front of a board, buy an analyst.
The Atlas is the layer that says what is happening inside an account and what it depends on, and it is worth having when that is the question you are stuck on.
8 questions the Atlas answers
Different teams read the same record and take different decisions from it.
Sales intelligence
Which accounts should we call this quarter, and what do we say?
Rank the accounts in a territory by whether a program you sell into is running now.
Account-based marketingWhich accounts, which seats, and which message for each?
Build a target list from accounts running a live program that matches what you sell.
Competitive intelligenceWho is inside our competitor’s accounts, doing what, since when?
Build the client list of a competitor that has never published one.
Product intelligenceWhat are they building, on what, and what are they walking away from?
See which components a company is sunsetting, and what is named as the replacement.
Market intelligenceWhere is the category going, across every company in it?
Count the accounts in a category adopting a platform, and the accounts leaving one.
Account expansionWhere can we grow inside the accounts we already have, and which are at risk?
Watch the accounts you already serve for a program in a function you do not yet cover.
Champion trackingWhere have our champions gone, and who just arrived in an account we care about?
Follow a champion who leaves a customer, and see where they land.
Partner co-sellingWhich partner is already inside this account, and what should we take to them?
Find which integrators and outsourcers are already placed inside a target account.
Ask for a demo
Name an account and we read it back to you on the call.
Tell us the company you want read and what you sell. Before the call we build the record. On the call you see it: the suppliers, the programs, the seats, the estate, the wallet, with the evidence behind each line.
Pick an account you know cold, because the useful demos are the ones where you can tell us which lines we got wrong.
The founder note
A living map of every company.
Every commercial decision is a bet on what is happening inside another company. For decades the picture came from analysts, who were accurate and expensive, or from databases, which were cheap and described a company by what it is rather than what is happening to it. The cost of reading fell 280-fold in two years, and what was left was knowing what to read for.
Read the note →Common questions
Common questions
Zylabs keeps a living map of what is happening inside companies and how they depend on one another. Where a database stops at what a company is, the Atlas says what it is doing, who serves it and what it depends on, with a method and a date on every line.
The Atlas is a current, evidenced picture of what is happening inside a company, covering who serves it, what it is building, who decides, what it runs and what it spends, with each record joined to every other company it deals with. Every line carries the method that produced it and the date it was read.
A database describes a company by what it is, using a sector code, a headcount band and an address, while intent data scores it on web activity that is disappearing as buyers move into assistants. The Atlas records what the company is doing and what it depends on, and it computes the things no source states rather than reporting only what someone else already published.
We read more than 60 channels on a schedule, split between the primary documents a company and its people write themselves and the records it is compelled to file or a third party is compelled to keep. We publish the classes of evidence, the grade on every line, the volume read for an account and the stated blind spot of every class, and we withhold the channel list itself, because that is the part a competitor would copy.
More than 40 million companies are entity-resolved and connected on the map, with more than a million dated relationships between them. If the company you care about has not been read to depth yet, we run it before your demo.
Checking a record is what the demo is for, so name an account you know cold and we will read it back to you line by line. Anything you correct goes into the record with its date, and every claim that rested on the line we got wrong is withdrawn with it.