Published Last updated

Research

Home-Services Marketing Benchmarks: Study Methodology

This page sets out how C&C Gorilla researches and reports home-services marketing benchmarks: what we measure, how each figure is defined, the standard a figure has to clear before it is published, the limitations that travel with it, and the studies currently in progress. Figures are published only when the sample behind them supports publication, which is why the standard is on the record before the results are.

How a figure gets published

The pipeline every benchmark figure has to pass through

The standard comes before the results, because a figure cannot be shaped to suit a headline once the method, the sample requirement and the disclosure are already on the record.

  1. Data sources

    Account, call tracking and CRM records rather than platform-reported conversions alone, with every field traced to the system that records it.

  2. Measurement framework

    The chain a figure has to follow: impression, click or call, lead, qualified lead, appointment, estimate, sale, revenue.

  3. Sample requirement

    A minimum number of accounts per published industry figure and campaigns per metric, with the sample stated beside every value.

  4. Publication gate

    Each gate applied at publication, and any figure that does not clear its gate withheld rather than estimated.

Two outcomes, and only two

  • Gate met

    Published with its disclosure

    Released with its sample and coverage, date range, industry breakdown, methodology, exclusions, limitations and definitions attached.

  • Gate not met

    Withheld, and the reason stated

    Not published, with the absence recorded rather than filled in. Where tracking was insufficient to attribute outcomes, the account is excluded rather than estimated.

No figure will be described as a guarantee, a target or a promise of performance, and if a release has to be corrected the correction is recorded rather than applied silently.

Why the standard comes before the results

Benchmark numbers are quoted constantly in this industry and are frequently unsourced, averaged across trades that do not behave alike, or drawn from a sample too small to mean anything. Publishing the method first means the figures cannot be shaped later to suit a headline.

  • A benchmark is only useful if the reader knows the sample, the period and the method behind it.

  • A cost-per-lead figure blended across HVAC, roofing and restoration describes no real business, so we do not produce one.

  • Publishing the definitions and the publication standard first is what makes the figures checkable when they arrive.

What the study will report, and how each point is defined

Each data point below is published only when the underlying sample is large enough to report without disclosing a single client’s performance.

Data pointHow it will be defined
Sample size
Date range
Methodology
Industries
Ad spend analysed
Lead volume
Cost per click
Cost per lead
Conversion rate
Call versus form performance
Lead quality
Booked jobs
Revenue
Impressions
Clicks
Calls
Forms
Qualified leads
Appointments and inspections
Estimates issued
Sales
Customer acquisition cost
Return on ad spend
Mobile versus desktop
Channel
Last updated

Publication standard

Our publication standard: what a figure must clear before it appears

These are requirements rather than targets, and they are applied at publication.

Nothing on this page is a result. The study has not been published, and every rung below is a requirement a figure has to meet rather than a value that has been measured.

  1. Gate 01

    Accounts per published industry figure

    At least 10 accounts contributing to the same industry and channel.

  2. Gate 02

    Campaigns per published metric

    At least 30 campaigns, so a single unusual campaign cannot define a range.

  3. Gate 03

    Spend behind a published cost figure

    A stated minimum spend analysed, reported alongside the figure.

  4. Gate 04

    Downstream data completeness

    Outcome figures published only where at least 70% of leads have a recorded downstream outcome. Below that, the figure is reported with its coverage stated.

  5. Gate 05

    Re-identification

    No figure published where a single account contributes more than 25% of the sample, or where a market and service line combination would identify one business.

A figure that does not clear its standard is withheld rather than estimated, and the reason is stated.

What will be published alongside every figure

A number without its sample, period and method is not a benchmark. This is the disclosure that travels with every release.

DisclosedDetail
Sample size and coverageAccounts, campaigns and the proportion of leads carrying a downstream outcome.
Number of businessesDistinct businesses contributing, so the reader can judge whether a figure rests on a handful.
Total spend analysedThe money behind the figures, per channel and per industry.
Date range and last updatedThe window measured and the date the figures were recalculated.
Data sources usedWhich systems each figure came from, including where a client supplied data by hand.
Industry breakdownEvery figure reported per trade, never as a single all-market average.
Methodology and exclusionsHow accounts were selected, what was excluded and why.
LimitationsWhat the results do not prove, published alongside the results rather than on request.
DefinitionsWhat counts as a lead, a qualified lead, an appointment and a sale in this study.
Author and reviewerNamed, once the author and review information exists. We will not publish a study under a placeholder name.
Citation guidanceHow to cite the study, so the sample and period travel with any figure taken from it.

