How do Credits work across storage, syncing, and profile enrichment?

Credits are the unit Bread & Butter uses to pay for the work that happens after someone becomes an Engaged Lead. Capturing a visitor, tracking their journey and scoring them never uses credits. What costs credits is keeping their profile once your plan's allowance is full, pushing their Intelligence Brief into your CRM, the research that turns an email address into that Brief, answering the questions you ask about your data in your AI assistant, and going out to find lookalikes: net-new people who look like your best customers.

How Bread & Butter credits work: visits and scoring never use credits; research on each new Engaged Lead is where most credits go; a fixed monthly fee with no surprise charges, and research waits until you choose to add credits.

In short: Credits pay for the work Bread & Butter does after someone becomes an Engaged Lead; capturing, tracking and scoring visitors never uses credits. The Pro plan includes 3,000 credits a month and Pro Plus includes 10,000, and each add-on block brings another 10,000. They pay for five things: storing profiles beyond your plan's allowance, sending Intelligence Briefs to your CRM, researching each lead, answering questions in your AI assistant, and finding lookalikes. Researching new leads is usually the large majority of it. Your plan is a fixed monthly fee, so using more never adds a surprise charge: when credits run low, capture never stops, research waits in a queue until you choose to add credits, and the rest resumes as soon as you do.

What a Credit Is, and Where They Come From

A credit is a fixed unit of work. Each paid plan includes a set number of credits a month, renewed at the start of each billing period:

  • Pro: 3,000 credits a month, and storage for 5,000 profiles.
  • Pro Plus: 10,000 credits a month, storage for 10,000 profiles, your AI assistant, connected to Claude or ChatGPT, for questions and reports, and Lookalikes.

If you need more, an add-on block gives you another 10,000 credits a month, and you can add as many blocks as your traffic needs. Credits added part way through a billing period go to work immediately, starting with anything waiting in the queue. Unused credits don't carry over when your credits reset. Pricing can vary by plan and by partner, and the current prices are shown in the app when you add credits.

The Free plan includes no credits, and credits can't be added to it. Profile enrichment, Intelligence Brief syncing, storage beyond the included profiles and everything else in this article are paid features. On Free you get unlimited traffic and visitors, 200 stored profiles, and the basic contact sync; to go past 200 profiles, you upgrade to Pro.

Where to see it. Open Account details, then the Usage tab. It shows the credits used so far this billing period, what's left, and every credit by what it paid for. Each captured lead also shows what its own research cost, so you can always see which profile cost what and why.

The Five Things Credits Pay For


How it is metered What it costs
Storing profiles Flat, per profile beyond your allowance, every month Pro includes 5,000 profiles and Pro Plus includes 10,000, then 2 credits a month for each additional profile. Free includes 200 with no way to add more; upgrading lifts the ceiling.
Syncing to your CRM Flat, per Intelligence Brief generated The basic contact sync (first name, last name, email address, profile link) never costs credits; which destinations a plan can send it to is set by the plan. A full Intelligence Brief costs 3 credits each time one is generated, and one charge covers every destination it goes to. The Intelligence Stream is part of an Enterprise license.
Profile enrichment Variable, by research effort On average 25 to 250 credits per profile, depending on how hard the person is to find and whether there is a company to research. The rest of this article explains why.
Questions in your AI assistant (MCP) Variable, by the size of the answer A quarter of a credit per question, plus 1 credit for every 100 rows in the answer. A question about one person costs 9 credits, because it brings back their full profile. Questions that reach further back in time cost more. Pro Plus.
Lookalikes Flat per lookalike, plus research 100 credits to source each lookalike and prepare its first email, plus the ordinary enrichment cost of researching that person and their company. Pro Plus.

Storage and syncing are entirely predictable. If you know how many profiles you hold and how many Briefs you generate, you know what they will cost. Enrichment is the one worth understanding, because it is usually the large majority of what an account uses. AI assistant questions cost what you ask for, and finding lookalikes is the one you can cap or switch off.

Storage

Nothing is deleted or trimmed when you pass your plan's allowance. Every profile is kept and your history stays complete; storage simply starts drawing credits.

Storage is charged every month, not once. Each profile beyond your allowance costs 2 credits a month for as long as it is stored, and the count is recalculated each month. A Pro account holding 6,000 profiles is 1,000 over, so storage uses 2,000 credits that month, and the same again the next month if the count hasn't changed.

