Lead Generation Virtual Assistant: How to Get Leads You Can Actually Use, Not Just Longer Lists

Here's the most common way a lead generation virtual assistant disappoints. The spreadsheet's full. Companies, names, email addresses. Your assistant has clearly worked hard.

Then you open the first record.

You can't tell why this business might need your service. You don't know if the contact still works there. And you've no idea whether they already asked you not to contact them.

That's the trap with a lead generation virtual assistant. A list can be produced quickly without producing any useful sales evidence. If the person who uses the list has to research every entry again, you've moved the data collection, not the decision about who's worth contacting.

This guide shows you how to set up research that's genuinely usable: clear criteria, evidence attached to every claim, sensible stopping rules, a clean CRM and respect for people who've said no.

Here, a virtual assistant (VA) is an independent freelancer or agency providing a research and record-keeping service.

Decide what decision the research should help you make

Write down the action you want to take after reading a finished record:

• Is this company a fit for our partner programme?
• Has this customer expanded?
• Does this prospect match our specific service?

Avoid starting with "50 names a day." A volume target doesn't tell anyone which facts matter.

If you sell a service for businesses with several locations, then the number and type of locations matters far more than a generic industry label or an impressive-sounding job title.

Describe the problem your service solves and the visible signs worth investigating. But keep those signs separate from assuming the prospect has the problem.

A company opening a new location might have new coordination needs. You don't know their internal situation from the announcement alone.

Be clear about what research can't tell you. Public sources rarely prove budget, buying authority, urgency or willingness to talk. Those usually need a conversation. Don't label a record "qualified" just because their website looks right.

Answer these before collection starts:

• Should the VA contact people?
• Is a published email address permission to market?
• How fresh does a source need to be?
• What happens to uncertain records?

Otherwise a large database fills up with assumptions.

Turn "ideal customer" into things you can actually check

An ideal customer profile is just a description of businesses likely to fit your offer. For research, turn it into checkable fields:

• business type,
• service area,
• operating model,
• current activity,
• clear exclusions.

Then split them:

• Must-have criteria. A company outside your service area is excluded straight away.
• Helpful signals. A recent expansion is interesting but not essential.

If everything is a must-have, your VA wastes hours hunting for facts that are rarely public.

Give each criterion an accepted source and a rule for uncertainty:

• The company's own locations page can show where it operates.
• A dated press release can confirm an announcement.
• A third-party directory can suggest a candidate, but shouldn't settle a current fact.

Use three simple outcomes: fits, doesn't fit, needs review.

"Unknown" should never silently become "yes."

Include exclusions that protect relationships: current customers, active negotiations, competitors, unsuitable sectors and anyone on your suppression list. A shiny new list that ignores your existing database creates embarrassing duplicate approaches.

And don't ask your VA to guess sensitive personal characteristics or collect unrelated personal details to make records look "complete."

Every claim in the record needs a trail back to its source

For each important fact, record:

• the fact,
• the source URL,
• the date it was checked,
• what the source actually supports.

A confidence label can help, but it shouldn't replace evidence someone else can check.

Source date and check date are different. A page viewed today might contain an announcement from three years ago. Record both where relevant, otherwise old information looks freshly confirmed.

Keep facts and interpretations apart:

• Fact: "The company lists several locations."
• Interpretation: "The company needs our software."

A good record can include both, as long as the second is clearly labelled as a reason to investigate, not proven demand.

For UK companies, GOV.UK's company information service shows public details like registered addresses, officers and filings. Useful for identifying the legal entity. But it doesn't prove a contact's current role or that they're interested.

Don't paste long passages into your CRM (customer relationship management system). Store a short supported note and a link.

And give uncertain records their own review queue, so they don't slip into an outreach list as if approved.

A worked example: researching fitness studios

Here's a hypothetical example.

A business offers maintenance coordination to independent fitness studios. The owner wants operators with multiple local sites, because that setup might suit the service. The VA does research and CRM preparation, not outreach.

The brief requires: - an official website, - evidence of more than one site in the service area, - an identifiable business contact route.

It excludes: - franchises managed through a central national contract, - existing customers, - anyone on the suppression list.

