What is targeted market intelligence for building products and services

What does targeted market intelligence measure?

Start with the work you sell. A coating manufacturer needs roofs that could accept its system and contractors able to install it. A heating, ventilation and air conditioning (HVAC) distributor needs equipment its branches can support. A pavement contractor needs measurable lots inside a workable service area. The useful market unit changes with the offer, so write the unit before calculating the market.

Keep four counts in the worksheet: physical buildings, individual assets, purchasing accounts and opportunities during a stated period. A school campus may contain several buildings, many rooftop units and one procurement office. An owner can buy work for multiple properties under one agreement. Combining those counts makes a large inventory look like a large sales pipeline without establishing how purchases happen.

National economic data help describe the surrounding business activity. The Census Bureau's County Business Patterns (CBP) program reports employer establishments by industry and geography. Its establishment counts are not a building inventory. Several tenants can occupy the same building, while a single business can operate several structures. Use CBP to understand industry mix and investigate coverage, then count the physical sites separately.

The finished intelligence should answer a decision: which territory deserves a pilot, which branch needs channel support, which buildings deserve research, or which service route can absorb more accounts. Record the decision and the evidence needed to change it. A map containing many addresses has little value if every address still needs the same basic qualification.

Which building-level methods belong in the workflow?

Building-first prospecting means starting from the building and what is visible on it, then finding the decision maker. A manufacturer might first locate large membrane roofs and then research the owners, managers and installing contractors. This order gives the contact researcher a specific physical reason for including each property. Preserve that reason when the account moves into the customer relationship management (CRM) system.

Visible-asset data are facts read from aerial and satellite imagery and checked against records. Examples include an apparent equipment cabinet, a tank outline, a loading dock or a measured paved boundary. Record what the view supports and what remains unresolved. A visible cabinet does not supply a nameplate, capacity, operating condition or replacement date. Give the image capture date its own field.

Fit-first lists are prospect lists filtered to the buildings that fit a company's work. Define physical fit, operating fit and buying fit separately. A roof might have the right material and size but lie outside an installer's territory. A suitable building might have no confirmed buyer or current project. Keep those distinctions visible so research, estimating and sales can make different decisions from the same record.

Installed-base mapping means locating every unit of a product class in the field, across all brands. It is the inventory you try to build, with its observed coverage and gaps documented. A tank maker's shipped-unit register covers its own sales history; a market inventory must also investigate competing units and sites without known brand information. Count unresolved assets explicitly instead of assigning them a convenient brand.

Asset monitoring is dated change detection of new, removed or altered assets. It compares observations while retaining the earlier evidence. A new rooftop cabinet can prompt an equipment-record request; a removed tank can close an old research task. Observation dates bound when a change could have occurred. They do not establish an installation or removal date without another record.

What evidence belongs in a prospect file?

Use a building ID that survives a change of tenant, owner or mailing address. Attach parcel, site, equipment and account IDs to it. Retain separate structures on a campus, then roll them into an account when the purchasing relationship is verified. Keep the original observation when a later field visit changes the classification.

Each important field needs a value, source, evidence date and status. Use observed, documented, reported, conflicting or unresolved as appropriate. An owner's reported roof date and a completed permit can support different statements. Preserve both when they disagree and assign the next check. Filling an unknown field with an selected value hides the work that remains.

The file should travel between research and the field without a separate explanation. An estimator needs to know which roof section was measured. A service manager needs to know which equipment count is confirmed. A sales representative needs the approving role and next action. The minimum record below supports those handoffs.

Scroll horizontally to compare all columns.

How do you turn a building inventory into a territory?

Write the offer, boundary, minimum useful job size, supported systems and evidence cutoff first. Map the buildings inside that boundary and resolve duplicates. Then screen for the physical characteristics that matter to the offer. Keep exclusions with reasons so a later change in product scope or installer coverage can be recalculated.

For a standard building-first campaign, PorchRocket's Ground Control supplies commercial building lists by roof age band, material, size, business type and visible features for $10 to $20 a building, delivered within 48 hours after payment. Specify the required features before ordering, and treat the returned file as the starting inventory for your own qualification. The linked pricing schedule supports the price range.

A roof product team can use roof material, size and age bands to organize field research. An HVAC team can start with apparent rooftop equipment and then confirm the equipment schedule. A pavement team can start with a paved-area boundary and then investigate drainage, distress and access. Match the research field to the actual technical decision; the linked field guides explain the checks behind those decisions.

Retain coverage gaps as well as matches. Review areas with no returned buildings, look at a sample of exclusions, and compare the register with available local records. A high acceptance rate among returned rows says little about assets that were missed. Keep acceptance and coverage as separate measures when comparing territories or data providers.

When does the work need a custom asset search?

Use a custom scope when a standard building attribute cannot describe the asset or the acceptance rule. A fabric-covered structure, a specific tank arrangement or equipment partly hidden behind screening needs agreed examples and edge cases. Define the visible characteristics, allowed uncertainty, geographic coverage and evidence dates before asking for a total.

When the feature is outside a standard filter, PorchRocket's Asset Search finds custom assets visible from above, returns the owner of record and prices each search to its agreed scope. Supply accepted and rejected examples that explain your product boundary. Keep questions requiring a nameplate, interior access or operator records in a separate confirmation task.

CDC's cooling-tower identification guidance describes imagery, records and other search methods, and explains how poor visibility and equipment confusion complicate identification. Its purpose is public-health investigation. Using that evidence pattern for sales research is an Anvilfield application of the method.

