How it works

Nobody else shows you this

Every provider in this category claims accuracy. Almost none will tell you how they arrive at it, because for most of them the honest answer is that a crawler found it and no person ever checked.

Here is exactly what happens between your brief and your delivered file.

01 · Why we publish this

Two reasons

The first is self-interested. A guarantee is only worth something if the buyer can see why it is possible. "100% accurate" from a company that scrapes the web is a marketing number. The same claim from a company that telephones every contact before delivery is a description of a method. You cannot tell those apart unless we show you the method.

The second is that the method is not the hard part. Anyone in this industry could describe what we do. Building it is different: telephone researchers who can hold a conversation with a switchboard in the right language, and years of accumulated judgement about which sources are worth trusting for which markets. Publishing the approach costs us nothing. Hiding it would cost us the ability to charge for the work.

02 · Before anything starts

Does the data exist?

Most data engagements begin with a purchase order. Ours usually begin with a question: whether the list you want can be built at all.

Some markets are thin. Some job titles do not exist in the way a client assumes they do. Some industries look substantial from the outside and turn out to contain forty companies rather than four hundred. Finding this out after you have committed budget is an expensive way to learn.

So before any engagement, we run a feasibility study. You tell us the market, the company profile, and the roles you need. We estimate how many companies and how many contacts we expect to find against those criteria.

If the numbers are there, you know what you are buying before you buy it. If they are not, we tell you, and you have saved a quarter's budget and a campaign built on a premise that was never going to work.

03 · The process

How a list gets built

Five stages, every one performed by a named team. The work branches where desk research cannot reach, and it gates before delivery: anything we cannot confirm goes back to research rather than into your file.

Why it is shaped like this. Most contact-data processes are a straight line, because they have to be: an automated source either finds a record or it does not, and there is no third option. Ours does two things a straight line cannot.

It branches. Where desk research cannot surface a contact, the work does not stop. It moves to a telephone team who call into the organisation directly, and then rejoins the flow. That branch is where the contacts come from that appear in no commercial database you already pay for.

And it gates. Nothing reaches your file until a person has confirmed it. Where a record cannot be confirmed, it is not quietly dropped; it goes back to the research team, who find out who does hold the responsibility and verify them instead. The gate is not us giving up on a contact. It is us refusing to hand you a guess. Every stage below is performed by a named team, and every stage produces something you can check.

STAGE01Requirement briefYour criteria, in writing02Web & desk researchBuilt from scratch, by people03Gap screeningWhat could not be found?04Telephone validationEvery contact, confirmed05DeliveryYour format. Yours to keepGATEGAPS FOUNDTelephone researchCalling in for what desk work missedREJOINSNOT CONFIRMEDBack to research. We findout who does hold the roleand verify them instead.PERFORMED BYAccount and research leadsWeb and desk researchersDomain specialistsTelephone researchersQuality and operationswith telephone researchers on the branchPRODUCESA WRITTEN BRIEFThe spec your deliveryis measured againstA SOURCE TRAILWhere every recordcame fromA GAP LISTWhat no onlinesource documentsCONFIRMATION + CONSENTA confirmation on every record. Whereopt-in applies, a date-stamped recordingTHE FILEIn your CRM format,guarantee attached
01

Requirement gathering

Your criteria, captured in writing.

Account and research leads · produces a written brief
02

Web & desk research

Built from scratch, by people.

Web and desk researchers · produces a source trail
03

Gap screening

What could not be found?

Domain specialists · produces a gap list
Gaps found

Telephone research

Calling in for what desk work missed. Rejoins the flow.

Telephone researchers
04

Telephone validation

Every contact confirmed.

Telephone researchers · produces confirmation and consent record
The gate

Confirmed, or back to research

Nothing reaches your file until a person has confirmed it. Where a record cannot be, it goes back to the research team, who find out who does hold the responsibility and verify them instead.

05

Delivery

Your format. Yours to keep.

Quality and operations · produces the file, guarantee attached
04 · Stage by stage

What actually happens

01

Requirement gathering

Nothing starts from a stored record, because we hold none. There is no database to filter and no product to log into. Every engagement begins with your brief.

What we need from you

The company profile or the named account list. The job titles, job roles, and job levels that matter. The geographies in scope. Anything to exclude, including your suppression file if you have one. And where relevant, the campaign the data is for, because a list built for an ABM programme looks different from a list built for a webinar invitation even when the criteria read the same.