Methodology principles

  • Every metric is calculated from account, call tracking and CRM data rather than from platform-reported conversions alone.

  • Results are reported per industry and per intent family, never blended into a single average.

  • Sample sizes are stated next to every figure, including where a metric rests on a small number of accounts.

  • Accounts are excluded where tracking was insufficient to attribute outcomes, rather than estimated.

  • Client identities are never disclosed, and any figure that could identify a single account is aggregated or withheld.

Limitations we will publish alongside the results

Every study has boundaries, and stating them is part of reporting honestly.

  • Results reflect the markets, service mixes and seasons the accounts operated in, not a universal market.

  • Spend levels vary between accounts, so cost figures are reported as ranges and medians rather than single averages.

  • Outcomes depend on sales process and capacity as well as marketing, which the study cannot isolate completely.

  • Data reflects periods already completed, and costs in a competitive market move over time.

Where the data comes from

Every field is traced back to the system that records it, because a figure that cannot be audited back to its source is not a benchmark. Where a source cannot capture part of the ladder, the limitation is stated rather than estimated around.

Website enquiry forms (this site)

Every enquiry submitted through this site is stored with its service, business and message, and can carry campaign parameters.

Limitation: Captures only enquiries that reach this site. Calls and third-party form tools are outside it.

Feeds: leads, forms, device, date_range

Google Ads

Client-owned accounts, read through the client’s own access.

Limitation: Platform conversions are not sales. Conversion actions must be defined and audited before the data means anything.

Feeds: spend, impressions, clicks, ctr, cpc, campaign_type, brand_vs_nonbrand, device

Meta Ads

Client-owned ad accounts.

Limitation: Attribution windows and modelled conversions differ from search. Meta-reported results are comparable across Meta accounts, not across channels.

Feeds: spend, impressions, reach, clicks, ctr, cpc, campaign_type, device

Call tracking

Where a call tracking system is in place, call data is the most reliable record of phone demand and the only place answer rates exist.

Limitation: Requires dynamic number insertion to attribute a call to a source, and call quality still has to be judged by listening or scoring.

Feeds: calls, call_answer_rate, call_duration, response_time

Client CRM or scheduling system

This is where lead quality becomes visible, because it holds outcomes rather than enquiries.

Limitation: Field mapping differs per system, lead records rarely carry campaign attribution, and matching a CRM record back to a lead is the hardest step in the chain.

Feeds: appointments, estimates, sales, revenue, qualified_leads, lead_to_appointment, appointment_to_sale

Client invoicing or accounting data

Invoiced value is the only honest revenue figure.

Limitation: Usually supplied as an export rather than a live connection, and job costing has to be supplied with it for gross profit to mean anything.

Feeds: revenue, gross_profit, roas, cac

Website analytics

Session and channel context, useful for organic and reactivation measurement.

Limitation: Consent, ad blockers and cross-device behaviour mean session counts understate reality, and analytics cannot see the sale.

Feeds: impressions, clicks, device, new_vs_returning

Reactivation sending platform

Sends, opens and attributed bookings for email, SMS and mail waves.

Limitation: A reactivation conversion is often a phone call, so without call tracking the outcome is invisible even though revenue results.

Feeds: send_volume, leads, sales, revenue

Gorilla reporting layer

The layer where platform, call and CRM records are joined into one view per client.

Limitation: Derived figures are only as complete as their weakest source. Where a join is missing, the gap is reported rather than estimated.

Feeds: cac, roas, qualified_cpl, lead_to_qualified, response_time

Methodology decisions we have made

Publishing a study means committing to answers on these questions. Ours are stated here, so a reader can hold any future figure to them rather than taking the method on trust.

How are industries grouped?

Never blended. Results are reported per trade and per intent family, and a blended all-industry average is not published.

How is client identity protected?

No client is named, and any figure a single account could be identified from is aggregated further or withheld.

What happens where tracking was insufficient?

The account is excluded from that metric rather than estimated. Coverage is reported so a reader can see how much of the sample a figure rests on.

Are platform-reported conversions used as results?

No. Platform data is used for delivery and cost. Outcomes come from call tracking, CRM and invoicing data.

How is spend reported?

As ranges and medians rather than single averages, because spend levels vary widely between accounts of different sizes.

Each decision above is applied to every account in a study rather than chosen per account, and the definitions travel with any figure taken from it, including how it is cited.

Research agenda

The studies in progress behind this methodology. Each states the question it answers, the data it draws on and the standard it has to meet before publication, so the subject matter and the bar are both public from the outset.