Studio 1. Found through a directory. The VA reads the official locations page: several sites in the target area. They record the URL and check date. The exact legal entity gets a separate check because the trading name is ambiguous.

Studio 2. A news article says it plans to open a second site. Its website still lists one. The VA marks it needs review. A future plan isn't proof of the current setup.

Studio 3. Fits all the criteria, but already has an active conversation in the CRM. The VA adds the research note to the existing record and alerts the owner. They don't create a duplicate or add it to a new outreach sequence.

The owner reviews the batch and chooses which records deserve more investigation.

None is called a guaranteed opportunity. The deliverable is a supported shortlist that saves repeated research while keeping "good fit" separate from "actually interested."

Set a stopping rule so research doesn't become endless browsing

Define the minimum complete record and a reasonable limit. Your VA should know when there's enough evidence for your next decision, and when more searching won't help.

A stopping rule can be simple:

• A must-have exclusion is confirmed? Stop that candidate.
• Required evidence and contact route verified? Stop. Don't keep collecting decorative facts nobody will use.

For uncertain fields, name the next source to try, and when it goes back for review. Your VA shouldn't spend half a paid batch hunting a private fact that'll never be public.

Review rejected and uncertain records too, not just accepted ones. If nearly every candidate fails the same condition, your market definition or search method might be wrong. Raising the quota would just scale an unproductive search.

In the fitness example, the "planned second site" studio could become a separate expansion watch category. That's a deliberate decision, documented, not something your VA decides quietly.

And agree the cost of extra research. The original fee shouldn't silently stretch because you keep asking for more detail after delivery.

Clean the CRM without wiping out history

Before importing anything, look at your existing data. Company names vary. Contacts change email addresses. A trading name might differ from the legal name. Matching on one familiar word can merge two unrelated businesses.

Define:

• what counts as an update (a verified new website can be added),
• what needs review (a suspected duplicate needs human inspection of both histories).

Don't let your VA bulk merge or delete records just to make the database look tidy.

Use a staging sheet for the first batch. Include the proposed CRM action, the matching existing record, the new information and the evidence. Review it before import, especially if your CRM triggers automations when fields or stages change.

Read your CRM's current import documentation and test a small sample first. Check how it handles unique IDs, blank fields and linked records. A perfect spreadsheet can still create duplicates or wipe useful data if the mapping's wrong.

Keep meaningful history. A past objection, an open complaint or an existing conversation shouldn't disappear under a newly researched job title.

And avoid a private research spreadsheet that slowly becomes a second CRM with different statuses. Our tools guide covers keeping one authoritative record.

Finding contact details is not permission to email them

This is where a lot of small businesses get caught out.

Researching information and using it for outreach are separate decisions. A publicly visible email address isn't blanket permission for any campaign. Platform rules, privacy law and marketing rules can apply even to business contacts.

In the UK, the ICO's guidance on business-to-business marketing explains that the Privacy and Electronic Communications Regulations (PECR) treat corporate subscribers (like limited companies) differently from sole traders and some partnerships. It also explains that data protection law still applies when personal data is used in a business context.

So don't reduce it to "B2B email is always allowed." The channel, recipient type, data, legal basis, transparency and objections all matter.

The ICO also discusses publicly available business contact data and the need to follow data protection rules. Collecting a named work email address is already processing personal data, before anything is sent. Research needs a lawful, defined purpose too.

Some of the ICO's guidance is under review following legislative changes, so use the current version and proper advice for the actual campaign.

For other countries, check the local rules. Don't import the UK distinctions. Your VA should follow an approved research and outreach policy, not decide legal eligibility because an address appeared on a website.

If you later delegate sending, make that a separate workflow with its own approved messages, suppression checks, identity, channel limits and reply handling. Don't wire a freshly filled research list straight into automatic outreach.

A suppression list is your business's memory

A suppression list records contacts or businesses that shouldn't receive specified marketing.

It's different from deleting an uninteresting prospect. Without it, the same person can be rediscovered and contacted again in next month's research batch.

The ICO's business marketing guidance recommends screening new lists against suppression records and explains people's right to object to direct marketing. Keep enough information to honour the restriction, not an archive of everything ever collected.

Tell your VA:

• where suppression is checked,
• how new objections are recorded,
• how the restriction reaches every tool. If your CRM and email tool keep separate lists, a clean record in one doesn't protect the other.

