What Is Sales Intelligence? Data Types, Use Cases and How to Judge the Data Behind It

Sales intelligence is the data a B2B sales team gathers about companies, the people in them and what those companies are doing right now. It tells a rep who to contact and why. IBM's definition is shorter: "the collection of data to improve sales performance and decision-making". Salesforce adds the insights built on that data, used across the sales process.

The reason it matters is timing. By the time a buyer first contacts a vendor, 61% of the buying journey is already over. The vendor they favoured before that call wins about 80% of the time. Sales intelligence is how you get into the first 61%.

This guide covers the six data types, the four steps that turn raw sources into a list, and four yardsticks for judging a vendor's data. The yardsticks come with Spona's own measurements: fill rates on 10,000 company records, verification pass rates and what changed in one quarter.

Get thousands of leads in 4 minutes.

Stop stitching tools together.
Start with verified sales data, delivered in minutes.

SCALE SALES NOW

No contracts · No setup · No onboarding

Sales intelligence at a glance: six data types in one table

Sources disagree on the count. Salesforce lists five types, IBM seven, and TechnologyAdvice six, adding engagement data. Six is the useful number, because each type below answers a different question a rep asks before a call.

Data type	Question it answers	Example fields	Where it comes from
Firmographic	What kind of company is this?	Industry, headcount band, revenue band, country, founding year	Company registers, websites, LinkedIn pages, filings
Contact	Who do I talk to?	Name, title, seniority, department, work email, phone	Professional profiles, team pages, email verification
Technographic	What software do they run?	CRM, CMS, marketing automation, payment stack	Website code scans, job posts, review sites
Intent	Are they researching this now?	Topic surges, content consumed, review-site visits	Publisher co-ops, review sites, your own site
Signals	What just changed?	Funding round, hiring, new location, first seen, leadership change	Filings, press, job boards, registers
Engagement and deal	How has this account dealt with us?	Emails opened, meetings held, stage, past deals	Your CRM, email and calendar

The first five describe accounts you do not know yet. The sixth comes from your own systems, which is where sales data and sales intelligence part ways.

A sales intelligence database is the store that holds these records. Each row is one company or person, ideally with a source and a last-checked stamp on every field. Those two stamps matter more than the row count, as the yardsticks below show.

TechnologyAdvice adds one warning worth keeping. Intent data is not a guaranteed buying signal. A research spike says someone at the company is reading, not that the company is buying.

Why B2B sales intelligence matters before the first call

B2B sales intelligence exists because the buyer does most of the work without you. Gartner finds that B2B buyers spend only 17% of their buying time meeting with potential suppliers. When they compare several vendors, any one rep may get 5% or 6% of that time.

Gartner puts the typical buying group for a complex B2B solution at six to 10 decision makers. And 77% of buyers call their latest purchase very complex or difficult.

6sense surveyed nearly 4,000 buyers, a fifth of them in the UK and Ireland, and measured the same shift from the buyer's side. First contact with a vendor has moved from 69% of the journey to 61%. Buyers initiated 79% of those engagements. And the vendor they favoured before contact wins the deal roughly 80% of the time.

So the call that matters happens before the buyer calls you. To make it, a rep needs the company, the person, a reason and a working email. Finding those by hand is where selling time goes. Salesforce's State of Sales puts the average seller at 40% of their time selling, and 48% say they lack the bandwidth for adequate cold outreach.

That survey covered 4,050 sales professionals in 22 countries, including the US and the UK.

How sales intelligence works: collect, match, verify, deliver

Every sales intelligence platform runs the same loop, vendor or spreadsheet. The difference is how well each step gets checked.

Sales intelligence loop in four steps, collect, match, verify and deliver, with the check each step adds to a record

1. Collect: pull records from public and licensed sources

The raw material is public or licensed: company registers, websites, job boards, press releases, filings, professional profiles and technology scans. BuiltWith alone tracks 127,000+ internet technologies across the web. Spona's own process mines 40+ sources per search.

