Warm vs Cold Leads: Conversion Rate Multiplier, Benchmarks, and ROI

The warm vs cold lead conversion rate multiplier is a ratio: the conversion rate for a defined warm cohort divided by the conversion rate for a comparable cold cohort. If 12 of 100 eligible warm leads become customers and 2 of 100 eligible cold leads become customers, the rates are 12% and 2%. The multiplier is 12% ÷ 2% = 6x. That is a worked example, not a benchmark.

Is the multiplier universally 5x to 10x? No. A 5x to 10x result can appear when warm and cold cohorts are measured at the same stage, but it is not a defensible universal B2B constant. Published percentages often mix different starting populations and conversion events, such as sent cold emails, qualified inbound form fills, SQLs, opportunities, booked meetings, or closed-won deals.

This guide treats the multiplier as something to calculate, not something to assume. Compare warm and cold cohorts only when they use the same eligibility rules, conversion event, denominator, attribution window, maturity period, and cost basis. Then report the stage-specific multiplier alongside held meetings, qualified opportunities, mature wins, cost per outcome, and ROI.

Table of Contents

30 sections Show

Warm, cold, and hot leads in plain English

“Warm” is an operational state, not a standardized industry category. Two companies can use the same word for different levels of engagement. The safest approach is to keep source, temperature, qualification, consent, and lifecycle stage in separate CRM fields.

HubSpot’s lifecycle-stage documentation treats stages such as MQL, SQL, opportunity, and customer as configurable parts of an organization’s own process. Salesforce’s product-qualified lead guidance also makes an important distinction: product activity becomes useful qualification evidence only when it is combined with meaningful usage thresholds and relevant firmographic context.

Term	Working definition for this guide
Cold lead	A person or account that may fit the target profile but has no verified prior relationship, recent reliable engagement, referral, product signal, or declared interest relevant to the current outreach. Cold does not mean low quality.
Warm lead	A person or account with recent, credible context that makes outreach more relevant. Warm does not automatically mean qualified, ready to buy, contactable, or legally marketable.
Hot lead	A fit person or account showing explicit current evaluation or need, such as a demo request, active buying conversation, procurement step, confirmed timeline, or agreed next action.
Inbound lead	A person or organization that initiated an identifiable interaction with the seller or its owned property. Inbound describes direction, not fit or temperature.
Outbound lead	A person or account first approached by the seller in the relevant motion. Outbound can be warm when there is a referral, prior relationship, reactivation event, or other credible context.
Product-qualified lead	A user or account that has reached a documented product-usage threshold and meets required firmographic or use-case criteria.
MQL	A lead that meets the marketing team’s documented qualification criteria and is ready for the next marketing-to-sales step.
SQL	A lead accepted or qualified by sales under documented criteria for direct progression.
Opportunity	An active potential transaction represented as a deal under the organization’s CRM rules.

For a narrower treatment of temperature, see BrandJet’s guide to the warm lead vs hot lead difference.

How to calculate the warm vs cold lead conversion rate multiplier

Choose one conversion event and use it for both cohorts. Calculate each cohort’s conversion rate, then divide the warm rate by the cold rate.

Warm cohort conversion rate ÷ cold cohort conversion rate = conversion multiplier

Cohort	Eligible leads	Customers	Conversion rate
Warm	100	12	12%
Cold	100	2	2%

In this hypothetical example, 12% ÷ 2% produces a 6x multiplier. Change the conversion event and the result can change. A held-meeting multiplier, opportunity multiplier, and mature-win multiplier are separate measurements, even when they use the same leads.

Why do warm leads convert better than cold leads?

Warm cohorts often contain information that reduces uncertainty. A referral can transfer trust, product activity can show relevance, and a positive reply can confirm reachability and willingness to engage. A current buying conversation can reveal timing, role, and need.

Measurement can create part of the apparent advantage. Warm leads often enter the reported funnel later: a qualified form fill has passed filters that a raw cold account has not. Unless the starting points are aligned, the comparison favors the warmer cohort by design.

Selection bias matters too. People who choose to reply, attend an event, use a product, or request a demo are not a random sample of the target market. Their higher downstream conversion may reflect fit, urgency, brand familiarity, budget, role, timing, or prior demand creation. It does not prove that the act of labeling them “warm” caused the result.