A few things to know:

  • Only profiles with an email address count. Visitors who haven't converted are tracked and scored, but they aren't stored profiles and never count toward your allowance.
  • Add-on blocks add credits, not storage. Each block gives you 10,000 more credits a month. It doesn't raise the profiles your plan includes; it pays for the profiles above it.
  • Storage is paid first. When credits run low, storage is the first thing they go to, because nothing else works without the profiles.

On a high-volume account storage can become a meaningful share of monthly usage, which is worth knowing when you see a recommendation to add credits.

Why profiles can't be deleted

Once storage starts costing credits, deleting old or poor leads can look like an easy saving. Bread & Butter doesn't allow it, because every profile is part of what your Ideal Customer Profile learns from.

The good profiles show what a buyer looks like. The bad ones show what a buyer doesn't look like: the leads that went nowhere, the wrong industry, the contacts you ruled out. Take the bad ones away and the model only sees half the picture. Traits your non-buyers share start to look neutral, Focused Score drifts toward the wrong visitors, and the lookalikes we find start to resemble leads you already rejected.

Keeping everything, good and bad, is what lets the ICP keep sharpening over time. If a lead is wrong for you, move it to the right stage instead: Dead / Do Not Contact for a real person who is disqualified, Spam / Bad Email for a fake one. It leaves your Focused Tab and Nurture queue, it's never contacted again, and the model keeps learning from it. How do Lead Stages work? has the detail.

Syncing

Every conversion sends the basic contact record with no credits involved: first name, last name, email address and a link back to the profile. That is the List tier, and it never draws on your balance at any destination. Which destinations you can send it to is a matter of plan rather than credits: on Free that is CSV export, Google Sheets, Mailchimp and Slack, while sending the List into a CRM is a paid feature. Integrations and Automation lists every destination and what each plan can reach.

The Intelligence Brief is different. Generating one and pushing it into the CRM record costs 3 credits per Brief. The charge is on generation, not on delivery, so a Brief that goes to your CRM, a Slack channel and an inbox on the same session close is one charge, not three. When a known lead returns and their Brief is refreshed, that refresh is a new generation and a new charge.

Questions in your AI assistant (MCP)

On Pro Plus, with the Bread & Butter MCP server (https://mcp.breadbutter.io) connected to Claude or ChatGPT (see Bread & Butter MCP server connections for how to connect), you can ask anything about your own data in plain language and get the answer straight from your account: which leads converted today, which sources brought the most Focused visitors last quarter, how a campaign compares with last year. Each question draws credits based on how much it asks of your data, because a quick look at today and a search through years of history are very different amounts of work.

  • A quarter of a credit per question, however small.
  • Plus 1 credit for every 100 rows in the answer.
  • A question about one person costs 9 credits, because it brings back their full profile: research, journey and Brief.
  • History costs more. Questions about the last 90 days are charged at the normal rate, questions reaching back up to a year at twice the rate, and anything older than a year at five times the rate.

Some examples:

  • "Show me the 10 leads that converted today" is a quarter of a credit plus a tenth for the 10 rows: 0.35 credits.
  • "What were my top 25 referrers in the last 90 days?" is a quarter plus a quarter: 0.5 credits.
  • "What was my top referrer on every Tuesday for the last six years?" returns 312 rows reaching more than a year back, so it is charged at five times the rate: about 17 credits.

Fractions add up as you go, and your balance goes down a whole credit at a time. The Usage tab shows the exact cost of each question.

The ready-made reports you run in your AI assistant work the same way, and they cost little: running the five main reports every week comes to about 60 credits a month. A report covering today costs less than the same report covering the last five years. Report requests made through the Bread & Butter API are charged the same way too. Reports Bread & Butter builds for you from the Reports page show their cost on each report before they run, and a scheduled run is skipped and noted in History if your balance is too low.

Downloading a report as a CSV file never costs credits, on any plan.

How this compares. At the add-on block rate, an AI assistant question costs less than a fortieth of what HubSpot charges for its cheapest AI action.

To keep answers fast for every customer, each account has limits: by default, a single answer returns up to 500 rows, and questions can use up to 1,000 credits an hour. If a question hits a limit, your assistant says so. The AI assistant is part of the Pro Plus plan; it isn't available on Pro or during the trial.

Nurture emails

The first email Nurture drafts for each new lead is written from the research, so it is part of the research. It is never charged separately.

Lookalikes

Lookalikes are the outbound side of the product. Bread & Butter learns your Ideal Customer Profile from your own pipeline and goes out to find net-new people who match it, people who have never visited your site. Each arrives with a Lookalike Score showing how closely they resemble your best customers. Lookalikes are part of the Pro Plus plan, because they need enough researched leads to understand your Ideal Customer Profile; they aren't available on Pro or during the trial.

