Sales intelligence tools: 5 layers, 5 different leaks, one you actually have

Sales intelligence tools: the 5 layers sold as one category

Sales intelligence tools are compared as if they were one product, and they are 5 different ones. A scored comparison that ranks them on a single total hides the only question that matters: which of the 5 is missing from your pipeline right now.

1. Contact data. Whether this person is reachable today, at a work email or on a mobile.
2. Account data. Who this company is: size, industry, location, structure, who works there.
3. Technographics. What they run, which is how you know whether your product fits before you write.
4. Buying signals. Whether this account is in market now: a job change, a funding round, a hiring sprint, a stack move.
5. Conversation and revenue intelligence. What happened inside the calls you already had, and what it predicts.

Those 5 fix 5 different leaks. A team that cannot reach anyone has a layer 1 problem and will get nothing from an intent platform at 50,000 dollars a year, a list price observed in September 2026. A team with excellent contact data and no close rate has a layer 5 problem and buying more contacts makes it worse. Buying the bundle because it is one purchase order is how a stack ends up costing more than the pipeline it produces. If what you are really shopping for is the enrichment layer alone, our comparison of 51 lead enrichment tools by job takes that apart tool by tool. This page sits inside our wider B2B data providers guide.

Sales intelligence is the information a seller has about an account before the first contact and during the deal. A sales intelligence platform, sometimes sold as sales intelligence software, is what supplies it. The words are used interchangeably by vendors; what changes between them is which of the 5 layers below they actually own.

Which layer you are missing, before you compare sales intelligence tools

The diagnosis takes 4 numbers you already have. Where your funnel loses the most is the layer to buy, and everything else can wait a quarter.

Run it on your last full quarter rather than on this month. One bad month moves every number, and the layer you buy on a bad month is rarely the one you needed.

If the diagnosis points at layer 1, the fix is measurable before it is bought: resolve 100 of your own contacts and read the hit rate. Run that test on your own accounts.

Layer 1: contact data, the only one you can verify before you buy

Every platform in this category publishes a contact count, from 200 million to 320 million and up. Those numbers are self-declared, they carry no verification date, and they are not counting the same thing. More importantly they are not testable: you cannot check 320 million records, and no vendor will let you.

What is testable is your own coverage. Take 100 real people from your own target accounts, not a sample the vendor supplies, and count how many come back with a deliverable address. That percentage is the only figure in this whole category that you produce yourself, and it takes an afternoon. It is also the argument for building your own contact database rather than renting one: a list you resolved is a list whose coverage you know.

This is the layer Derrick is built for. Email Finder resolves a work address at 5 credits per email and Phone Finder a mobile at 150 credits per phone, both billed only when a result comes back, so a person who cannot be resolved costs nothing. That billing shape is the concrete difference with a subscription to an index: you pay for what resolved, not for what the index claims to hold. Both sit on the paid plans, from 9 EUR a month, and the free plan and its 100 credits per month cover the import and enrichment side. Measure your coverage on your own accounts.

Layers 2 and 3: account data and technographics

Account data is what turns a list of names into a list of accounts you can qualify: headcount, industry, country, structure, and who actually works there. Technographics is the layer above it, and it is the one that filters hardest, because what a company runs tells you whether your product has a place before you have written a word.

Both are cheap relative to what they save. Enrich Companies fills the firmographics at 1 credit per company, Find Similar Companies turns one good account into a matched list at 1 credit per company, Find a company's people lists current and former staff at 1 credit per person without needing a Sales Navigator seat, and Website Technologies detects the stack at 2 credits per website. Our page on choosing an enrichment provider sets out the 8 criteria to run those against.

The mistake at this layer is buying breadth. A database that covers every industry on earth is worth less to you than one that covers your 3 segments properly, and the coverage figure that matters is again the one you measure on your own accounts rather than the one on the pricing page. Our geographic coverage page explains why match rates move so much between markets.

Layer 4: buying signals, and the timing problem

Intent is the layer with the widest gap between what it promises and what it delivers, because most of it is inferred rather than observed. A signal that says an account read 3 articles about your category is a probability. A signal that says their VP of Sales changed jobs last week, or that they just closed a funding round, or that they are hiring 12 SDRs, is an event. Events are checkable. Inferences are not.

The practical distinction is worth holding onto when you compare platforms: ask which signals are observed events and which are modelled scores, and ask what the recency is. A signal that arrives 6 weeks after the event is a report, not a trigger.

