Disclosure: This tutorial is based on Gumloop’s public documentation and pricing pages checked on October 26, 2026. Prices and credits can change, so check the vendor’s site before subscribing. Some links on this site are affiliate links; see our Affiliate Disclosure page for details.
What is Gumloop?
Gumloop is an AI agent platform that builds agents to use your tools and data to do real work. In the vendor’s own description, you can build agents that understand your business, work across your tools, and take work from idea to outcome. That is a useful way to think about it for lead research: instead of opening ten tabs, copying details by hand, and pasting them into a spreadsheet, you set up an agent or flow once, give it a clear job, and let it repeat that job in a consistent way.
You do not need to write code to get started. The vendor presents understanding the task itself as the main prerequisite for building an agent, and its documentation says agents can then be automated with triggers, schedules, and the API. In practice, that means you describe the outcome you want, connect the tools the agent is allowed to use, and decide when it should run. The current vendor site puts agents at the centre of the product, and its pricing page lists older workflows as legacy, so this tutorial describes a workflow pattern you can build with an agent or flow rather than naming a specific node that may have been renamed.
Gumloop also offers a wide set of connections, including email, chat tools, spreadsheets, customer tools, and a real browser that can open pages and pull data from sites your team already uses. Its own site lists use cases such as lead generation, lead qualification, content creation, CRM work, and data analysis. Those are exactly the kinds of repetitive, research-heavy jobs where a small team can save the most time, provided the output is checked before it is used.
What you can automate with Gumloop
Lead research is a good first project because it is repetitive, easy to define, and easy to check. The same pattern also works for several neighbouring jobs, as long as you keep a human in the loop for anything that will be sent to a customer.
- Lead research and enrichment: collect information about a company or person, then use AI to summarise whether they match your ideal customer and what to say to them first.
- Web research into spreadsheets: read company pages, profiles, or a source sheet, extract the fields you need, and send a tidy row of results to a spreadsheet for review.
- Content drafting workflows: turn research notes, product details, or a brief into a first draft, an outline, or a set of social posts that a person can then edit.
- CRM updates: add new leads, update existing records, and keep notes in one place instead of copying the same details between tools by hand.
- Document and data processing: read files, sheets, and connected sources, summarise them, and turn scattered information into a short, usable summary.
The lead research workflow, step by step
Start small. Run this workflow on ten to twenty leads first, check every result, and only then scale it up. A workflow that looks impressive on a hundred leads is useless if half of the details are wrong, and lead research is exactly where small errors become embarrassing emails.
- Define your ideal lead: write down the industry, company size, role, and location you are targeting, plus any deal-breakers. This definition becomes the scoring rule in step five, so be specific rather than writing a vague wish list.
- Create a new flow in Gumloop: start a new agent or flow and give it a clear name, such as Lead Research. Write a short description of the job in plain language: who it should research, what it should return, and what it should never do, such as sending an email on its own.
- Add a step that collects company or profile pages to research: bring in the pages, profiles, or source list you want the workflow to work through. Keep the source list small and clean at first, because a messy input list produces messy research no matter how good the AI step is.
- Add an AI step that extracts and summarises the fields you need: ask for the same fields every time, for example name, company, role, contact page, and fit notes. Asking for a fixed set of fields makes the results much easier to compare, filter, and send to a spreadsheet or CRM.
- Add a step that scores each lead against your criteria: give each lead a simple score or label based on the ideal-lead definition from step one. A three-point scale, such as strong fit, possible fit, or poor fit, is usually more useful than a complicated score nobody can explain.
- Send the results to a spreadsheet or CRM: send the scored results, with the source page attached, to a sheet or CRM where a person can review them. Do not send outreach directly from the workflow until you have checked the data and the wording.
- Run the flow on a schedule or on demand, and review a sample of results manually before using them: schedule it for a regular research batch, or run it when you add a new list. Before using any batch, open a sample of source pages, check names, roles, and companies, and fix your instructions where the AI misunderstood the brief.
The most important habit in this whole tutorial is that last check. AI research is fast, but it can also be confidently wrong about a job title, a company size, or a contact page. A five-minute manual review protects your reputation and, over time, teaches you exactly which instructions need to be tightened.
Gumloop pricing
Pricing was checked on October 26, 2026 on Gumloop’s own pricing page and in its credits documentation. The Pro plan starts at $37 per month and includes 20,000 credits per month. The documentation breaks that total down as 7,400 credits at the $0.005 list price plus 12,600 bonus credits. Enterprise pricing is custom, with a custom credit allocation. The pricing page lists only Pro and Enterprise, so there is no free plan shown alongside them.
