How to Use n8n with ._Template 180 Sales Intelligence Account Research
Your sales team spends 6–10 hours a week on account research that a workflow could do in seconds. Someone opens LinkedIn, checks the company website, pulls headcount from a data provider, skims recent
Your sales team spends 6–10 hours a week on account research that a workflow could do in seconds. Someone opens LinkedIn, checks the company website, pulls headcount from a data provider, skims recent news, and pastes fragments into a CRM note before the first outreach email ever goes out. Template 180 — Sales Intelligence Account Research — replaces that manual grind with an n8n workflow that enriches any company the moment it lands in your pipeline. This article shows you exactly how to run it.
The Problem: Account Research Doesn't Scale With Headcount
Manual account research fails for three structural reasons. First, it's linear — one rep researching one account at a time, with no leverage. Add more leads and you either add more reps or let accounts sit unresearched. Second, it's inconsistent. Every rep pulls different data points, formats notes differently, and stops digging at different depths, so your CRM fills with uneven records that can't be scored or segmented reliably. Third, it's stale by the time you use it. A company that raised funding, changed CEOs, or doubled headcount last month gets contacted with data pulled six weeks ago.
For a busy founder or a lean ops team, the real cost isn't the hours — it's the opportunity. Reps researching are reps not selling. And the accounts that deserve the deepest research (your ICP fits, your enterprise targets) are exactly the ones that get shortchanged when a rep is racing through a list of 40 companies before lunch.
The Solution: An Automated Enrichment Pipeline
Template 180 turns account research into an event-driven pipeline. You feed it a company domain or name, and it returns a structured intelligence profile: firmographics (industry, employee count, revenue band, HQ location), technographics (what tools the company runs), recent signals (funding, hiring, news mentions), and a plain-language summary your reps can read in ten seconds.
The design principle is simple: research should happen once, automatically, and be reusable by everyone. Instead of a rep doing ad-hoc digging, the workflow standardizes the data points collected, writes them back to your CRM, and optionally scores the account against your ICP so reps prioritize instead of investigate. The workflow runs on n8n, which means it's self-hosted, auditable, and costs you nothing per run beyond the enrichment APIs you already pay for.
Step-by-Step Setup in n8n
Here's how the workflow is wired. Import the template JSON into your n8n instance, then configure these nodes in order.
1. Trigger node. Choose the entry point that matches your stack. Use a Webhook node if you want your CRM (HubSpot, Pipedrive) to fire the workflow when a new company is created — set the HTTP Method to POST and copy the production URL into your CRM's outbound webhook settings. For batch enrichment of an existing list, swap in a Schedule Trigger node running nightly, followed by a CRM query node that pulls accounts where an enriched flag is empty.
2. Normalize input. Add a Set node (or an Edit Fields node in newer versions) to standardize the incoming payload down to a single clean field: domain. Strip https://, www., and trailing slashes here so downstream API calls don't fail on formatting. A tiny Code node with return items.map(i => ({ json: { domain: i.json.domain.replace(/^https?:\/\/(www\.)?/, '').split('/')[0] } })) handles the edge cases cleanly.
3. Firmographic lookup. Use an HTTP Request node pointed at your enrichment provider (Clearbit, Apollo, or People Data Labs all work). Set Authentication to "Header Auth" and store the API key in n8n Credentials — never hardcode it in the node. Pass {{ $json.domain }} as the query parameter. Enable "Retry On Fail" with 3 attempts and a 5-second wait, because enrichment APIs rate-limit aggressively.
4. Recent signals. Add a parallel HTTP Request node hitting a news/signals endpoint (or a simple Google News RSS fetch) using the company name returned from step 3. This is what makes the profile feel current instead of a static database dump.
5. Merge and shape. Wire both HTTP Request outputs into a Merge node set to "Combine" by position, then a Set node that maps the raw API fields into your CRM's exact field names — employee_count, industry, tech_stack, last_signal, research_summary. Being explicit here is what keeps your CRM clean.
6. Optional AI summary. Drop in an AI Agent or Basic LLM Chain node connected to a Claude model (claude-sonnet-5 is the right balance of speed and quality for this). Prompt it: "Given this firmographic and signal data, write a 3-sentence account brief a sales rep can act on. Lead with why this account fits our ICP." This converts raw data into something a rep actually reads.
7. Write-back. Finish with your CRM node (HubSpot, Pipedrive, or a generic HTTP Request PUT) that updates the account record and flips the enriched flag to true so it never gets processed twice.
The Payoff: What Changes When This Runs
Once live, the workflow reshapes how your team operates. Reps open an account in the CRM and the research is already there — no tab-switching, no data providers, no guesswork. Response time to inbound leads drops from hours to seconds, which matters enormously because the odds of qualifying a lead fall off a cliff after the first five minutes.
Consistency is the quieter win. Every account gets the same depth of research, formatted identically, which means your account scoring and territory segmentation finally run on trustworthy data. And because the pipeline is event-driven, it scales to 10 accounts or 10,000 with no additional labor — you're paying for API calls, not headcount. For a lean team, that's the difference between a founder-led sales motion and a repeatable machine.
Common Pitfalls to Avoid
Skipping deduplication. Without the enriched flag check in step 7, a re-triggered webhook will re-enrich the same account, burning API credits and overwriting good data. Always gate the workflow so each domain processes once.
Hardcoding API keys. Keys pasted directly into HTTP Request nodes leak the moment you export or share the workflow. Use n8n's Credentials store exclusively — it encrypts them at rest and keeps them out of your workflow JSON.
Ignoring rate limits. Enrichment APIs will silently return 429 errors under load. If you don't enable "Retry On Fail" and add a small delay, batch runs will drop accounts without warning. For large lists, insert a Loop Over Items node with a batch size of 10 and a short wait between batches.
Over-collecting data. It's tempting to pull every field an API offers. Don't. Collect only the data points your reps actually use to make a decision. A bloated record is as useless as an empty one, and it makes your CRM slower to read at the exact moment speed matters.
No fallback path. Some domains won't resolve — they're too small, too new, or the API has no record. Add an IF node after the firmographic lookup that routes empty results to a "needs manual review" queue instead of writing blank fields over your account. A graceful miss beats a corrupted record.
Template 180 isn't magic — it's leverage. It takes the one task that consumes your team's most expensive hours and turns it into infrastructure that runs while everyone sleeps. Import the workflow, wire the seven nodes, and let your reps go back to doing the one thing automation can't: closing.