Each lookalike costs 100 credits to source and to prepare the first email that travels with them. The research on that person and their company is then charged as ordinary Profile Enrichment, at the variable rates explained below.

The research is not an optional extra on top of the lookalike. It is how we know the lookalike is worth sending you.

A name and an email address out of a database proves nothing. The research is what establishes the person's actual role, the company they work for, its industry and size, and whether any of that really matches the profile of your best customers. Without it we would be handing you a contact and inviting you to find out for yourself, which is the thing this feature exists to replace.

So a lookalike is never delivered unresearched. If the research can't confirm the match, the lookalike doesn't reach your account and you aren't charged for it. The charge is per lookalike that arrives, fully researched, with its first email ready.

Lookalikes arrive a few at a time rather than as a bulk import, which keeps the spend steady and visible rather than landing as one large bill. To control the spend, set a monthly cap on lookalikes, or switch them off, in Settings under Lead Gen Autopilot.

What Enrichment Actually Researches

To see why enrichment can't be a flat rate, it helps to know what the research is. Profile Enrichment doesn't stop at finding a social media link. It builds a complete picture for your sales team:

  • Individual context: the person's name, current role, recent promotions, publications, and public touchpoints such as skills, hobbies and volunteer causes.
  • Company intelligence: the organization's market positioning, core offerings, recent launches, funding, corporate donations and key leadership for multi-threading.
  • Timely sales hooks: recent press, news and awards, for the person or the company, to give your reps a real opener.
  • Journey synthesis: connecting that external footprint to what the person did on your site, so the Brief reads as one story rather than a profile stapled to a pageview list.

Why Enrichment Costs Vary

Some people have a wide public footprint and some don't, so the compute needed to research them varies. It works like a taxi meter: the harder the engine has to work, the more credits are used. There are three cases.

1. Rapid matches: the fewest credits

If a person has an easily accessible footprint, the profile comes together almost immediately. Minimal effort, very few credits.

2. Deep searches: more credits

If a person is hard to find, the engine doesn't give up. It queries alternate databases, local records, news sites and professional registries until it has a confident match. The more sources it has to work through, the more credits it uses.

3. Ghosts: the most credits

Occasionally a lead submits an email address with no digital footprint at all, or the data is so conflicting, thousands of people with the same name, say, that no single person can be confirmed. These empty profiles often cost the most, which surprises people.

Proving a negative is expensive. To establish that a lead doesn't exist online, or that the data is too ambiguous to trust, the engine has to exhaust every available source, which can take up to 30 minutes of continuous processing. We prioritize accuracy over filling fields, so the engine will never guess to make a profile look complete. And an empty, verified result is still useful: it means your reps don't spend 30 minutes of their own day hunting the same ghost.

Why Corporate Email Domains Cost More, and Why That Is Good News

Across an account, this is usually the biggest factor in what enrichment costs, and the expensive outcome is also the better one.

A corporate domain means the company research always happens. A personal email address only sometimes leads to it.

The engine always tries to research two subjects: the person, and the organization they work for. With a Gmail, Outlook or Yahoo address, the domain gives us nothing to start from. Often the person's public footprint still tells us where they work, and when it does we research that company exactly as we would from a work address: size, industry, positioning, funding, launches, press and leadership, connected back to the person and their role. But sometimes the footprint doesn't lead to an employer, and the research stops at the individual.

With a corporate domain there is no maybe. The domain names the organization, so the company research always happens. That is why, generally speaking, an account capturing mostly corporate domains spends more per lead than one capturing mostly personal addresses at the same volume: not because a personal address is always cheaper, but because a corporate one is always the full job.

What the extra spend buys. A richer profile, yes, but more importantly the raw material your Ideal Customer Profile is built from. An ICP is a description of the kind of organization that buys from you: size, industry, growth stage, geography, buying triggers. A profile with no company attached can tell Focused Score how the person behaved on your site, which matters, but it can't tell the model what kind of business they came from. Every capture that resolves to a company adds a data point to the picture of who your buyers are, and when you mark a deal Won/Closed the model learns from the company as well as the journey.

This compounds. More captures with a company behind them means better firmographic coverage, a sharper ICP, more accurate scoring, and a sales team that prioritizes better. The cost is higher at the front and the return is a model that gets more useful every month.

