Clevenio
Clevenio, trading as Sales Led Oy and founded in 2023 in Jyvaskyla, Finland, sells a proposition the large US data platforms cannot: the whole market rather than a sample of it. Its argument is that list building is a symptom of incomplete market visibility, and that a finite market can simply be mapped. The company builds from each country's official business register rather than from scraped or user contributed lists, then uses large language models to cross reference those registers against public data across the web and attach the right decision makers to the right companies, refreshing continuously so a newly registered company reaches the customer's CRM without anyone noticing it.
Two products carry it: a Finnish database drawing on the full trade register plus telecom operator data for phone numbers matched to named individuals, financial statements, industry codes, headcount and group structure; and a European database covering the Nordics, Baltics, DACH, Benelux, Poland and the UK from each local register plus professional networks and the open web.
Around the data sit seven features: buying signal triggers, CRM sync into HubSpot, Pipedrive and Salesforce, an audience builder, multichannel sequences, a lightweight CRM built for outbound, a REST API, and an MCP server that is live for every customer and marketed as the first thing on the homepage. Around 500 B2B sales teams across the Nordics, sold to CSOs, reps, marketing and RevOps.
Capability Axes
Capability grades
17 of 17 axes rated · 6 graded A or B
A genuine middle case, which is rare on this axis, and the vendor's own method statement is what puts it there: public business registers give the foundation, and large language models cross reference them with public data across the web to match the right decision makers to the right companies, monitored continuously.
The models are therefore load bearing for the differentiator, because a register extract without decision makers attached is a free download from PRH or Bolagsverket, and the decision maker layer is what customers buy. Against that, roughly half the platform is mechanical and would survive intact: sequences, the lightweight CRM, CRM sync, buying signal triggers, the REST API and the audience builder are all conventional software, and the register foundation has independent value. Not A because a saleable product remains after removal; not C because what remains is not the product the vendor sells.
The autonomy here is unusual because it points at the customer's own system of record rather than at a prospect. The vendor states the design goal plainly: nobody has to notice a new company for it to end up in your CRM. Model matched decision maker records are pushed automatically into HubSpot, Pipedrive or Salesforce and refreshed continuously as companies and roles change, so a machine judgement about who works where becomes a CRM record without a human confirming it.
Nothing published describes a review step, a confidence threshold below which a match is withheld, or an audit trail of what was written and on what basis. Sequences and triggers add conventional send side automation with no described approval. The mitigating factor is scope: the product does not compose messages autonomously or handle replies.
The mechanism is published, which no other data vendor in this index does. Clevenio states on its homepage that public business registers give the foundation and that large language models cross reference them against public data across the web to match decision makers to companies, with the whole dataset monitored continuously.
That is a data vendor telling buyers that its contact records are model produced rather than verified by a researcher, and it is the kind of admission most vendors in this category work hard to avoid making. Not A because the disclosure stops at the method: no model provider, family or version is named, nothing states where inference runs, and critically no accuracy or error rate is published for the matching itself, which is the single number that matters here, since a mismatched decision maker means a rep phones the wrong person at the right company and the buyer has no way to size that risk.
Top of band on attribution and short on substantiation. Twelve customer logos are named organisations rather than anonymous images, four testimonials carry full names, job titles and companies including Fondion, Villi.io, HeiaHeia and Right People Group Spain, a case studies library and a wall of love exist, and roughly 500 B2B sales teams is a specific and plausible scale claim for a three year old company.
What is missing is any quantified outcome with a method: the headline data claims of complete coverage of the Nordic markets, high email deliverability and decision maker match rates carry no published methodology, and the cost of inaction argument on the homepage rests on unsourced category statistics, that reps spend 70 percent of their week on non selling work and that the fix returns around two selling hours per rep per day. Third party reviewers also note the absence of independent reviews on the major platforms, so there is little outside check on any of it.
Recorded as observed rather than concluded: multichannel sequences run across email, phone and other channels, and nothing located on the vendor's own site sets out a consent basis, suppression mechanism, opt out route or do not call handling for the outreach the product performs.
The gap is more consequential for this vendor than for a US equivalent, because a Finnish controller selling data on named individuals across the Nordics, Baltics, DACH, Benelux, Poland and the UK is operating under national supervisory authorities that publish their own and not always consistent guidance on legitimate interest for B2B contact, and under ePrivacy rules on electronic marketing. Affiliated content elsewhere does work through legitimate interest reasoning for B2B outreach in some detail, but the ownership of that property could not be confirmed and it is not relied on here.