What a good collector keeps is the source URL for every value. Without it, steps 3 and 4 have nothing to check against.

2. Match: one record per company, one per person

Sources disagree. One lists "Acme Ltd", another "ACME Limited", a third only the domain. Matching resolves them to one company record, usually keyed on the domain, and one record per person keyed on the work email. Duplicates go before anything gets counted.

The output is a deduplicated set, which is the only base you can measure fill rates on. Spona's fill rates below come from 10,000 deduplicated company records for that reason.

3. Verify: test every email before it ships

A collected email is a guess until a verifier tests it: syntax, domain, mail server and a mailbox check. The industry's own numbers show why this step exists.

In one year, ZeroBounce processed 11 billion+ addresses. It found only 62% valid. More than 9% sat on catch-all domains, which accept every address and confirm none.

Verification is not a one-time job. ZeroBounce also finds that at least 23% of an email list decays within a year. A vendor should state when it last checked each email, and recheck before a send. Our guide to B2B data validation covers the checks field by field.

4. Deliver: put the record where the rep works

Actionable sales intelligence is a record that arrives where the rep works, with a reason to call attached. Delivery comes as a CSV, an API response or a CRM sync. The reason to call is the signal from step 1: the funding round, the hire, the new location.

Automated sales intelligence means this loop runs on a schedule, not when a rep remembers. Spona's cadence, for example, is weekly signal ingestion, a recheck after 30 days and a stale flag after 90.

How AI improves sales intelligence

AI does three jobs in this loop. It reads unstructured sources, so a job post becomes a hiring signal with a date and a press release becomes a funding round. It matches records that spell the same company three ways. And it turns a plain description of your buyer into filters, which is how a chat-based search works.

Adoption is already mainstream: 55% of sales professionals use AI for prospecting. What AI cannot do is make an unverified email valid. The verify step stays.

Sales intelligence vs CRM, business intelligence and conversation intelligence

"Sales intelligence CRM" is a common search, and the answer is two different things that work together. Sales intelligence describes accounts you do not know yet. A CRM records what you have done with accounts you do know. The table sets five systems that reps often confuse side by side.

System	What it holds	Question it answers	Where the data comes from
Sales intelligence	External data on companies, people, technologies and signals	Who should we contact, and why now?	Registers, websites, job boards, scans, verification
CRM	Your accounts, contacts, activities and deals	What have we done with this account, and what is next?	Your reps, forms, email and calendar sync
Business intelligence (sales analytics)	Aggregated metrics: pipeline, win rate, cycle length, quota	How is the team performing, and where does it leak?	CRM, billing and marketing systems
Conversation intelligence	Call and meeting recordings, transcripts, talk ratios, objections	What happened on the call, and what should we coach?	Dialer and video meeting tools
Market intelligence	Market size, competitors, pricing trends	Which markets should we enter?	Analyst reports, surveys, filings

The CRM row is where sales intelligence earns its keep. Validity surveyed 631 CRM admins in the US, UK and Australia. 24% said less than half of their CRM data is accurate and complete. 31% said poor data costs at least 20% of annual revenue. And 48% had seen decay accelerate.

Salesforce's State of Sales adds the AI angle: 51% of sales leaders with AI say disconnected systems slow their AI initiatives. And 74% of sales professionals are focusing on data cleansing. A sales intelligence feed into the CRM is one answer to both: it fills blanks and refreshes stale values on a schedule. Our data hygiene post covers the merge rules.

Sales intelligence use cases, ranked by 1,000 Spona searches

The best evidence of what sales intelligence is used for is what teams search for. Spona classified 1,000 of its searches by the filters each one led with, one class per search, so the shares add to 100%.

Ranked bars of 1,000 Spona searches by filter combination: industry, headcount, country and seniority lead at 22%

The plainest use case leads. 22% of searches combined industry, headcount, country and seniority: a classic ideal customer profile with a role on top. Another 16% started from a role and a country and asked for a verified email. Together the two firmographic-plus-role patterns made 38% of searches.

