Scalantec co-founder Nicolas Schell joined Sales Pepper, the podcast hosted by Dominic Klingberg. The topic: the data enrichment tool Clay, for which Nicolas also runs the Berlin Clay community, and the concrete cost and practical questions most beginners run into first.
How Nicolas found Clay
Nicolas comes from a sales background himself, with up to 150 cold calls a day and plenty of manual research for personalized outreach. The push toward Clay came from a former manager who asked him to review the existing sales tech stack. Nicolas went on to test over 50 different tools before he landed on Clay. Today, Scalantec builds three core use cases for clients on top of it: classic B2B outbound (roughly 80 percent of the work, according to Nicolas), CRM enrichment, and signal-based approaches.
Why Clay specifically
Nicolas describes Clay as an orchestration tool, the centerpiece of every go-to-market strategy Scalantec builds for clients. Three things stood out to him in particular:
- Over 150 native integrations: unlike a fixed subscription to a single database (in the past, tools like Hunter or Apollo), Clay charges usage-based via a credit system, only for the data you actually pull.
- Native connections to outreach tools and CRMs: Lemlist, Instantly, Smartlead, and Outreach connect just as directly as Salesforce or HubSpot.
- The "Claygent": an AI agent inside Clay that can independently visit websites and research information, for example pulling a CEO's name from a legal notice page.
The Clay community
Nicolas runs one of now over 16 city-based Clay communities, which Clay itself supports with a quarterly events budget. That includes smaller "Clay Days" with 20 to 25 attendees working on their own tables together, as well as larger "Clay Demo Nights" with 70 to 80 attendees where more complex use cases get presented. In the worldwide "Global Clay Comp," Nicolas placed 11th.
Common beginner mistakes and how to avoid them
Asked about frequent pitfalls, Nicolas names several recurring patterns:
- Auto-update on enrichment columns: left switched on, a column refreshes automatically the moment the underlying field changes, which can quietly burn through a huge number of credits. Nicolas's advice: turn it off, consistently.
- Missing manual review: Clay is no miracle tool, in Nicolas's view. He recommends first thinking through what you'd research manually if you had unlimited time, then spot-checking the automated results against that.
- Tier limits: on the cheapest plan (around 134 US dollars, according to Nicolas), sequencing tools like Instantly or Lemlist and CRM connections aren't available yet, that requires a higher tier.
Using Clay more cheaply
Nicolas names two concrete cost levers:
- Your own API key instead of Clay credits for the Claygent: with your own access to Gemini, Claude, or ChatGPT, LLM tasks run far cheaper than through Clay credits (according to Nicolas, sometimes around a tenth of a cent per task, versus 1.6 to over 2 cents per Clay credit). Anyone who hasn't used their own API key yet should contact Clay support directly, according to Nicolas, to get onto the right tier.
- Your own data-provider subscriptions: if you're already using tools like LeadMagic, Prospeo, or Better Contact, you save by routing them through your own subscription instead of through Clay credits.
Who Clay pays off for, and who it doesn't (yet)
On a very tight budget, Nicolas recommends sticking to manual channels first (phone, hand-written emails, and LinkedIn messages), since realistically running Clay well costs at least a few hundred euros a month, on top of outreach tools and databases. Clay also doesn't make much sense without a clearly defined ICP or without product-market fit.
On the other hand, anyone in the enterprise or mid-market segment who needs to research companies barely findable on LinkedIn, for example via commercial registers, or anyone in the SMB space who wants to personalize high outbound volume, finds a real lever in Clay, according to Nicolas. One example he gives: automated account research for a smaller total addressable market, where Clay pulls together things like ownership structure and strategic statements from annual reports and feeds the result straight into the CRM or into a document for sales.
Cold outreach in practice: what still works today
Nicolas names several concrete rules Scalantec applies in its own outreach for clients:
"I'd never go in with a hard sales pitch in cold outbound. I try to keep the whole thing as conversational as possible, like I'm talking to a buddy."
- Short messages: ideally under 75 words, and under 40 words for Outlook recipients, since Microsoft filters noticeably more aggressively than Gmail, in Nicolas's experience.
- Avoid obvious AI tells: sentences of the same length and long em dashes give away AI-generated copy instantly, according to Nicolas.
- Dedicated sending infrastructure instead of the main domain: around ten similarly named secondary domains with two inboxes each, warmed up for three to four weeks before use, to protect the reputation of the company's actual domain.
- Real personalization instead of standard triggers: lines like "Congrats on your funding round" or "I saw you're hiring" mostly no longer work, according to Nicolas.
As an example of real relevance, he cites Scalantec client Seven Senders: for its solution automating the claims process on lost parcels, Scalantec researched in advance which carriers a target company works with and how high its goods value is, and addressed that exact point directly. More on the work with Seven Senders in the full case study.
One more rule of thumb: never pair a LinkedIn connection request with a pitch in the very next message.
"Please, please don't do that. I get so many connection requests, and then the second message is straight into a mile-long pitch. Don't do it."
FAQ: Clay in B2B sales
When does Clay pay off for a company?
Mainly with a clearly defined ICP, existing product-market fit, and a budget of at least a few hundred euros a month. Especially valuable for account research in enterprise segments and for high-volume outbound in the SMB market.
How can you reduce Clay costs?
By using your own API key with Gemini, Claude, or ChatGPT for the Claygent instead of Clay credits, and by using your own subscriptions with data providers like LeadMagic, Prospeo, or Better Contact instead of pulling that data through Clay credits.
What's the most common beginner mistake with Clay?
Leaving auto-update enabled on enrichment columns, which can quietly burn through large amounts of credits, plus too little manual review of the automatically generated results.
What still makes cold outreach messages successful today?
Short, conversational messages with real personalization tied to a specific pain point instead of generic triggers, plus a dedicated, properly warmed-up sending infrastructure kept separate from the main domain.
How would this work in your team?
In a free 30-minute call we look at what it would look like on your stack.