Recorded as observed rather than concluded: a privacy policy, cookie policy and terms are published under a clearly identified Finnish legal entity, Sales Led Oy, but the policy bodies were not read this pass, so no controller designation, lawful basis, retention position or data subject rights process has been verified. The specific question this product creates is the sharpest privacy question of any EU domiciled vendor in the index and it deserves naming precisely.
Clevenio holds direct phone numbers and email addresses for named decision makers across at least ten countries, assembled by model inference from public web data and, in Finland, from telecom operator data matched to the individual. Article 14 of the GDPR applies squarely when personal data is obtained from sources other than the data subject, and no notification, no self service lookup and no removal route was located anywhere on the site. That is the piece Apollo publishes and which this vendor, operating inside the jurisdiction that wrote the rule, does not appear to.
Top of band, and on sourcing alone this is the most authoritative record in the index. The company layer comes from each country's official national business register rather than from scraping or user contribution, which means registration numbers, financial statements, industry codes, headcount, ownership and group structure are government sourced and independently checkable, and the vendor names this as its foundation rather than burying it.
The person layer is disclosed separately and honestly: matched by large language models from public data across the web and, in the European product, from professional networks, which is as close as anyone here comes to naming LinkedIn as a source. The Finnish product adds a genuinely unusual disclosed route, phone numbers drawn from telecom operator data and matched to the named person. Publishing three distinct source classes and the method that joins them is exemplary. Not A on the same ground that holds Privacy down: Apollo earns the A partly by publishing an actual Article 14 discharge and a self service removal path, and neither was located here.
Low exposure and mostly by construction. The primary data source is official public registers, which are published for exactly this kind of use, so the bulk of the dataset carries no platform dependency at all. Integrations are official CRM connectors into HubSpot, Pipedrive and Salesforce, and the developer surface is a documented REST API plus an MCP server the vendor operates itself.
There is no browser extension, no credential custody and no automation performed against a third party platform on the buyer's behalf. Not A because the person layer is described as drawn from professional networks and the open web, which is the contested surface in this category, and no conformance position is stated on it anywhere.
Recorded as observed rather than concluded: nothing located addresses model training, tenant separation, retention or what happens to customer CRM data, sequence content and reply text that passes through the platform. One structural point is worth recording because it inverts the usual concern.
The models here operate on public register and web data to build a database that is shared by every customer by design, so the cross tenant question that dominates elsewhere in this index barely applies to the dataset itself; nobody is surprised that another customer can see the same company. What is unaddressed is the other direction, whether anything a customer does inside the platform feeds the models or the shared dataset.
No impersonation and no synthetic identity anywhere in the product: sequences are composed by and sent from real reps, and the company makes no claim to autonomous agents contacting anyone. What sits behind it is quieter and worth naming. The person who receives the call is in the database because a language model matched them to a company from public web data, and in Finland their direct number may have come from telecom operator data.
They did not provide it, were not notified, and have no visible route to see or remove the record. No Article 50 position exists and none is really needed, since the AI here shapes who gets contacted rather than what is said to them, which is a different and less examined shape of the same problem.
Top of band on agent readiness and narrow on everything else. The MCP server is live for every customer, included in both plans rather than sold as an upgrade, and carried as the first banner on the homepage, naming ChatGPT and Claude explicitly so a buyer can query their own mapped market from inside an AI tool.
That is the third MCP server graded in this index after ActiveCampaign and Clari, and by some distance the most prominently positioned; a three year old Finnish company is ahead of most of the corpus on the surface that will matter next. A documented rest API for developers sits alongside it, and there is a page addressed to AI systems in the footer. Not A because the conventional connector set is three CRMs and nothing else, with no sequencer, dialer, marketing platform or verified marketplace listing.
Recorded as observed rather than concluded: no hosting provider, region, residency option or sub processor list was located on the vendor's own surface. What is established is jurisdictional, that the operating entity is Sales Led Oy, a Finnish company, and the product is bilingual English and Finnish with a Nordic customer base.
A third party review platform reports that data is processed in the EU and handled in line with GDPR, which is consistent with everything else about the vendor and is very likely true, but it is a third party's account of the vendor's claim rather than a statement read on the vendor's own page, and this index grades what is published. Re verify at the privacy policy.