Derrick covers the observed side. Signal tracks leads and accounts for job changes, funding rounds, hiring sprees and stack moves and fires the alert when it happens, from 20 EUR a month, and it bills 1 credit per signal that actually fires rather than per account monitored. Company Hiring Signal shows who is hiring and for which roles at 1 credit per company, and Google News Scraper returns the most relevant recent article per company at 1 credit per news item. For modelled, aggregated intent across a whole category, the platforms built for that do a job Derrick does not: 6sense, Demandbase and Bombora sit on that layer, and it is a genuinely different product from event detection. The provider benchmark shows how far the same gap between claim and measurement runs across the whole category.

Layer 5: conversation and revenue intelligence

This layer is not a data purchase at all, which is why it sits oddly inside comparisons of sales intelligence tools. Conversation intelligence records and analyses the calls you already had; revenue intelligence forecasts from your own pipeline. Gong is the reference product here, and nothing Derrick does overlaps with it: one looks outward at accounts you have not met, the other looks inward at conversations you already had.

It is worth naming because of what it implies for a budget. If your reply rate is healthy and your close rate is not, no amount of contact data or intent fixes that, and every euro spent on layers 1 to 4 is spent on the wrong problem. That is the single most common misallocation in this category, and the scored comparisons cannot catch it because they only rank what they sell.

What sales intelligence tools cost, and the number nobody publishes

Entry prices in this category run from a few hundred dollars per user per year to well over 50,000 dollars a year for the account-intelligence platforms, and a full stack is commonly quoted between 41,000 and 173,000 dollars a year. Those are list figures observed in September 2026 and they are not the number that decides anything.

The number that decides is cost per usable record, and it is missing from every pricing page because it depends on your segment rather than on the vendor. It has 2 shapes, and they behave in opposite directions.

Work out both on the same 100 accounts before you sign. On a segment where a vendor has thin coverage, a subscription that looks cheap on the pricing page can be several times the cost per usable record of a pay-per-result tool, and nothing on either pricing page will tell you that. Our pricing comparison works through the arithmetic.

The vendor scores that are not benchmarks

Comparisons in this category increasingly publish a scoring framework: so many features across so many categories, each platform rated out of a total. It looks like a benchmark and it is not one, for a reason that is easy to check and rarely mentioned. The scoring is usually published by one of the platforms being scored, and that platform usually wins.

The features chosen are the axis. A vendor strong on multichannel engagement weights multichannel engagement; a vendor strong on intent weights intent. None of this is dishonest, and none of it is transferable to your decision either. 3 questions turn a score back into something usable.

• Who published it, and are they in the ranking? If yes, read the category weights rather than the totals.
• Is any line measurable by you? Coverage on your accounts, bounce rate on your sends, signal recency on accounts you know. Those you can reproduce.
• What is not scored? Export rights, contract length, what happens to your data when you leave. These never appear in a feature matrix and they cost the most later.

How to choose sales intelligence tools without buying the whole stack

The layered view produces a different buying order than a scored comparison does, and it is usually cheaper by a factor.

On the modelled-intent layer and on conversation intelligence, those platforms do work Derrick does not do at all. If your diagnosis points at one of those two layers, that is where your budget belongs, and no amount of contact data will substitute for it. On the data layer, layers 1 to 3, Derrick does better, and the reason is concrete rather than a claim about quality: it charges for records that resolved instead of for access to an index, it runs on the list you already hold rather than on one you rent, and it works at every scale, from a 40 row test to a nightly API job. Start where your diagnosis pointed, measure the layer for one quarter, and add the next one only when the first has stopped being the constraint. If that first layer is the data one, the cheapest way to find out is to run it on your own accounts rather than to read another comparison. Test the data layer on your own list.

Key takeaways

• Sales intelligence tools bundle 5 layers that fix 5 different leaks. Buying the bundle buys 4 layers you may not need.
• Published contact counts, from 200 to 320 million and up, are self-declared and untestable. Coverage on 100 of your own accounts is not.
• Observed events beat modelled intent for triggering outreach, and recency matters more than the score.
• A poor close rate is a layer 5 problem. More contact data makes it worse, not better.
• Vendor scoring frameworks are usually published by a platform inside the ranking. Read the weights, not the totals.
• Cost per usable record is the number that decides, it is on no pricing page, and it takes 100 accounts to produce.

Derrick covers the data layers: contact resolution, company data, staff, technographics, and observed buying signals, from the Google Sheets sidebar, an AI assistant over MCP or the REST API on the Standard plan at 20 EUR a month, with a web app arriving. Email Finder and Phone Finder bill only when a result comes back, so a record that does not exist costs nothing. They sit on the paid plans, from 9 EUR a month, which is also what the coverage test needs, since resolving 100 addresses spends 500 credits. The free plan and its 100 credits per month cover the import and enrichment side, enough to build and qualify the test list before you spend a credit on resolution. Start with the free plan, or read the rest of the B2B data providers guide.

We publish one email every 2 weeks with what we measure on B2B data coverage, decay and cost. Subscribe from the homepage.