Record the scope where it makes sense. A request to stop marketing emails shouldn't be read as permission to start marketing by phone. And operational customer messages have their own rules.

In the fitness example, a studio on the suppression list isn't added to outreach. The VA can note it was excluded under the process, without copying the whole objection history into an unrestricted sheet.

Give your VA enough to avoid prohibited contact, not every private conversation that led to the restriction.

Every CRM stage should be triggered by something that actually happened

Define each stage by evidence:

• Researched: must-have criteria checked.
• Approved for outreach: an authorised person accepted it under the campaign rules.
• Contacted: a permitted message was actually sent and logged.

"Engaged" shouldn't mean any reply. An out-of-office, a rejection or an unsubscribe isn't buying interest. Define what counts, and how other replies are classified.

A booked call isn't a qualified opportunity either. The call might reveal no need or no budget. Keep that sales decision with the person authorised to make it, unless your VA has specifically been engaged for qualification.

Give every active record a next action and an owner. A positive reply with no follow-up owner can sit forever while the dashboard looks great.

In the fitness example, the owner's first conversation establishes whether maintenance coordination is a real problem. Only then does the record move beyond "researched fit." The locations page supported the selection, not the claim of demand.

Our VA KPI guide explains how to measure research quality without rewarding raw list size.

Review a batch as if you were about to use it

Open a sampled record as though you're about to act on it. Can you:

• understand why it fits?
• verify the important facts?
• see what's uncertain?
• see the next permitted action?

...without repeating the whole search? That's the real acceptance test.

Check the rejected records too. A VA can make the accepted list look strong by throwing out difficult but valuable cases for the wrong reason.

When errors repeat, fix the underlying rule:

• Planned locations recorded as open? Improve the evidence definition.
• Current customers showing up as new leads? Fix the CRM matching step.

"Find better leads" fixes neither.

Agree how often facts get rechecked. A legal company ID changes rarely. A contact's job role changes often. Recheck time-sensitive details before use, rather than marking the whole record "verified" forever.

And keep data under your control: store research in business-owned systems, restrict exports, document approved tools and helpers, and arrange return or deletion when the service ends.

Your lead research brief

LEAD RESEARCH BRIEF

Decision this research supports:

Must-have criteria (and accepted source for each):

Helpful signals:

Exclusions (customers, active deals, competitors, suppression):

Outcomes: fits / doesn't fit / needs review

Record fields:
  Fact / source URL / source date / check date / what it supports
  Interpretation (labelled):
  Contact route:
  Next action and owner:

Stopping rule:

Next source to try when uncertain:

CRM: staging sheet first / matching rule / no bulk merges

Suppression check (where, and which tools):

Outreach: included? (Y/N) Approved policy (link):

Recheck schedule for time-sensitive facts:

The useful output isn't a spreadsheet proving someone was busy. It's records another person can act on with less repeated research and fewer false assumptions.

Outsourcing lead research works when the evidence, the uncertainty and the next decision all arrive together.

Related: how to give a VA safe access and how to delegate to a virtual assistant.

Frequently asked questions

What does a lead generation virtual assistant do?

They research businesses that match your criteria, verify key facts from reliable sources, record evidence and uncertainty, prepare or update CRM records and respect suppression lists. Outreach is a separate task you might or might not delegate.

How many leads should a VA research per day?

Avoid quotas at first. The right volume depends on how much evidence each record needs. Agree what a complete record looks like and a sensible stopping rule, then measure how long that takes.

Is it legal to email business contacts a VA found online?

It depends on the country, channel and recipient type. In the UK, rules differ for companies and for sole traders and some partnerships, and data protection law still applies. Get advice for your campaign rather than assuming business email is always allowed.

Should my VA have full access to my CRM?

Only to what the work needs. Use a staging sheet for first imports, avoid bulk merges or deletions, and check automations that might trigger when fields change.

What is a suppression list?

A record of contacts or businesses that shouldn't receive specified marketing. It stops someone who objected from being rediscovered and contacted again in the next research batch.

For a separate project correcting existing records, the CRM data cleanup guide explains field rules, duplicate review and controlled pilot updates.