Recorded as observed rather than concluded: no security page, trust centre, certification claim, penetration test or vulnerability disclosure route was located, and the footer's Legal section contains exactly three items, terms and conditions, privacy policy and cookie policy, with no security entry alongside them.
For a company founded in 2023 the absence of a SOC 2 or ISO audit is stage appropriate rather than negligent, and the honest reading is that certification has not yet happened rather than that it is being concealed. What is at stake meanwhile is worth stating: the platform holds write access into the buyer's CRM, pushing records automatically into HubSpot, Pipedrive or Salesforce, and it holds sending access for sequences.
The second pricing page with no price graded in this session, and this one is sharper than Clari's because the vendor labels it transparent. The page carries the eyebrow transparent pricing above the headline Pricing built for growth, and its meta description promises simple, transparent pricing. Both plans, the Finnish database and the European database, display the words ask for pricing where the number belongs, each followed by a Book a meeting button.
There is no tier, no seat rate, no unit, no floor, no range and no currency anywhere on it. Third party reviews from mid 2026 record a homepage line placing list building from 199 EUR per month and report earlier figures between 49 and 99 USD; none of those reconcile with each other or appear on the current page, which indicates repackaging rather than disclosure.
What the page does do well is describe precisely what each plan contains, seven itemised inclusions each, with API and MCP access in both. A buyer can therefore establish exactly what they would get and nothing at all about what it costs.
Recorded as observed rather than concluded: terms and conditions are published but were not read, and no post termination data right, retention window, deletion timeline or export path was located. Two structural points cut in the buyer's favour and neither is claimed by the vendor as a commitment. The product's whole purpose is to write records into the customer's own CRM, so the operative output already lives in a system the buyer owns and would survive the relationship ending.
And both plans include REST API and MCP access as standard, which is a working extraction route rather than a support ticket. Against that, the mapped market, coverage history and target market radar analytics are the vendor's, and continuously refreshed data stops being refreshed the day the contract does.
Recorded as observed rather than concluded: multichannel sequences send email and nothing published addresses mailbox warmup, domain authentication, inbox placement, bounce or complaint thresholds, volume governance or sender reputation.
The vendor's argument on this axis is indirect and made through data quality rather than sending discipline, that continuously verified records against official registers mean fewer sends to dead addresses, which is a real contributor to deliverability and not a substitute for the controls themselves. The phone side is the larger channel here in any case, since the product is explicitly built to get reps on the phone, and telephony discipline is likewise unaddressed.
Unusually precise in both directions. The geography is stated market by market rather than as a region: Finland as a dedicated product, and a European product covering the Nordics, Baltics, DACH, Benelux, Poland and the UK, each mapped from its own local business register. The buyer is segmented into four seats with a page each, CSOs and sales leaders, sales reps, marketing teams and RevOps, and the site runs in English and Finnish.
Roughly 500 B2B sales teams, twelve named customer logos, and a positioning that is explicit about being a poor fit for anyone whose pipeline sits outside the covered markets, which is the anti ICP candour Boomerang was credited for.
Not A because the coverage is deliberately regional, because the evidenced customer base is concentrated in Finland and the Nordics rather than spread across the European footprint the pricing page claims, and because there is no enterprise apparatus described.
Compared With
Editorial comparisons are published only where the index assesses two vendors as direct competitors for the same buyer. Each carries a verdict, the buyer conditions that favor each vendor, and a graded side by side.
Pricing
What this vendor charges, what it commits to in writing, and where the bill can move. Figures the vendor publishes itself are labeled Vendor Published. Figures labeled Estimated come from other sources and the vendor has not confirmed them.
- ›Nothing loads. The pricing page serves 32 kilobytes of navigation and menus and no prices, no plans and no units at all to a machine reading it.
- ›The one thing it does put in the readable part is an announcement that its data now works directly inside AI assistants for every customer.
- ›That is worth pausing on. A data company building specifically so assistants can read its product, on a pricing page those same assistants cannot read at all.
- ›From the navigation you can see there are two separate databases, one Finnish and one European, plus an outreach tool. Whether those are sold together or separately is the first question and nothing answers it.
- ›This is a European vendor selling European data, so ask what currency the contract is in, and ask on what basis the records were collected, because you take on that position when you use them.
How the price works
What you are charged for, and what makes the bill go up.
Not retrievable. The pricing page renders client side and the served document, at 32 kilobytes, contains only site navigation, a language switcher and an announcement banner. No tier name, figure, band, metering unit, seat concept, trial term or structured data offer is present in any form.