Every new account starts with a 14-day free trial of Pro. That is the free way to try Gumloop, but it is not a free plan: the trial requires a card, it is a one-time offer per customer, and it rolls into a paid Pro subscription unless you cancel during the trial. Cancelling during the trial stops the card from being charged when the trial ends, according to the credits documentation.
Credits are Gumloop’s currency for agent work, and they are not a fixed price per message. One credit is worth $0.005, so $1 buys 200 credits. Model calls bill at cost, each successful connector tool call costs 1 credit plus that tool’s own charge, active compute costs 5 credits per session-minute, and an orchestration fee of 8% is added to those three parts. The documentation gives a useful warning: the same agent might cost 2 credits for a quick question and 200 credits for a deep research task. If you run out of credits, agents stop unless you enable pay-as-you-go overage, billed at $0.005 per credit, and Pro overage is capped. Pro credits do not roll over month to month, except on Enterprise plans. That is why the first few runs matter so much: they tell you what a typical lead batch actually costs before you commit to a larger list.
Gumloop pros and cons
Pros:
- No-code builder for AI-heavy workflows: you can build agents and flows around research, enrichment, and drafting without writing scripts, as long as you can describe the task clearly.
- Generous credits on Pro: 20,000 credits per month gives a small team real room to test lead research batches, especially when a quick task may cost only a handful of credits.
- Faster than stitching together single-purpose tools: one agent can collect pages, summarise them, score the results, and send them to a sheet or CRM, instead of moving data manually between four separate tools.
Cons:
- Credit-based pricing takes a few runs to predict: because costs vary with the model, the tools used, and how long the agent works, you cannot know the exact cost of a new workflow until you have tested it.
- Very new workflows still need human checking: AI research and drafting can contain errors, so every batch needs a manual sample check before outreach, CRM updates, or publication.
- Advanced users may prefer full code or self-hosted automation: if you want complete control over hosting, code, and infrastructure, a visual agent platform will feel more constrained than building the same workflow yourself.
Gumloop vs Zapier, Make and n8n
Zapier and Make are broader general automation tools with more app integrations in the usual sense: they are built to move data between apps, trigger actions when something happens, and connect the everyday tools a business already uses. If your main need is simple, reliable app-to-app automation, that breadth is a real advantage, and Gumloop is not the obvious first choice for that job alone.
n8n sits at the other end of the spectrum. It is self-hostable and developer-friendly, which suits teams that want to run their own infrastructure, write code where it helps, and keep tight control over how workflows run. That flexibility comes with more setup and maintenance, which is a poor trade for a small team that just wants lead research done this week.
Gumloop’s angle is AI-first flows without code: scraping and web research plus language-model steps, such as extraction, summarising, and scoring, inside the same agent or flow. That makes it a stronger fit than a general automation tool when the hard part is not moving a row from one app to another, but understanding what is on a page and deciding whether a lead is worth pursuing. The trade-off is the credit model described above, and the need to check AI output before acting on it.
FAQ
Is there a free way to try Gumloop?
There is a free trial, but no free plan was shown on the pricing page checked on October 26, 2026. Every new account starts with a 14-day free trial of Pro, which requires a card, is a one-time offer per customer, and rolls into a paid Pro subscription unless you cancel during the trial.
Do you need to code to use Gumloop?
No. Gumloop is a visual builder and agent platform: you describe the task, connect the tools the agent may use, and set triggers or a schedule. Understanding the task is the main prerequisite, not writing code, although advanced users can still use the API and developer tools if they want to.
Is Gumloop good for lead research?
Yes, for research and enrichment patterns, with human review of outputs. Its own credits documentation uses researching a sales lead as its worked example, and its site lists lead generation and lead qualification among its use cases. The workflow in this tutorial follows that same pattern: collect sources, extract fixed fields, score against your criteria, and review a sample before using the results.
Verdict
Gumloop is a strong choice for a small team that spends too much time on repetitive research and wants an AI-first way to automate it without hiring a developer. The Pro trial makes it easy to test a real batch, and the 20,000 monthly credits are a generous starting point, but the variable credit cost and the need for manual checking mean it is not a set-and-forget tool. Start with one workflow, measure its real cost, and only then build more.
Early verdict: 8/10 for small teams automating repetitive research and content tasks.
Start with Gumloop, and for how we evaluate tools, see our Review Methodology. If you are comparing AI business tools, also read our Crevio review.