The defensible inference is that credible context can improve relevance and reduce friction. The size of the improvement must be measured locally. It should not be borrowed from a generic percentage range or conversion multiplier.

Warm vs cold lead conversion multiplier: what the evidence shows

Dimension	Cold motion	Warm motion	What to measure
Starting context	Little or no verified recent context	Recent relationship, engagement, referral, product, or declared-intent evidence	Evidence type, timestamp, confidence, and source
Acquisition cost	Data, list building, enrichment, verification, infrastructure, research, seller labor, and compliance operations	Content, events, referrals, communities, product experience, paid acquisition, signal capture, routing, nurture, and seller response	Fully loaded cohort cost, not only media or data spend
Contactability	Depends heavily on data accuracy, channel reach, deliverability, and applicable rules	May have better contact details or permission, but not always	Valid records, delivered messages, completed calls, objections, and suppression
Speed	Usually prioritized by fit and account value	Explicit requests and positive replies usually justify a clear response SLA	Time from qualifying event to first useful response
Sequence	Often needs more education, trust building, and testing	Can use a shorter context-led sequence when the signal is strong	Attempts, elapsed time, channel mix, and stop rules
Personalization	Public company, role, market, and problem context	Volunteered or clearly known interaction context	Research time, response quality, and privacy risk
Sales cycle	Can begin before a buyer has defined a project	May begin after some education or evaluation has occurred	Median days by matched cohort and mature outcome
Conversion	Starts earlier and includes more unreachable or uninterested records	Often enters after engagement or qualification	Shared stage-to-stage rates only
Attribution	Outreach can receive first-touch credit even when brand or content assisted	Prior content, event, community, product, or partner costs can be hidden	Sourced, influenced, and finance-recognized views
Scale	Limited by reachable market, data quality, reputation, and sales capacity	Limited by demand creation, signal volume, routing, and response capacity	Marginal cost and performance as volume rises

A warm lead is not automatically cheaper. Warm demand can require substantial upstream investment. A cold motion is not automatically inefficient. It can be the rational way to reach a narrow group of strategic accounts that will not appear through inbound demand.

Is the 5x to 10x conversion multiplier universal?

How to read these benchmarks

Every rate below is labeled by cohort, period, channel, numerator, and denominator. Vendor datasets describe the vendor’s measured customer or client cohort, not the whole B2B market. All replies are kept separate from positive replies. Booked meetings are kept separate from held meetings. Qualified inbound form fills are not treated as website visitors. CRM stage rates are not multiplied across unrelated studies.

Disclosure: BrandJet sells outreach and intent-signal software. Vendor datasets are identified as such, and the worked ROI scenario uses hypothetical inputs. No cited rate is presented as a universal market average.

Before using any benchmark, answer these questions:

1. What entered the cohort: an email, person, account, inquiry, MQL, SQL, or opportunity?
2. What event counted as conversion?
3. Was the denominator attempted, sent, delivered, connected, qualified, booked, held, or mature?
4. What channel and source produced the cohort?
5. What company types, markets, roles, and contract values were included?
6. What period and attribution window were used?
7. Were duplicate people and accounts removed?
8. Were negative replies and opt-outs included in the reply count?
9. Were meetings booked or held?
10. Which acquisition and labor costs were included?

These labels prevent a common error: treating unlike percentages as if they sit on one universal funnel.

Warm vs cold conversion benchmarks by funnel stage

The following sources are useful because they disclose enough methodology to interpret a specific metric. None provides a universal warm-versus-cold lead conversion rate.

Sources checked July 14, 2026. These examples are separated by channel and stage because unlike percentages do not belong in one funnel.

Cold email reply evidence

Funnel transition	Cohort and period	Reported result	What it does not measure
Sent cold email to reply	Belkins client campaigns in 2025: 7,530,489 cold emails and 34,393 unique replies. Automated replies were excluded, bounces were discussed separately, and open tracking was disabled.	0.45% reply rate per sent email, according to the Belkins 2026 cold email study.	Positive replies, booked meetings, held meetings, opportunities, wins, or the whole market