The product structure visible from navigation rather than from pricing comprises a data foundation, a target market monitoring capability, integrations, an outreach function, and two separately named databases covering Finland and Europe more broadly. Whether the two databases are licensed together or separately is not established.
The served portion carries an announcement that a model context interface is live for every customer, allowing the vendor's data to be used directly inside named conversational assistants, described as available to all customers rather than as a paid addition.
The site is published in two languages and the vendor is European, so pricing is likely denominated in euros. Nothing served confirms the currency and no figure is recorded in any currency.
The visible purchase route is a demonstration booking rather than a self serve signup.
No credible third party estimate was located, so none is recorded.
What the contract says about your data
What the vendor commits to in writing once your data is in the product.
Not established. The pricing page renders client side and served navigation only, so no legal or security links were present in the retrieved markup. No processing agreement, sub processor listing, certification claim, retention period or residency statement was located.
That is a retrieval limitation rather than a finding, and it compounds the pricing limitation rather than standing separately: the same rendering behavior that hid the plan cards would hide a legal footer.
The custody question is more consequential for this vendor than for most comparable ones, and the reason is geographic. This is a European data supplier whose stated product is Finnish and wider European business contact data, sold to European buyers. That places both the vendor and its customers squarely inside the European regulatory regime for the whole of the transaction, rather than the common pattern in this index of a United States platform supplying European buyers.
The consequence is that lawful basis is not a procurement formality here but the substance of the product. A buyer should establish on what basis the Finnish and European records were assembled, what notice the individuals in them received, and how removal requests are handled, because the buyer inherits that position when they act on the data.
One further consideration follows from the vendor's own announcement that its data is available through a model context interface for use inside conversational assistants. That means records can flow into a third party assistant's context, and a buyer should establish what that route retains and where it processes.
Getting started
What it costs and what is included before the product is running.
Not established. Nothing about fees, trials, minimums, contract length or onboarding was present in the served document, which carried navigation only.
The purchase route visible in the served portion is a demonstration booking rather than a self serve path, which suggests a sales assisted motion, though that inference rests on the absence of a signup route rather than on any published statement.
Two cost questions are implied by the product structure and neither is answerable from what was served. The vendor names a Finnish database and a European database as separate items in its navigation, so whether those are licensed together, separately or as tiers of one another is the first commercial question and it is unresolved. And an outreach capability is named alongside the data, so whether the platform is sold as data, as an outreach tool, or as a bundle determines what it is being compared against.
One further item belongs in a buyer's model and is published rather than inferred. The vendor states that its data is available through a model context interface to every customer, described as included rather than as an upgrade. If that is genuinely inclusive it removes an integration cost that several vendors in this index charge for, and it is worth confirming at the point of quote.
A buyer should ask for the two databases priced separately, and should establish the currency of denomination before comparing against any figure in this index.
What to watch for
Where this pricing can surprise a buyer who has not read it closely.
The pricing page renders entirely client side and serves navigation only, which places it in the same category as ActiveCampaign though at a fraction of the size.
The served document is 32 kilobytes and contains the site navigation, the language switcher and the announcement banner. It carries no tier name, no figure, no band, no metering unit, no seat concept, no trial term and no structured data offer. Everything a buyer would read arrives after the scripts run.
What is worth recording is what the vendor chose to put in the served portion, because it is not pricing. The top of the document announces that a model context interface is live for every customer, so that its data can be used directly inside named conversational assistants. That is the first vendor in this index to lead its pricing page with a machine consumption announcement, and it sits oddly against a pricing page a machine cannot read.
That juxtaposition is the finding. A data supplier building deliberately for retrieval by assistants, whose own commercial terms are invisible to those same assistants, is the clearest single instance of the split this property exists to document. Aimdoc published a machine callable gateway beside a price only its metadata carried; this vendor announces machine access to its product on a page that tells a machine nothing about what that access costs.
The product structure is partially visible from navigation rather than from pricing: a data foundation, a target market monitoring capability, integrations, an outreach function, and two separately named databases covering Finland and Europe more broadly. The existence of two distinct databases suggests they may be licensed separately, which would be the first commercially relevant question, and nothing served addresses it.
The page is published in two languages, which for a European vendor implies pricing may be denominated in euros. Nothing served confirms the currency, and no figure is recorded in any currency rather than an assumed one.
No dollar figure is recorded in the numeric field and no estimate is recorded in the display field.