Today
The team runs on unstructured, often thin account research. Tabs, a ChatGPT summary, a half-finished doc. Every rep does it their own way, so nothing is shared and nothing is coachable.
Here's how to improve it with Legatus.
Account research
Let Legatus, your AI agent, research the account end to end and match it against your ICP, the problems you solve and the customers you want.
Account signals
Every account comes back the same way. Facts first, sourced line by line. Then the verdict: does it fit your ICP and is it worth working now? Reasoning shown, so nothing runs on faith.
Full intelligence on Formalize
Julie B. is owner
Why this account?
Formalize crossed 160 employees while running a live pivot from single-product whistleblowing into a full GRC platform — across four countries, under Series B board pressure to deploy €30M into ARR growth fast. The scale and the transformation are happening at the same time.
Why now?
The Munich office is under a named end-of-2026 headcount target (30 employees). The CRO and CFO have both departed since the Series B closed. There is no dedicated People leader in evidence at a company now running four offices across four countries under board pressure to grow ARR fast.
Why us?
We work exclusively with Nordic scale-ups in the 100–500 employee range going through exactly this inflection — and both Heartwood founders built people functions inside Nordic scale-ups before starting the practice. Formalize’s current profile matches the pattern we know best: post-Series B, multi-country expansion, no dedicated People leader, leadership team under pressure.
ICP scoring
Account details
Ownership
Account state
Research history
Account playbook
Then structured as a playbook. The account story, why now, who to talk to and the first move. A rep who has never seen the account can make a credible move tomorrow.
Julie B. is winning
Formalize crossed 160 employees while running a live pivot from single-product whistleblowing into a full GRC platform — across four countries, under Series B board pressure to deploy €30M into ARR growth fast. The scale and the transformation are happening at the same time.
Why now
The Munich office is under a named end-of-2026 headcount target (30 employees). The CRO and CFO have both departed since the Series B closed. There is no dedicated People leader in evidence at a company now running four offices across four countries under board pressure to grow ARR fast.
Why us
We work exclusively with Nordic scale-ups in the 100–500 employee range going through exactly this inflection — and both Heartwood founders built people functions inside Nordic scale-ups before starting the practice. Formalize’s current profile matches the pattern we know best: post-Series B, multi-country expansion, no dedicated People leader, leadership team under pressure.
Core challenge
No dedicated People leader during multi-country expansion
Formalize is scaling past 160 employees across four countries with no VP People, CPO, or CHRO in post. The people systems — compensation, performance, onboarding, hiring frameworks — that a company at this stage needs to operate are either informal or absent, and the gap compounds every month as the Munich ramp adds headcount and the CRO and CFO seats remain unfilled.
Other challenges
By hand vs Legatus AI
Work 3x the accounts. Same headcount.
Check the math behind itHow long does account research take?
Account research is the work of building a full picture of a target account before connecting: the facts, the fit, the timing. By hand it takes 1 to 3 hours per account, and that's just the facts. A complete picture with strategy and content behind it runs 4 to 6 hours.
That number is why research stays thin. Nobody can afford it for account 346 on the list, so teams cut corners: real research for the top five, a few tabs and a ChatGPT summary for the rest. Legatus runs the same account in 14 to 21 minutes. Check the math behind it.
How to automate account research
Account research only becomes valuable when it's systematic and every rep uses it. One rep researching well is a skill. Every rep working from the same picture is a system, and that takes three changes.
Same depth on every account: an agent doesn't cut corners, so account 346 gets the same treatment as the top five. One shared picture: research lands on the account, not in a rep's head or a half-finished doc, so everyone sees the same thing and managers can coach against it. Facts separated from judgment: the research carries what's true, and the verdict on whether the account is worth working is made against your ICP, the same way every time.
How to use an AI agent for account research
The difference between an agent and a chat prompt is what stands behind the answer. A prompt is a one-off: a rep writes it, gets a tidy summary built from generic internet content, and can't verify a line of it. Convincing enough to trust, which is the danger.
An agent works differently on all three counts. It runs while you work, not when a rep remembers to ask. It's trained on your company: your ICP, the problems you solve, the customers you want. And it goes and reads: the funding announcement, the job postings, the CEO's stage talk, the strategy report. It comes back with findings, each one sourced, so a rep can check any line before building on it. That reading-and-judging layer separates research from enrichment, and it's the work no team has the capacity to do by hand across a full list.
What should account research include?
Start by understanding the account story behind the company, then build the buying roster of relevant people. From there it's reading work: industry insights, challenges, timing signals, company news, hiring signals. Most research stops before the layer that matters: every source listed, the material behind it, what's uncertain or missing, how fresh each finding is. A full account picture includes what isn't known. That's what makes the rest trustworthy.
It can give you a summary, and that's the problem: a one-off, built from generic internet content, with nothing sourced. It reads well and can't be checked. For a quick orientation it's fine. For research a rep will act on, you need findings that stand behind their sources. That's the difference an agent makes.
By hand: firmographics, funding, hiring, leadership, then the reading: recent talks, reports, posts. Plan 1 to 3 hours per account. On Republik you open the account's playbook instead; the research is already done, sourced and matched against your ICP.
A template structures the work, it doesn't do it. If you build one, cover the company, the people, hiring and timing signals, challenges and the sources behind it all. On Republik that structure is the research output itself, filled in on every account you name.
Two checks, in order. Does it match your ICP, the criteria an account must meet to belong on your list. And is it worth working now, meaning a real problem you can point to and a reason to move on it today. Legatus scores both on every account and shows the reasoning, so the verdict is checkable, not a gut call.
Agent training is part of onboarding: your ICP, the problems you solve, the customers you want more of, the references you've earned. That's what lets it judge fit and timing instead of just collecting facts.
Public sources: announcements, job postings, filings, talks, posts. Every finding carries its source, so any line can be verified. Accounts are refreshed quarterly or when you're ready to work them, so you never open on stale data.