Why specificity helps rather than hinders

There is no penalty for being unusually precise. A brief that reads "heads of RPA implementation at manufacturing firms in DACH above 500 employees, excluding our existing customer list" produces a better list than one that reads "operations leaders in Europe," and it does not cost more. Vague briefs produce vague lists, and the cost of that surfaces later, in a campaign that underperforms for reasons nobody can trace.

Where a brief is ambiguous, we ask rather than assume. That conversation is usually short and it is always cheaper than a delivery built on the wrong premise.

What this stage produces: a written brief. It becomes the specification your delivery is measured against, which matters when you check the work. Accuracy is not an abstract quality. It means this record matches this criterion in this brief.

02

Web and desk research

Our research team builds the list from scratch.

Where we look

Company and corporate websites. Third-party business sources. Professional and social networks. Company press releases and news. Financial statements and annual reports. Blogs, industry groups, and communities. Online newspapers, magazines, and the trade press. Company reports and filings. Directories, associations, and industry bodies. Paid sources, where they add something we cannot get elsewhere.

Why this is not the same as crawling those sources

A crawler reads a page and extracts what matches a pattern. A researcher reads a page and works out what it means.

The difference shows up constantly. A company announces a restructure and the head of a department you need no longer holds that title, but somebody now holds that responsibility and their new title does not contain any of your keywords. A subsidiary operates under a different trading name from its parent. A job title in one market means something entirely different in another. An organisation buries its leadership in a PDF annual report rather than a team page.

None of that is recoverable by pattern matching. All of it is routine for a person reading carefully.

What this stage produces: a source trail behind every record. We know where each one came from, which matters if you ever need to answer that question.

03

Gap screening and telephone research

The list is then screened for gaps, and this is the stage that separates our output from anything assembled at scale.

What a gap looks like

An account where the right person could not be found. A role that no online source documents. An organisation that publishes almost nothing. A market where the professional networks everyone else relies on have thin penetration. A person who does the job you need done and has no digital footprint whatsoever.

Every list has these. The difference is what happens next. A crawler-based source returns the list with the gaps in it, because there is nothing else it can do. Most of the time nobody notices, because you cannot see the contacts that are missing.

What we do instead

Gaps go to a telephone research team who work them directly, calling into the organisation to find out who holds the responsibility your brief describes. Sometimes that means the switchboard. Sometimes a department. Sometimes somebody who knows the person you need and is happy to say so.

This is not a fallback for when the desk research underperforms. It is a designed stage, and it is why our clients get contacts that appear in no commercial database they already pay for.

Where it matters most

Public sector undertakings and government bodies, which publish little and follow no standard structure. Emerging markets, where commercial coverage thins sharply. Highly specialised industries with no meaningful web presence. And privately held firms of any size in any market, which have never had a commercial reason to be visible.

What this stage produces: a list with no holes in it, including contacts that exist nowhere else.

04

Telephone validation and opt-in

Every contact is called before it reaches you. Not a sample. Not a statistically representative subset. Every one.

What we confirm

That the person still works there. That their job title and responsibilities match what our research indicated, which is not the same thing as their title being unchanged. That the work email format we hold is correct. And that the phone numbers connect, whether that is a direct dial, an extension, a switchboard, or a mobile.

Who we speak to

It depends on the organisation. Sometimes the contact directly. Often the switchboard, a colleague in the same department, or the HR or administration team. What matters is that a person at the organisation confirms it, not that we reached the contact themselves.

What happens when a contact has moved on

We do not mark the record dead and move to the next one. We find out who now holds that responsibility and research them instead. A list returned with gaps where people used to be has not done the job it was commissioned for.

Where opt-in is in scope

For clients marketing into Europe, this is also where consent is captured. Our team explains who is asking, on whose behalf, what the contact would be agreeing to receive, exactly how their data will be used, and their right to withdraw consent and how to exercise it. Every one of those elements is in the script, because consent that the person did not understand is not consent.

Permission is sought by telephone and then confirmed by email, giving a double opt-in specific to one client. Not a general marketing consent, and not drawn from a pool that was mass-opted-in for whoever buys it.