Signal-led searches came next. Technology, funding and hiring filters led 27% of searches between them. Local-business searches, a category plus a city plus the owner, made 10%.

The Shopify row, at 8%, is a request, not a product: Spona has announced Shopify data and does not sell it yet. And one search in 20 began by excluding what was already in the CRM, which is sales intelligence used to find new accounts.

Six signals with real counts

Spona defines a signal as "a dated event Spona ingests and lets customers filter on". Six work today, each shown here with one real search and what it returned.

Signal	Example search	What came back
Funding	US software firms funded in the last 90 days	1,200 companies, 4,800 verified founder or finance contacts
Hiring	US manufacturers with 1 to 49 staff hiring in the last 30 days	28,400 companies
Technology	UK HubSpot users with 20 to 200 staff	5,600 companies, 14,200 verified sales or marketing contacts
Local business	Dental practices in the Chicago metro area	4,100 locations, 3,200 companies, 3,660 verified owner or office-manager contacts
New business	US construction companies with 1 to 49 staff first seen in the last 90 days	18,400 companies
New location	US restaurant companies with 1 to 49 staff and a location first seen in the last 90 days	14,600 companies

The outside world explains the counts. North American startups raised $92 billion in a recent quarter, $61 billion of it for AI companies. Europe raised $25 billion in the same quarter, and the UK led with $7.5 billion. Each round is a dated event with a named company behind it; our guide to recently funded startups shows how to work them.

Hiring moves faster. US employers made 5.2 million hires in one recent month, ending it with 7.1 million job openings.

New business filings arrive in bulk. The US logged 534,947 business applications in one month, seasonally adjusted. The UK register added 815,277 incorporations in the latest financial year, against 787,120 dissolutions. A first-seen signal turns that flow into a list of companies still choosing their suppliers.

How to judge sales intelligence data: four yardsticks

Ranking guides call good sales intelligence data "accurate and current" and stop there. The UK government's data quality framework gives you something to measure: six dimensions, accuracy, completeness, uniqueness, consistency, timeliness and validity. Three of them, plus provenance, turn into questions a buyer can ask a vendor, with Spona's own numbers as a reference point.

A fill rate is the share of records with a value in a field. Spona measured it on 10,000 deduplicated company records, at the company level only. Contact-level fill rates are not part of this study.

Field group	Fill rate per field
Identity	Company name 100%, country 99%, domain 97%, city 92%
Profile	Industry 94%, business description 93%, employee band 90%, company LinkedIn URL 88%, founding year 72%, revenue band 62%
Technology and funding	Technology evidence 38%, latest funding round 8%, funding date 8%, investor names 7%

Read the low rows with their denominators. The funding fields use all companies as the denominator, not only funded companies, so 8% means roughly one company in twelve has a known round. Revenue and headcount may be estimates and need their own source flags.

The pass mark is a rate for every field you filter on, not one global number. The vendor question: "What is your fill rate per field on a sample of my ICP?"

Validity: how many emails pass verification?

Validity is whether a value works: an email that accepts mail, a phone that rings. Spona checked 10,000 unique candidate addresses per country and sent to 5,000 verified ones.

Panel	Pass	Catch-all	Other rejected	Hard bounce on sends
US (10,000 checked, 5,000 sent)	78.0%	14.0%	8.0%	1.5% (75 bounces)
UK (10,000 checked, 5,000 sent)	80.0%	12.0%	8.0%	1.2% (60 bounces)

Pass, catch-all and other rejected share one denominator, all unique addresses checked, and the three categories do not overlap. Other rejected covers invalid, unknown, disposable and suppressed addresses. Hard bounce means permanent delivery failures divided by attempted sends to the verified subset, excluding catch-all addresses. Inbox placement is a separate metric.

Spona's recommended operating target is hard bounces below 2% on a freshly checked list, with a recheck when a send exceeds it. Role and vertical splits remain unmeasured. The vendor question: "What hard bounce rate did your last verified list produce, and on how many sends?"