In progress · 1 of 5

Home-Services Lead Quality Report

Lead volume is the number most agencies report and the number most owners act on. The ratio between raw leads and sold jobs is where the difference between a good month and a bad one actually lives, and it is almost never published by channel.

Question answered

Across home-service trades and channels, what proportion of raw leads become qualified leads, appointments, estimates and sold work?

Data required

  • Lead records with channel attribution
  • Qualification outcome per lead, using one stated definition
  • Appointment and estimate records where the trade uses them
  • Sold or not sold, with the reason recorded where possible

To be published it must clear

  • At least 10 accounts per published trade and channel
  • At least 30 campaigns contributing to each channel figure
  • At least 70% of leads carrying a recorded downstream outcome
  • One qualification definition applied across every account

In progress · 2 of 5

Call vs Form: Conversion Differences in Home Services

Most businesses optimise for the cheaper lead. If calls convert materially better than forms, cost per lead is the wrong metric and cost per sold job is the right one, which changes how budget should be allocated.

Question answered

Do phone enquiries convert to sold work at a different rate than form enquiries, and how does that differ by trade, urgency and average job value?

Data required

  • Call and form leads recorded separately, never combined
  • Call answer rate and response time
  • Qualification and sale outcome per lead, by contact method
  • Average job value by contact method

To be published it must clear

  • Call tracking in place for every contributing account, since untracked calls would bias the call side
  • At least 30 campaigns with both call and form leads present
  • Answer rate reported alongside, so unanswered calls are not counted as failed demand

In progress · 3 of 5

What Home-Service Businesses Actually Spend, and What It Returns

Percentage-of-revenue rules are repeated constantly and rarely tested. Reporting actual spend as a share of revenue, next to the return it produced, replaces a rule of thumb with a distribution.

Question answered

What do home-service businesses of different sizes actually spend on marketing, and what revenue and gross profit does that spend produce?

Data required

  • Total marketing spend, including management fees and production, not only ad spend
  • Invoiced revenue for the same period
  • Gross profit where job costing exists
  • Service mix, because a 20% spend on maintenance work is not the same as 20% on installs

To be published it must clear

  • At least 10 accounts per size band
  • Twelve consecutive months per account, to remove seasonality
  • Spend and revenue drawn from the same accounting period
  • Gross profit reported only for accounts that supply job costing

In progress · 4 of 5

Reactivation: Revenue From Customers Already Owned

Reactivation is usually the cheapest demand available to a service business and the least measured. A study comparing reactivated revenue against new-customer acquisition cost turns it from a tactic into a budget line.

Question answered

What revenue does a structured reactivation programme produce from an existing customer list, and how does it compare with acquiring new customers?

Data required

  • Customer list size, with recency and service history
  • Reactivation wave sends by channel: email, SMS, mail
  • Booked work attributed to each wave, and its invoiced value
  • Comparable new-customer acquisition cost for the same period

To be published it must clear

  • At least 20 accounts with lists of sufficient size and history
  • A stated definition of what counts as reactivated work
  • Hold-out or control where available, since some reactivated customers would have returned anyway

In progress · 5 of 5

Seasonality: How Lead Cost and Quality Move Through the Year

Budgets are usually set annually and spent evenly, while demand in seasonal trades does not arrive evenly. Monthly cost and quality curves would show whether spending should be pulled forward into shoulder months.

Question answered

How do cost per lead, lead quality and close rate move through the year in seasonal home-service trades, and what does that imply for budget timing?

Data required

  • Monthly spend, leads, qualified leads and sold work across at least one full cycle
  • Trade and market for each account
  • Weather or season context where available
  • Enough accounts per trade to separate a pattern from one unusual year

To be published it must clear

  • Twenty-four consecutive months per account, so one unusual season does not define the pattern
  • At least 10 accounts per trade
  • Cost figures reported as ranges, because regional markets differ

How results will be released

When the study is published, it will carry the same structure as the tables above, with the sample stated and the limitations attached. Versions will be dated so earlier figures can be compared with later ones.

  • Publication will be announced on this page and linked from the relevant industry pages.
  • No figure will be described as a guarantee, a target or a promise of performance.
  • Agencies and competitors are free to cite the study, provided the sample and period are cited with it.
  • If a release has to be corrected, the correction will be recorded rather than applied silently.

Where to go from here

Keep reading

No next step required. These are the pages that go deeper on the same question.

Talk to us

A conversation about your own numbers, with no obligation.

Questions about this study

Want your own numbers rather than an industry average

We would rather measure your actual cost per booked job, by channel and by service line, than compare you to a figure averaged across trades that behave differently.

Book a Growth Strategy CallHow the Growth Engine works