The twelve-attempt rule

We make up to twelve attempts to reach a contact for opt-in. If we still cannot speak with them, the record is marked as no answer. Not assumed. Not inferred from silence. Not quietly added to your list because they did not pick up.

Failure to opt out is not consent, and we do not treat it as consent. That standard costs us contacts on every campaign. It is the reason the ones we do deliver hold up.

What this stage produces: a researcher's confirmation against every record. For opt-in work, a date-stamped voice recording of the permission, retained for eighteen months and retrievable on request.

05

Delivery

Your list arrives in your own in-house format.

Format

NetSuite, Salesforce, Sage, or whichever CRM or marketing automation platform your team actually works in. Structured to your specification, ready to import without cleanup. Nobody on your side spends a day reformatting before the data is usable, and nobody introduces errors doing it.

Ownership

On delivery, the data is your property. Not licensed. Not leased. Not access that expires when a subscription lapses. There is no product database. Nothing is aggregated, nothing is resold, and no historical archive accumulates. Client data is retained only for the 75-day guarantee period and then deleted. Consent recordings are held for eighteen months because a consent record that cannot be produced is worthless.

What this stage produces: the file, in your format, with the guarantee attached to every record.

05 · What arrives

What arrives in the file

Not a name and a guess at an email address. Every record carries, where applicable and where it exists:

Standard delivery specificationRev. 2026.08

About the company

  • Company name
  • Total revenue
  • Employee size
  • Physical address of headquarters
  • Phone number
  • Fax number
  • Website
  • Company email address
  • NAICS and SIC codes with industry

About the contact

  • Full name
  • Job title
  • Direct dial or extension
  • Work email address
  • Physical address of site

In your format

  • NetSuite · Salesforce · Sage
  • or your own template
The numbers connect. Phone data is the weakest field in almost every commercial database, and the one your sales team discovers is wrong within an hour of receiving a file. Direct dials, extension numbers, company switchboards, and mobile numbers are all verified as working before delivery. We know they connect because we called them. That is not a claim a database assembled by crawling can make about any phone number it holds, and it is the reason a KAT-i list behaves differently the first week your team works it.
06 · The guarantee

What we guarantee, and what happens when we miss

Accurate on delivery

Every contact is telephone-verified before it reaches you, and guaranteed accurate and relevant to the criteria in your brief. Not a percentage with an acceptable margin of error. If a record does not match the brief, it was not delivered correctly, and we correct it.

Covered for 75 days

People change jobs, and sometimes they do it the week after we deliver. If a contact we supplied changes role within 75 days, or an email address or phone number we provided stops working in that window, we replace the contact or credit it back to you. The risk of ordinary decay in that period sits with us, not with you.

Yours to keep, permanently

On delivery the data becomes your property, and stays yours regardless of whether you work with us again. There is no product database. Nothing is aggregated, nothing is resold, and no historical archive accumulates. Client data is retained only for the 75-day guarantee period and then deleted.

One thing we will tell you that most providers won't. Owning the data forever does not make it accurate forever. B2B contact data decays continuously, whatever its source and whoever built it. After a period, records need re-verifying, and contacts opted in for one campaign will eventually need opting in again.

Any provider implying otherwise is either not being straight with you or does not understand their own product. We tell you when a list needs refreshing rather than let you run a campaign against data that quietly stopped being true.

07 · Timelines

How long it takes

We quote a timeline against your specific brief before work begins, and we do not quote a turnaround we cannot hold. What drives the number is worth understanding, because it explains why two engagements of the same size can differ significantly.

Volume matters least

Once a research method is established for a market, scale is largely a question of capacity.

Difficulty matters most

A few hundred contacts at well-documented enterprises in North America moves quickly. The same volume across privately held manufacturers in emerging markets, or inside a sector with no commercial data at all, takes considerably longer, because each record requires primary research rather than desk work.

Depth matters

Mapping twenty stakeholders inside one organisation is slower per account than finding one contact at twenty organisations, even at identical record counts.

Opt-in adds time by design

Twelve attempts per contact, in the contact's own business hours, cannot be compressed without abandoning the standard that makes the consent worth having.

Region affects working hours

Verification happens in the contact's business hours, not ours. A global list moves at the pace of its slowest time zone.

If a deadline is fixed, tell us at the briefing stage. We would rather scope to what is achievable than deliver late or, worse, deliver on time by skipping the verification that the whole engagement depends on.