Qualified inbound booking evidence

Funnel transition	Cohort and period	Reported result	What it does not measure
Qualified inbound form fill to booked meeting	More than one million B2B SaaS inbound form fills among RevenueHero customers during calendar 2025. Every observed company used RevenueHero’s qualification and instant-scheduling system, and RevenueHero says the aggregate is weighted by form volume.	Median 62%, top quartile 72%, and top decile 78% among qualified form fills, according to the RevenueHero 2025 inbound benchmark.	All visitors, all inquiries, held meetings, opportunities, customers, or companies not using the product

Self-reported CRM-stage evidence

Funnel transition	Cohort and period	Reported result	What it does not measure
Lead to SQL	Norwest survey of 177 sales and marketing leaders from VC- and PE-backed B2B companies, including North American and Israeli portfolio companies plus North American third-party respondents, fielded from July 21 to August 19, 2025. Results were self-reported.	Mean 37% and median 36% in the Norwest 2025 B2B benchmark report.	A standardized lead definition, observed CRM data, or a warm-versus-cold split
SQL to opportunity	Same Norwest survey and limitations.	Mean 41% and median 40% in the Norwest 2025 B2B benchmark report.	Comparable SQL or opportunity definitions across companies
Opportunity to proposal	Same Norwest survey and limitations.	Mean 43% and median 42% in the Norwest 2025 B2B benchmark report.	Opportunity-to-win, proposal acceptance, or source-specific performance
Proposal to win	Same Norwest survey and limitations.	Mean 47% and median 45% in the Norwest 2025 B2B benchmark report.	Opportunity-to-win across all stages, cohort maturity, or source-specific performance

The Norwest survey also shows why labels need local definitions. Only 25% of respondents described MQLs through lead scoring. Other respondents used explicit high intent, a combination of engagement and persona, any form or event response, or no MQL stage at all, as reported in the same Norwest methodology-rich survey.

Do not multiply results across these evidence sets. They come from different populations, systems, channels, and stage rules. This evidence does not provide a methodology-rich market benchmark for delivered-email-to-reply, reply-to-meeting, MQL-to-SQL, opportunity-to-win, or account-to-pipeline. Belkins reports replies per sent email, not per delivered email. Norwest’s lead-to-SQL rate is not an MQL-to-SQL rate, and its proposal-to-win rate is not an opportunity-to-win rate. Preserve those labels. Calculate account-to-pipeline from your own eligible, deduplicated account cohort.

Warm vs cold lead cost

The most useful cost comparison is not cost per raw lead. It is fully loaded cost per held meeting, qualified opportunity, and expected gross profit.

A cold motion can include contact data, list construction, enrichment, verification, sending infrastructure, telephone data, research, copy, seller time, deliverability monitoring, suppression, CRM tools, quality control, and compliance review. BrandJet’s B2B lead database can support data and enrichment work, but the cost model should still include the labor and systems required to turn data into an eligible cohort.

A warm motion can include content production, paid acquisition, events, communities, partnerships, referral incentives, product onboarding, customer success, social monitoring, signal capture, routing, scheduling, nurture, and sales response. These costs do not disappear because the final touch was inbound. BrandJet’s guide to generating warm B2B leads covers practical acquisition methods, while the ROI calculation below keeps their costs visible.

Allocate shared costs over a documented period. Use the same policy for both motions. If cold acquisition includes seller salaries but warm acquisition excludes content and event labor, the comparison is biased before any conversion is measured.

Warm vs cold lead ROI model with editable assumptions

Use the following variables:

Variable	Meaning
N	Eligible, deduplicated contacts or accounts
c	Contactable or delivered proportion
r	Positive-response proportion among contactable or delivered units
m	Held-meeting proportion among positive responses
q	Qualified-opportunity proportion among held meetings
w	Closed-won proportion among mature qualified opportunities
V	Average first-year contract value or expected deal value
g	Gross-margin proportion
C	Fully loaded acquisition and sales cost for the cohort

Use held meetings rather than booked meetings when attendance matters. Use mature opportunities for win-rate calculations.

Contactable or delivered units
= N * c

Positive responses
= N * c * r

Held meetings
= N * c * r * m

Qualified opportunities
= N * c * r * m * q

Expected wins
= N * c * r * m * q * w

Cost per held meeting
= C / held meetings

Cost per qualified opportunity
= C / qualified opportunities

Pipeline per 1,000 eligible units
= 1,000 * c * r * m * q * V

Expected gross profit before acquisition cost
= expected wins * V * g

Expected net contribution
= expected gross profit before acquisition cost - C

Expected ROI
= expected net contribution / C

Break-even opportunity win rate
= C / (qualified opportunities * V * g)

Break-even positive-response rate
= C / (N * c * m * q * w * V * g)

If your process begins at a later stage, start there. For example, a qualified inbound form-fill model should not invent an email-reply stage. An account-based model should replace people with eligible accounts and deduplicate all contacts within each account.

