Most outbound copy fails for one reason. The sender doesn't know the lead. They know the title, the company, the LinkedIn headline, and not much else. So the email reaches for something generic: "I see you're scaling..." or "Just wanted to share..."
Twain does that work before it writes. It researches the person and company, connects what it finds to your offer, and drafts the outreach from that context. High and Ultra show the qualification in Thesis; Ultra also builds a Strategy for the sequence. Your rep starts with the evidence, the reasoning, and a draft, instead of assembling all three from scratch.
The saved demonstration below reverses the usual roles: Twain is the prospect, and the lead is Twain's CEO, Mohamed Chahin. The seller is Merge, used here as an example of a company with a CRM integration offer. The excerpts follow Twain's Salesforce integration work through qualification and strategy to a cold call script for Mohamed, and all of them are saved results from the same lead.
Why research changes the message
A job title tells you roughly what someone does. It doesn't tell you which part of your offer matters to them, what they have already tried, or why they would want to hear from you now.
That is the work behind AI account research. Twain brings information about the account and contact together with your agent and campaign brief. The same fact can matter differently to two sellers: a hiring push might create a reason to discuss recruiting, onboarding, or sales capacity. Your offer and the prospect's situation determine which connection belongs in the message.
Twain uses that connection to shape the outreach. A recent post gives the message a useful opening when it explains why your offer matters to that person. Finding the post is only part of the work; deciding what to do with it is what turns research into a conversation.
Start with the person and company
Research brings the person, company, and signals behind the outreach into one lead view, with source references you can follow. When you supply previous interactions, Relationship adds that history to the context, so the next message can continue the conversation your team has already started.
For Mohamed, Research found his Salesforce export announcement, a shared example connecting Twain with Clay and HubSpot, and his comments about using Twain's research through MCP in Claude. One of the saved signals reads:
Mohamed announced native integration support enabling users to export Twain campaigns and deep prospect research directly into Salesforce.
The report links that signal to his LinkedIn announcement. It establishes that CRM connectivity matters to the product. The other signals add context, but a message about CRM integrations does not need to mention every piece of research Twain found.
High and Ultra also show Absent signals. They make gaps in the research visible, so a missing buying trigger does not quietly become an assumption in your pitch.
How Thesis connects the research to your offer
Thesis, available in High and Ultra, builds the qualification for each prospect. It brings together their situation, persona fit, why the timing matters, and why your company is relevant. That gives the writer a reason to approach this person, and gives your rep the reasoning behind it.
This saves a separate piece of work after research. A rep would otherwise have to read the signals, compare them with the offer, and explain why this contact belongs in the campaign. Thesis puts that reasoning beside the lead, including the distinction between an account that fits and a person whose role makes them worth approaching.
Here is the connection the saved Thesis makes between Twain's product and the seller's offer:
The overlap: Twain delivers its value through CRM context and export targets, while Merge sells a Unified API for read and write access across third-party systems. The open question is whether Twain's small team builds connector N plus 1 when a customer uses a long-tail CRM instead of HubSpot, Salesforce, or Attio.
The persona assessment records a partial fit: the founder role and CRM context are established, while integration pain is unproven. That distinction changes the outreach. The rep has a relevant question about future CRM requests instead of a pitch that assumes Twain's existing integrations are broken.
How Strategy turns the qualification into outreach
Ultra adds Strategy to decide how to approach the prospect. It chooses which evidence should lead, how to frame the offer, which proof is relevant, and how to develop the sequence across touches. The decisions are specific to the person and campaign, so the sequence has a direction before the first message is written.
That matters when several facts could make a plausible opening. The newest signal is not necessarily the strongest reason to write, and a proof point that fits one buyer may mean little to another. Strategy makes the selection and explains the approach, rather than leaving the rep to turn a list of facts into a sequence.
For this lead, the saved sequence strategy says:
Lead with the fused insight that Twain built bidirectional HubSpot and Salesforce sync, alongside Attio, Clay, and n8n connectivity, and ask about the next CRM a customer runs. It leads because it proves accurate research, respects what Twain already built, and makes the long-tail question without inventing a problem. Use the primary ask for a 15-minute technical fit conversation.
The same report sets aside the seed round and pricing as irrelevant to this approach. It keeps the MCP thread for a later angle. That selection is useful: the rep gets a focused opening and a way to develop the conversation, instead of a message crowded with everything the research found.
Sequence contains the messages or scripts built from that work. The reasoning remains alongside them, so your rep can see why Twain chose the approach and refine it. The same process grounds cold email personalization: the detail in the opening supports the offer in the rest of the email.
See the research in the message
Here is the saved Ultra cold call script for Mohamed, unchanged. It calls the HubSpot and Salesforce sync a recent release because it was recent when this demonstration was saved:
The key question is: "Are requests for CRMs beyond HubSpot and Salesforce showing up in your pipeline yet, or is that covering it for now?"
The script leads with the existing integrations, asks whether other CRM requests are appearing, and offers a technical fit discussion. The question gives Mohamed room to say the current coverage is enough. The rep has a researched reason to call and a concrete question ready to use, without having to translate the research into a script.
The caller fills in their name and reviews the script before using it. Twain prepares the outreach; the rep makes the call and follows the conversation. Our cold-calling walkthrough shows how call and voicemail steps fit into a campaign.
Choose the depth for the work
Choose how much of the research and outreach planning you want Twain to do:
- Base researches the account and contact, then writes email outreach from your instructions. You get Research and Sequence.
- High adds deeper research, Absent signals, and qualification through Thesis. Use it when you want the reasoning behind the fit as well as multichannel drafts.
- Ultra adds Strategy to choose the evidence, framing, and progression of the sequence. Use it when you want Twain to make more of the outreach decisions for each prospect.
The mode determines which tabs and reasoning you receive. Check its displayed credit cost before generating.
Try Ultra on one lead in an existing campaign to see the full Research → Thesis → Strategy → Sequence flow. Follow the evidence into the message and see which research and planning decisions Twain has made for you. Your campaign brief supplies the reason for reaching out and the offer; refine it before generating the rest of the list if either needs to change.