08 · Briefing

Write a brief that gets you a better list

The quality of a delivered list correlates closely with the quality of the brief behind it. A few things help more than people expect.

Name the campaign, not just the criteria

Knowing that a list is for a competitive displacement campaign rather than an event invitation changes which contacts inside an account are actually useful. Same criteria, different right answer.

Be specific about seniority, not just title

Titles vary enormously between organisations and between markets. "The person who owns the budget for this" and "the person who evaluates it technically" are more useful instructions than a list of titles, and we can work to both.

Send your suppression file

It costs you nothing and guarantees everything delivered is new. It also stops you paying for records you already have.

Tell us what a good record looks like to you

Some teams need direct dials above all else. Some need verified emails and nothing more. Some need site addresses for field sales routing. We can prioritise, but only if we know.

Flag the exclusions

Current customers, current opportunities, markets you cannot service, competitors, companies below a size threshold. Every exclusion you supply upfront is a record you do not pay for and do not have to clean out later.

Tell us if you are unsure the data exists

That is what the feasibility study is for, and asking early is much cheaper than finding out after.

09 · Limits

Limits build more trust than claims

We don't sell off-the-shelf lists

There is no catalogue and no database to buy from. If you need ten thousand generic contacts by Friday, a subscription platform will serve you better and we will say so.

We don't do consumer data

No individual consumer contact information, and no sensitive personal data. Business contact information only.

We don't guarantee responses

We guarantee that a contact is accurate, relevant to your brief, and reachable. Whether they reply depends on your offer and your messaging, which are not ours to control.

We don't treat silence as consent

Twelve attempts, then the record is marked no answer. It does not quietly enter your list.

We don't build lists that cannot be built

If the feasibility study says the universe is not there, we tell you rather than delivering something padded to hit a number.

We don't keep your data

There is no product database. Nothing is aggregated, nothing is resold, and no historical archive accumulates. Client data is retained only for the 75-day guarantee period and then deleted.

We don't quote turnarounds we cannot hold

If a deadline requires skipping verification, we will say so rather than agreeing and cutting corners you cannot see.

10 · Questions

Questions about the process

Do you really call every contact, or a sample?

Every contact. Sampling tells you what a batch looks like on average. It does not tell you whether the specific record your SDR is about to dial is correct.

What if you can't reach anyone at the organisation?

We work through the alternatives: switchboard, department, colleagues, administration. If the contact has moved on, we find out who now holds that responsibility and research them instead, so the account is not lost. Only where nobody at the organisation matches your criteria does a record not ship, and in that case we tell you which accounts came up empty rather than padding the file to hit a number.

Can we see the source for a given record?

We maintain a source trail through the research process. If you need provenance for specific records, particularly for a compliance review, ask and we will discuss what we can provide.

Do you use paid data sources?

Where they add something our own research cannot, yes, as one input among many. Nothing is delivered on the strength of a purchased record alone. Everything is verified by telephone regardless of where it originated.

How do you handle contacts who have moved companies?

During verification we identify who now holds the responsibility and research them instead. Where a contact has moved and the client is interested in where they went, that mapping is something we do as a separate service.

What happens if we dispute a record?

Tell us which record and why. If it does not match the brief or the contact details do not work, we replace it or credit it. We do not require you to prove it beyond identifying the issue.

Can you work inside our existing process?

Yes. Suppression files, your own data template, your naming conventions, your field structure, your delivery cadence. The output is designed to arrive in your systems, not ours.

Who actually does the work?

Our own teams. Web and desk researchers, domain specialists, and telephone researchers. Nothing is outsourced to a third-party data broker.

Where is our data stored?

Data is processed on secure Microsoft, Box.com, and Google Workspace business systems during the research and verification period. Once the list is completed, telephone-verified, and delivered to the client, our copy is deleted from active systems. We do not retain client data beyond the delivery and guarantee period. There is no product database. Nothing is aggregated, nothing is resold, and no historical archive accumulates.

Can we test the process before committing?

Two ways. A feasibility study estimates what exists against your criteria before you spend anything. Or send a requirement brief and we will run a trial and deliver sample data for you to check for accuracy, relevance, and fit with your systems.

Two ways to start

Tell us the market you want to reach and we will estimate what exists against your criteria. Or send a brief and we will build real sample data for you to check.