Timeliness: how much changes in one quarter?

People move. The median US worker has been with their employer 4.1 years. For workers between their mid-twenties and mid-thirties, it is 3.0 years. Every move changes a title, often an employer, and usually an email.

Spona measures this on a stable panel of 10,000 records per country, compared at the start and end of one quarter. In the US panel, 8.0% changed title or employer. Separately, 6.5% had an email that changed or became invalid. The UK panel showed 7.0% and 6.0%.

US and UK panels of 10,000 records: 8.0% and 7.0% changed title or employer in one quarter, email changes drawn apart

The two measures can overlap, so Spona reports them separately and never adds them. One quarter is also not a year: do not multiply by four. And the panel does not establish how often Spona contacts change jobs. It shows what a buyer should expect to re-verify.

The cadence that follows: verify every email before delivery, and again before a send if the check is over 30 days old. Refresh high-activity contacts every 30 days and the wider inventory every 90.

Our data decay guide covers the causes. The vendor question: "When did you last check each record, and is that a field I can filter on?"

Provenance and lawful use: where did the record come from?

Provenance is not one of the six dimensions, but it is how you prove them. It means a source URL and a last-checked timestamp on every record. Spona's own rule is that both should be mandatory fields, and its API returns them. Without them you cannot audit a record, answer a complaint or show where an address came from.

Lawful use has two rule sets. In the US, the FTC is blunt: CAN-SPAM makes no exception for business-to-business email. Each separate email in violation can cost up to $53,088, and you must honor opt-outs within 10 business days.

In the UK, the ICO says the PECR rule on marketing email does not apply to corporate subscribers. Sole traders and some partnerships count as individuals, so they need consent or the soft opt-in. Using a business contact's personal data for direct marketing still falls under the UK GDPR. Live B2B calls need screening against the TPS and the CTPS.

Our guide to data vendors has the longer due-diligence list. The vendor question: "Does every record carry a source URL, and can you show me the lawful basis for the UK contacts?"

Where Spona fits in sales intelligence, and where it does not

Spona is a sales intelligence tool built around one workflow. You describe your ideal customer in a chat. The AI mines 40+ sources, verifies every lead and delivers a scored list in about 4 minutes. You pay per lead, in credits, and every search starts with a free sample.

A lead costs $0.008 to $2.00, depending on the search type and the plan. Bulk Search runs $0.008 to $0.09 a lead for broad email or LinkedIn outreach. Standard Search runs $0.024 to $0.55, and the top of that range buys a mobile number. Deep Search runs $0.48 to $2.00 for leads scored, verified and enriched in real time.

Each new search session costs a one-time 1,000 credits. That is $8 to $10, by plan, and the pricing page lists the rates.

The Free plan carries 3,000 one-time credits. Pro costs $99 a month and Premium $499. Expert costs $999 a month, and Enterprise starts at $20,000 a year. The signal filters in the table above, funding, hiring, technology, first-seen and new-location, work on Pro and above.

The API runs on Expert and returns JSON, with a source URL and a last-checked field on every record. CSV export works on every paid plan, and Clay takes Spona data by CSV or API. HubSpot is the only native CRM integration. You can also bring your own list and pay only for what Spona adds: company data for 3 credits, a verified email for 5.

Where it does not fit. Spona covers four of the six data types: firmographic, contact, technographic and signals. Intent data and conversation intelligence are not on its feature list, and engagement data stays in your CRM. Teams on Salesforce or Pipedrive load Spona by CSV or API; neither has a native connector, and the same holds for Smartlead and Instantly.

The gaps in the data show in the yardsticks. Funding fields are filled for 8% of companies or fewer, and the fill rates cover company records only. The pass rates and the change panel are country-level only, with no role or vertical splits. There is no building-size filter for local-business searches.

Signal Search as its own search type, AI cold email, AI cold calling and the dialer are announced, not available. For a provider-by-provider view, our best B2B data providers post compares 12 of them by data type and price.