Hypothetical worked scenario

The following inputs are hypothetical assumptions, not benchmarks. Replace every value with your own mature cohort data.

Input	Hypothetical cold cohort	Hypothetical warm cohort
Eligible units	1,000	1,000
Contactable or delivered rate	90%	95%
Positive-response rate	3%	7%
Positive response to held meeting	35%	40%
Held meeting to qualified opportunity	45%	50%
Qualified opportunity win rate	20%	25%
Average contract value	$25,000	$25,000
Gross margin	75%	75%
Fully loaded cohort cost	$9,000	$18,000

Output	Hypothetical cold cohort	Hypothetical warm cohort
Positive responses	27.00	66.50
Held meetings	9.45	26.60
Qualified opportunities	4.25	13.30
Pipeline created	$106,312.50	$332,500.00
Expected wins	0.85	3.33
Expected gross profit	$15,946.88	$62,343.75
Cost per held meeting	$952.38	$676.69
Cost per qualified opportunity	$2,116.40	$1,353.38
Expected ROI	77.19%	246.35%
Break-even positive-response rate	1.69%	2.02%
Break-even opportunity win rate	11.29%	7.22%

In this illustration, the warm cohort costs twice as much to create and operate. It still has a lower cost per qualified opportunity because the assumed downstream stage rates are higher. That result is not guaranteed. Lower contract value, margin, contactability, meeting attendance, qualification, or win rate could reverse it.

The model makes the assumptions visible so finance, marketing, and sales can test the same economics.

A practical lead-temperature framework

Do not rely on a fixed score copied from another company. Start with three gates:

1. Identity and record integrity: Is the person or account resolved correctly and deduplicated?
2. Fit: Does it meet minimum account, role, geography, use-case, and commercial criteria?
3. Contactability and permission: Is the planned channel reachable, lawful, and consistent with consent, objections, and suppression?

Consent and contactability should not add warmth points. They are operating constraints.

After the gates, evaluate the evidence:

Dimension	Questions to record
Relationship	Is there a direct relationship, trusted introduction, customer history, partner connection, or prior sales conversation?
Recency	When did the strongest reliable signal occur, and is it still relevant to the buying cycle?
Source	Was the record referred, inbound, outbound, event-sourced, product-led, partner-sourced, or reactivated?
First-party behavior	Did the person submit a form, use the product, reply, attend a session, request material, or repeatedly interact with high-value content?
Declared intent	Did the person explicitly describe a need, timeline, evaluation, comparison, budget process, or desired next step?
Third-party research	Is the account researching relevant topics above its normal baseline, and is that signal corroborated?
Buying role	Is the person a user, evaluator, influencer, champion, budget owner, approver, or unknown?
Fit and use case	Is there a credible problem-product match and enough potential value?
Confidence	Is the evidence direct, inferred, probabilistic, anonymous, stale, or identity-matched?

Use a signal-confidence hierarchy

High-confidence signals include an explicit positive reply, a trusted referral, a confirmed project, a meaningful product-usage milestone, a procurement request, or an agreed next step.

Medium-confidence signals include a substantive event conversation, repeated high-value first-party behavior tied reliably to a person, relevant product use below the PQL threshold, or account research corroborated by another signal.

Low-confidence signals include an email open, isolated page view, generic download, social like, job change, funding event, profile view, or anonymous third-party surge. These can support research or routing, but they should not prove buying intent.

Apple’s Mail Privacy Protection documentation explains why senders may not be able to determine reliably whether an email was opened. Google’s sender guidance also says Google does not track open rates itself and cannot verify third-party open-rate accuracy. An open is weak telemetry, not strong person-level intent.

The Bombora Company Surge page describes account or domain-level research activity. 6sense’s intent-data explanation similarly describes anonymous behavior mapped to an account. Such data can help prioritize accounts, but it does not prove that a named person performed the research, controls the budget, wants contact, or has granted consent.

BrandJet’s social listening feature can help surface public context. Use that context to form a relevant business hypothesis, not to recite a surveillance trail. The guide to finding warm B2B leads on social media should be applied with the same confidence rules.

A workable local classification is:

• Cold: fit may exist, but no credible recent context has been verified.
• Warm: at least one strong relationship or engagement signal, or two independent medium-confidence signals, with reasonable recency and fit.
• Hot: explicit active need or evaluation plus an identifiable next step, timeline, or buying process.
• Reactivated: a dormant record has produced new credible evidence and returned to active treatment.

These are recommended operating rules, not industry standards. Calibrate them against your own mature outcomes in CRM lead management.

When to use cold, warm, or blended acquisition

Use cold acquisition when you need to enter a new market, reach a small set of strategic accounts, test an ICP, or create demand where inbound volume is insufficient. It works best when account value justifies careful research and when data, deliverability, and seller capacity are controlled.

Use warm acquisition when you have declared demand, referrals, existing relationships, meaningful product activity, substantive event interactions, or credible reactivation signals. The main operational risks are slow response, weak qualification, hidden upstream cost, and overinterpretation of activity.

Use a blended motion when outbound creates awareness or permission and later first-party behavior creates better context for the next touch. Preserve the original source and record the new warming evidence. Do not relabel every previously contacted account as warm and claim that the new label caused improved results.

A multi-channel outreach workflow can coordinate email, calls, and social touches, but the measurement should still report each channel’s attempts, connections, replies, meetings, and opt-outs separately. For warm cohorts that already show explicit interest, focus on moving from context to a useful next step. BrandJet’s guide to converting warm leads to customers covers that execution layer.

How to compare warm and cold cohorts fairly

A 30-day test can measure top-of-funnel operations. A full sales-cycle read is needed for opportunity, win, gross-profit, and ROI conclusions.

Step 1: Freeze the cohort

Choose one primary unit, either person or account. Define eligibility, exclusions, source, temperature rule, and entry date before launch. Deduplicate at both person and account level. Exclude employees, active opportunities, suppressed records, invalid contacts, and any other ineligible records consistently.

Step 2: Match important characteristics

Stratify or match warm and cold cohorts on ICP segment, company size, geography, role, product, use case, contract-value band, seller experience, and entry month. A high-fit cold account can outperform a low-fit warm inquiry. Mixing them makes temperature look more important than fit.

Step 3: Record an explicit event ledger

Track eligible units, valid units, attempts, delivered messages, completed dials, all replies, positive replies, booked meetings, held meetings, qualified opportunities, pipeline, wins, recognized revenue, gross profit, opt-outs, spam complaints, and negative replies.

Step 4: Read the right outcomes at the right time

At day 30, assess contactability, delivery, replies, booked meetings, held meetings, early opportunities, objections, complaints, and cost. At the end of one mature sales cycle, assess qualified pipeline, wins, losses, cycle length, gross profit, and ROI. Do not count open opportunities as wins or losses before they have had time to mature.

Step 5: Publish the denominator

For every rate, report the cohort, entity, source, temperature rule, period, numerator, denominator, sample size, attribution window, booked or held status, reply definition, included costs, exclusions, and known limitations.

Confounders that can distort the comparison

Selection bias: Warm groups self-select through engagement or relationship. Treat outcome differences as associations unless the treatment itself was randomized.

Deliverability and list quality: Invalid data, filtering, bounces, and poor reputation suppress cold results before a human can respond. Report attempted, delivered, bounced, blocked, and complaint outcomes separately.

Attribution windows: Warm opportunities may have long assisted paths through content, events, communities, ads, or product use. Preserve first-touch, sourced, influenced, and finance-recognized views.

Duplicate leads: Multiple contacts from one account can inflate denominators and give several touches credit for one opportunity. Deduplicate at both levels.

Sample size: A few extra wins can create a large percentage swing in a small cohort. Publish event counts and confidence intervals, not only rates.

Sales capacity: Slow routing or overloaded sellers can depress warm performance. Match response capacity, territory, seller experience, and SLA.

CRM-stage inconsistency: One company’s MQL can be another company’s inquiry or SQL. Publish stage-entry and stage-exit rules before comparing teams or external benchmarks.

Seasonality and maturity: Holidays, budget cycles, product launches, and incomplete sales cycles can distort the result. Use parallel periods and mature cohorts.

How to warm cold accounts ethically

The goal is to create useful context, not to manufacture a score.

Research the organization’s public priorities and form a relevant, falsifiable business hypothesis. Publish content that helps buyers understand the problem, tradeoffs, implementation, and risk. Run opt-in webinars, workshops, and roundtables. Seek introductions through customers, partners, investors, and advisers without pressuring the intermediary. Participate usefully in professional communities without harvesting members for automated outreach. Invite permission for a relevant follow-up after events. Reactivate old leads only when there is a credible new reason to reconnect.

Personalization should use one or two relevant facts, then state the hypothesis and invite correction. Do not tell someone that a hidden system detected an open or anonymous page visit. Do not combine personal details into a surveillance-style opening. Do not claim that a social action, job change, funding event, vendor score, or account-level research surge proves budget, urgency, authority, or consent.

Regional rules differ, and warmth does not override them. This is a high-level operational summary, not legal advice.

• In the United States, the FTC’s CAN-SPAM compliance guide says the law applies to commercial email, including B2B messages, and requires accurate headers, nondeceptive subject lines, a valid postal address, and a clear opt-out mechanism.
• In the United Kingdom, the ICO’s B2B marketing guidance explains that PECR treatment differs between corporate subscribers and individual subscribers such as sole traders and some partnerships. Identity and opt-out requirements still apply, and UK GDPR rights can apply when personal data is processed.
• In Canada, the CRTC’s anti-spam guidance says commercial electronic messages generally require express or qualifying implied consent, sender identification, contact information, and a working unsubscribe mechanism.
• In Australia, ACMA’s spam guidance requires consent, accurate sender identity, contact information, and an easy unsubscribe method. It also prohibits using or supplying address-harvesting software and lists produced using that software.
• In the European Union and EEA, GDPR Article 21 in the official regulation text gives people the right to object at any time to processing for direct marketing. Electronic-marketing rules also depend on applicable national law.

Deliverability rules are separate from legal permission. As of July 2026, Google’s email sender guidelines require SPF or DKIM for all senders and additional authentication, alignment, unsubscribe, and spam-rate controls for senders that transmit roughly 5,000 or more messages per day to personal Gmail accounts. Google says spam rates should stay below 0.30% and recommends keeping them below 0.10%. These platform requirements do not replace legal review.

FAQ

What is the warm leads vs cold leads conversion rate multiplier?

It is the warm cohort’s conversion rate divided by the comparable cold cohort’s conversion rate. If warm leads convert at 12% and cold leads at 2% using the same starting point and conversion event, the multiplier is 6x.

Do warm leads really convert 5x to 10x better than cold leads?

Not universally. A 5x to 10x difference can occur in a particular funnel, but it should not be treated as a general B2B benchmark. The result changes with lead definitions, funnel stage, channel, qualification rules, attribution window, and the conversion event being measured.

How do you calculate the warm vs cold lead conversion multiplier?

Calculate each cohort’s rate using the same numerator and denominator, then divide the warm rate by the cold rate. For example, a 15% warm opportunity rate divided by a 5% cold opportunity rate produces a 3x multiplier.

Which conversion rate should I use when comparing warm and cold leads?

Use the business outcome you need to improve, such as held meeting rate, qualified opportunity rate, or mature win rate. Both cohorts must start at an equivalent stage. A cold email reply rate cannot fairly be divided into a qualified inbound booking rate.

Can I compare inbound lead benchmarks with cold outreach benchmarks?

Usually not directly. Inbound and cold-outreach datasets often begin at different funnel stages and use different denominators. Compare them only when the cohort definition, starting event, conversion event, observation period, and qualification rules are sufficiently aligned.

Build your own warm vs cold conversion benchmark

Start with one segment and one motion. Define cold, warm, and hot operationally. Freeze the cohort, deduplicate it, record the evidence behind each label, and keep source separate from temperature. Track delivered messages, connections, positive replies, held meetings, qualified opportunities, mature wins, full cost, and gross profit.

After 30 days, use the early funnel to diagnose reach, relevance, routing, and capacity. After one mature sales cycle, calculate cost per held meeting, cost per qualified opportunity, pipeline per 1,000 eligible units, gross profit, ROI, and break-even rates. The result is a planning benchmark tied to your funnel rather than a generic warm-versus-cold percentage.