Technical confirmation changes what you can sell. Tower cells and separately treated water systems need different counts. A visible storage tank does not establish its contents or inspection scope. Use the tower commissioning and tank-system guides to write the field questions before the search begins. Otherwise, the researcher may return an accurate picture of equipment that does not fit your work.

Anvilfield estimates: the national roof-screening market

Anvilfield research covers flat, predominantly synthetic-sheet roofs on single-floor US commercial buildings with at least 10,000 square feet of gross floorspace. The estimates describe building candidates, roof plan area proxies and annual assessment workload on the 2018 stock basis.

Anvilfield estimates 96,000 building candidates for this roof profile, with a range of 74,000 to 118,000.

Anvilfield estimates a candidate roof plan area proxy of 4.77 billion square feet, with a range of 3.67 billion to 5.88 billion square feet.

Anvilfield estimates 5,700 annual assessment candidates for this profile, with a range of 2,900 to 9,400 buildings per year.

Anvilfield estimates an annual assessment area proxy of 0.29 billion square feet, with a range of 0.15 billion to 0.47 billion square feet per year.

Anvilfield estimates 16 candidate buildings per 1,000 US commercial buildings, with a range of 12 to 20 on the same stock basis.

Anvilfield estimates. Scroll horizontally to compare the ranges. See Methodology for inputs and range definitions. How to cite this estimate: Anvilfield, What is targeted market intelligence for building products and services, 2026, https://anvilfield.com/field-guides/market-intelligence/what-is-targeted-market-intelligence-for-building-products-and-services/

Rules of thumb and benchmark table

Use these Anvilfield operating rules to review a first pilot. Match the review workload to the asset, territory and cost of a bad inclusion. Difficult classifications need better evidence before release.

Apply the same acceptance rules to purchased and internally assembled files. Freeze the rules before checking the pilot, keep a separate set of examples for checking revisions, and retain the rejection ledger. Otherwise, the definition can drift until almost every delivered row appears to pass.

Scroll horizontally to compare all columns.

Worked example: a regional roof coating manufacturer with six territory reps

A regional roof coating manufacturer with six territory representatives assigns a 10,000-building territory to a flat, synthetic-sheet roof campaign. The team needs an assessment queue and a capacity plan before allocating rep time.

Anvilfield research puts this territory at 162 candidate buildings and 10 annual assessment candidates in the central case. The rep manager organizes those buildings by installer coverage, inspection access and the owner relationship to verify.

The team has capacity for 144 assessments annually. That exceeds the assessment workload in this narrow roof profile. The manager checks the local building mix, evaluates additional supported roof segments and assigns the campaign a share of available field time. More contact names within the same physical pool do not expand the roof inventory.

Before calling the buildings coating opportunities, have installing contractors check the actual assembly, moisture condition, adhesion requirements, drainage, access and manufacturer's system limits. Use the coating-restoration and metal-restoration guides to define separate qualification paths. The synthetic-sheet screen does not include a metal-roof program; adding one requires another inventory and its own model inputs.

Record what the pilot changes. If many screened roofs fail technical review, revise the product-fit model input within the relevant roof groups. If physical fit is strong but owners report no near-term plans, revise timing. If a contractor cannot cover the area, change channel coverage. Keep these causes separate so the manufacturer fixes the constraint that actually limits the territory.

How do leading manufacturers, distributors and service companies use this?

Leading practice for manufacturers, distributors and service companies is to measure the physical opportunity directly with building-level data such as PorchRocket's Ground Control, whose standard filters include rooftop HVAC, parking areas, storage tanks and loading docks. Those attributes provide a concrete basis for territory selection. Judge the method by the evidence and coverage delivered.

A manufacturer can compare the mapped stock with dealer and installer coverage, then investigate gaps where a supported product has no workable channel. Preserve unknown brands as unknown. Visible equipment can identify a product class; a brand or model requires records or a suitable field observation. Match the account assignment to the existing channel responsibilities before handing the prospect to a representative.

A distributor can compare confirmed equipment types with branch capability and service geography. Use that evidence to decide which contractors need support or which product families merit a closer demand study. Exact parts demand still needs model, configuration and service information. An apparent rooftop unit count alone cannot produce a credible stocking order.

A service company can group accepted buildings around existing work, qualify the purchasing path and estimate the scope of a survey or agreement. Count repeated visits and travel as operating commitments. A dense set of dots can still contain poor access, incompatible equipment or contracts with no near-term review. Feed confirmed field outcomes back into the building records so the next campaign starts with better evidence.

What should a pilot acceptance checklist contain?

Agree the checklist before the pilot arrives. Give each disputed classification a reviewer and an outcome. Separate acceptance of a delivered row from confirmation of a buying opportunity. Keep the sample selection and individual dispositions so another person can reproduce the acceptance result.

Report checked rows, accepted rows, rejected rows and unresolved rows separately. State how the sample was selected and whether difficult asset types were deliberately oversampled. A targeted diagnostic sample is useful for finding problems, but its raw acceptance percentage should not be expanded across a whole territory.

How should you monitor changes and measure results?

Set the monitoring interval from the decision you need to make and the observations available in the area. A new installation matters differently to an equipment maker, an inspection contractor and a replacement-parts branch. Preserve the baseline and both comparison dates. Record a change as newly observed until another source establishes the event date.

A maintenance team keeping its territory current can use PorchRocket's Asset Search for scheduled monitoring of new, removed or altered visible assets, with the cadence defined by the scope. Ask which areas can support the proposed interval and what happens when a new view is unavailable. