How to Qualify B2B Accounts: Criteria and Steps

Most outbound problems start before the first email. A list holds companies that look right on paper, and nobody checked whether they fit, whether anything suggests they are in a buying moment, or whether the data about them is any good. Account qualification is the step that answers those three questions for each company before anyone writes to a person.

This guide covers how to set criteria, how to gather evidence for them, how account qualification differs from lead qualification, and what to check before a qualified account turns into outreach.

Who this is for

This guide is for founders, sales leads and marketers who run outbound or account-based work and have a list of companies they do not trust yet. It suits you if:

You sell to other businesses and choose targets by company, not by individual sign-up.
You have some closed deals, won or lost, in a CRM or a spreadsheet. A handful is enough to start, as long as you read the results with care (see step 2).
You want a written definition of a "good account" that a second person could apply and get the same answer.

If you have no past deals at all, start with the free assessment described in step 1 and treat everything it returns as a hypothesis.

Lead qualification and account qualification are different jobs

The two get mixed up because they often happen in one tool. They answer different questions.

Account qualification asks whether a company is worth pursuing. The unit is the organization. The inputs are things like industry, size, the countries it operates in, the tools it uses and what it publishes on its own site. The output is a decision about the company: pursue, hold or drop.

Lead qualification asks whether a specific person is worth contacting, and whether now is the time. The unit is a person. The inputs are role, seniority, contact data and whether they have replied or engaged. Scoring individuals this way is what most teams mean by lead scoring.

Run account qualification first. A well-scored person at the wrong company is still the wrong target, and a decision about the company gives every later step a reason to exist.

Some tools in the stack already work at one level or the other. Website visitor tracking in Outlevo shows which companies visit, which pages they read and whether they fit your ICP, and it never identifies a person [[1]](#sources). LinkedIn Ads audiences are built from company names and websites, not from people [[2]](#sources). Contact data and verified emails belong to the lead stage. Keep the two stages separate in your records as well, so a company's status does not change just because one contact went quiet.

Steps

1. Write down who you already sell to

Criteria need a source. The most reliable one is your own customer base. Before you set rules, write a short description of the companies you win: what they do, how large they are, what they use, and why they bought. The guide on how to define an ideal customer profile walks through this in detail.

If you would rather start from your website than from memory, the free ICP assessment reads up to 5 public pages of a site [[3]](#sources). It returns facts labelled as verified observations or AI inferences, and up to 5 experimental segment hypotheses [[3]](#sources). Treat the experimental segments as questions to test, not as criteria.

You decide which countries you sell into. Outlevo asks which countries a customer sells into and never proposes markets [[4]](#sources). Do the same in your own process: state the countries as a business decision, then qualify within it.

2. Count companies, not deals

Past deals are the best evidence of what a good account looks like. Counting them correctly matters. Win/loss analysis in Outlevo counts companies, not deals [[4]](#sources). One company with five renewals or five reopened opportunities is still one company, and counting it five times makes a narrow pattern look broad.

Sample size matters too. Fewer than 10 won companies is labelled very low confidence [[4]](#sources). With a sample that small, use the pattern to shape a hypothesis and check it against new accounts. Do not turn it into a hard rule that excludes everything else. The guide on running a win/loss analysis on few deals covers this case.

You can import past deals from a CSV or your CRM and get a win/loss analysis that proposes ICP hypotheses [[5]](#sources). You can bring the history in three ways:

A CSV import takes up to 200,000 rows and maps your columns to domain, name, country, size, industry and deal fields such as stage, amount, outcome and loss reason [[6]](#sources).
With HubSpot, Outlevo learns your ideal customer from the deals you won and lost [[7]](#sources). It reads company name, domain, owner and lifecycle stage, plus deal pipeline, stage, amount and closed-lost reason [[7]](#sources).
With Salesforce, Outlevo reads accounts, opportunities and their stages for the ICP and win/loss analysis [[8]](#sources). This integration is a pilot, and the connection needs the Outlevo package in your org, which Outlevo installs with you [[8]](#sources).

The closed-lost reason is worth a look on its own. Lost deals often show a criterion you were missing, such as a company type that keeps saying no for the same reason.

3. Turn the pattern into three kinds of criteria

A usable qualification has three parts, and each should be scored separately. These are the same three scores a Readiness Card carries: fit, intent and data quality [[9]](#sources).

Fit is whether the company looks like the ones you win. Write fit criteria as plain, checkable statements. Examples of fields you can use are industry, company size and country, which are the company fields a CSV import maps [[6]](#sources). Add the things only you know: a tool they must run, a team they must have, a problem their site describes. Separate must-haves from nice-to-haves. A must-have that fails drops the account. A nice-to-have only moves the score.

Intent is whether something suggests the company is in a buying moment. Typical evidence is a visit to your pricing page, a hiring post, a change on a competitor's page, or a reply from a past contact. Intent decays, so record when you saw it. A signal from last quarter says little about this week.

Data quality is how much you trust what you know. An account can fit perfectly on a guess and still be a poor target to act on. Scoring data quality separately stops a thin record from looking like a strong one.

Keep the number of criteria small. Five to eight that you check every time beat thirty that nobody applies. If two people disagree on whether an account meets a criterion, rewrite the criterion until they agree.

4. Require dated, sourced observations

A score without evidence is an opinion. Set a minimum before any account counts as qualified. In Outlevo, a Readiness Card needs at least three observations, each with a source and a date [[9]](#sources). You can use the same rule by hand: no account is marked qualified until three observations are written down, each with where it came from and when you saw it.

When a fact comes from a website, keep the exact words. Facts read from a website are kept only when their quote is found on the page [[9]](#sources). A quote makes a fact checkable. "Mentions SOC 2 on its security page" is a claim someone can check on the page. "Seems security-minded" is not.

5. Keep facts and guesses apart

Qualification often mixes what you saw with what you assume. Label them differently. Outlevo labels guesses as hypotheses and never presents them as findings [[9]](#sources). AI-proposed ICP segments are labelled experimental until a person approves them [[4]](#sources).

In a spreadsheet, a simple status column does the job: "observed" for things with a source and a date, "hypothesis" for the rest. A hypothesis can still help you decide, but it should never be the only reason an account passes.

6. Add signals, and decide what counts

Fit tells you who to watch. Signals tell you when to act. Pick a short list of signals that matter for your product and decide in advance what each one does to the score.

Website visits are one source. A single script on your site shows which companies visit, which pages they read and whether they fit your ICP [[1]](#sources). Browsers that send Do Not Track or Global Privacy Control, and automated browsers, send nothing at all [[1]](#sources). Treat the numbers as a floor, not a full count.

Competitor activity is another. Outlevo's competitor tracking keeps quoted facts from competitors' public pages, changes, win/loss by competitor and competitor signals on accounts [[10]](#sources). For a wider view of the category, see competitor tracking software for B2B.

7. Set thresholds and have a person approve them

Decide what each outcome means before you score a single account. A simple version has three outcomes:

Pursue: all must-haves pass, fit is strong, and the record has enough dated evidence.
Watch: fit is good but intent or data is thin. Check again after a set interval.
Drop: a must-have fails.

Have one named person approve the rules. This mirrors how Outlevo handles outreach: nothing is sent outside a policy that a person approved, and policies set who can be contacted, on which channels, how often and with what budget [[11]](#sources). Qualification rules deserve the same ownership, because they decide who ends up in front of a prospect.

8. Write the result where reps will see it

A qualification that lives in one document gets ignored. Put the status on the company record in your CRM.

With HubSpot, Outlevo writes status, fit and intent scores, ICP version, reason, next action, due date and a Readiness Card link to its own properties [[7]](#sources). Standard fields are never edited, and nothing is written until an owner or admin approves the field mapping [[7]](#sources).

With Salesforce, Outlevo writes only to its own Account fields, a Readiness Card object and Tasks [[8]](#sources). A rep's Accept or Reject on the Account is read by the 15-minute sync [[8]](#sources). That feedback is useful: every rejected account is a data point about where your criteria are wrong.

Whatever tool you use, record the reason next to the status. "Dropped: no sales team, per careers page, seen March 3" teaches the next person more than a bare "Dropped".

9. Move to people only after the account passes

Once an account passes, move to the lead stage: find the right contacts, verify their addresses, and decide the channel. The enrichment entry covers how contact data gets added to a company record.

What to check

Before you treat a list as qualified, run through these checks.

Evidence

Does every qualified account have at least three observations, each with a source and a date? [[9]](#sources)
Are website facts backed by a quote from the page? [[9]](#sources)
Are hypotheses labelled as hypotheses, and absent from the reasons an account passed?

Criteria

Could a second person apply your criteria and reach the same result on ten accounts?
Is each criterion tied to something in your won or lost deals?
Did you count companies, not deals, and note when the sample is small? Under 10 won companies is very low confidence [[4]](#sources).

Process

Does a named person approve the rules and any change to them?
Is the reason for each decision stored next to the status?
Do rep decisions (accept or reject) flow back to the criteria?

Contact rules

A qualified account is a decision about a company. It is not permission to contact anyone there. Check the rules for each place you send to and each channel you use. Three reference points, each checked on 2026-10-22:

On US email, the FTC states: "The law makes no exception for business-to-business email." [[12]](#sources)
On legitimate interest, GDPR Article 6(1)(f) covers cases where "processing is necessary for the purposes of the legitimate interests pursued by the controller or by a third party". [[13]](#sources)
On corporate subscribers in the UK, the ICO says: "You can email or text any corporate body (a company, Scottish partnership, limited liability partnership or government body)." [[14]](#sources)

These are starting points, not legal advice. Ask a lawyer about your own situation.

Security of the data you hold

Qualification records include names, emails and notes. Outlevo protects every customer table with row-level security in the database, and encrypts personal data such as email addresses at the field level [[15]](#sources). Staff support access is read-only, time-limited and logged [[15]](#sources). If you keep the same records in your own sheets, apply similar care to who can open them.

What it costs to qualify in Outlevo

Costs depend on how deep you go, and the credit prices differ by step [[16]](#sources):

Light qualification costs 1 credit [[16]](#sources).
Qualification costs 3 credits [[16]](#sources).
A signal check costs 1 credit [[16]](#sources).
Deep research costs 30 credits [[16]](#sources).

Plans: Free is USD 0 per month, Builder is USD 29 per month with 320 credits a month, Growth is USD 149 per month with 1650 credits a month, and Managed is priced per contract [[16]](#sources). Connecting and syncing a CRM uses no credits, and neither does importing or exporting a CSV [[7]](#sources) [[8]](#sources) [[6]](#sources).

A practical pattern is to run the cheaper light check across the whole list, then spend on the deeper checks only for accounts that pass. That keeps the heavy research for the companies that already look right.

If you build on the API, metered calls quote their credit cost first and run only when confirmed [[17]](#sources). Test keys run against a sandbox with fake companies and contacts, and are not charged [[17]](#sources). That lets you try a qualification flow in code before you spend anything.

Common mistakes

Qualifying on fit alone. A perfect-fit company with no sign of timing may be worth watching, not contacting.
Treating one big customer as a pattern. One company is one data point, however large the deal.
Letting criteria grow. Each extra rule makes the process slower and the results less consistent.
Skipping the reason. Without a recorded reason, nobody can tell later whether a rule was wrong or the account changed.
Merging account and lead status. A company that fits is still a fit when one contact bounces.

Next step

Pick one segment and a short list of accounts from your current list. Write your must-have and nice-to-have criteria, then qualify each account by hand, with at least three observations that each carry a source and a date [[9]](#sources). Mark each as pursue, watch or drop, and write the reason. Compare your picks with what your reps would choose, and fix the criteria where you disagree.

If you want a starting hypothesis from your own site, run the free ICP assessment. To see how the scoring, evidence and CRM write-back fit together, read about account qualification and how Outlevo works, or look at a Readiness Card.