It is also a signal on its own. A high share of corporate domains usually means your marketing is reaching people at work, evaluating you on company time under their real work email. That is what a healthy B2B funnel looks like. A mix that is heavily personal addresses is worth a look for its own sake: it can mean your content is attracting the wrong audience, or that people aren't yet willing to give you their work address.

A higher enrichment bill driven by corporate domains is not a problem to solve. It is what a good funnel costs.

Reading Your Usage

Because the cost per search varies, we show you exactly where credits went. On each captured lead you will see the credit cost of that search alongside a status badge explaining it:

  • Profile Found (Low Credits): rapid match completed.
  • Profile Found (High Credits): deep search was needed to locate the footprint and sales hooks.
  • No Footprint Verified (Highest Credits): exhaustive search completed; no footprint exists, or the data was too ambiguous to confirm.

For the whole account, the Usage tab in Account details shows every credit this billing period by what it paid for.

Why Your Monthly Total Stays Predictable

No single lead's enrichment cost is knowable in advance, but most businesses see a consistent mix of easy and hard matches, and a consistent mix of corporate and personal email domains, because their traffic and audience don't change dramatically week to week. After a billing period or two that mix averages into a stable number. Your total enrichment spend becomes a reliable line to plan around even though the per-lead cost never is.

The cost moves with your marketing: how many people your campaigns bring in, how many of them become Engaged Leads, and how many come back. A launch month uses more, a quiet month less.

Your first 30 days give you an observed baseline rather than a guess. By the end of your trial you will know roughly how many profiles you add and what your mix looks like, and can decide from real numbers which plan fits and whether to keep everything running.

When Credits Run Low

Your plan is a fixed monthly fee with a set number of credits each month. Using more never adds a surprise charge to your card. Bread & Butter tells you ahead of time when you're about to run out, and adding credits is always your decision.

Credits are always spent in the same order, on every account:

  1. Storing captured profiles. Nothing else is possible without them.
  2. Syncing Intelligence Briefs to your CRM.
  3. Profile Enrichment.
  4. AI assistant questions (MCP).
  5. Finding lookalikes, with whatever is left.

AI assistant questions come after the research, so asking questions about your data never slows down capture, syncing or the research on new leads. When there are no credits left for a question, your assistant tells you and links you to adding credits.

Finding lookalikes is last deliberately. It is the one thing here that depends on everything above it: it can only find people who resemble your best customers once the research on those customers is complete. Sourcing new lookalikes against a half-finished profile would spend credits to produce worse targets. Everything above it keeps your own pipeline working, so if something has to stop, this is the right thing to stop first.

This order is the same on every account and is not something you configure.

What keeps running. Everything that doesn't cost credits. Visitors are still tracked, forms still convert, every step of every journey is still recorded, and every integration keeps sending new leads. An integration set to the Intelligence Brief or Intelligence Stream keeps sending the same Brief or Stream, with the lead's name and email address filled in, the researched details left empty, and a one-line notice to add Bread & Butter credits in their place. Your CRM stays current and nothing downstream breaks. Focused Score and your Focused Ratio keep working, though while enrichment is paused they lean on on-site behavior alone rather than behavior plus profile fit, so ICP matching is less precise than usual. You still see who is engaging most; you see less of who resembles your best customers.

What pauses. The research, AI assistant questions, and finding lookalikes.

Queued research runs within your billing period. Research that couldn't run is queued rather than skipped. Add credits before your billing period ends and the engine goes back through the whole queue, so your ICP catches up rather than having a gap in it. When your billing period ends, anything still queued is dropped, and your renewed credits go to the new period's leads.

The add-on recommendation is for a full month. When you run low, Bread & Butter recommends how many credits you need a month based on your recent usage, sized for a full billing period so the same thing doesn't happen next month. If research is waiting in the queue, an add-on bought part way through a period is charged and granted in full rather than prorated, so there are enough credits to clear the queue and keep researching new leads for the rest of the period. The credits go to work straight away, starting with the queue.

AI assistant questions and finding lookalikes don't queue. A question that couldn't run can simply be asked again once credits are added. Lookalikes stop being sourced until enrichment is running again, because a lookalike found against a stale profile is not worth having.

A Note on Accuracy

The research engine is powerful, but it is not flawless. Because it synthesizes live, unstructured data from across the open web, it can occasionally hit outdated source information or misread context. We recommend your reps give the generated insights a quick read before sending anything, the same way they would check any research done on their behalf. The Brief is designed to speed up the work, not to replace a human glance before the send button.

Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.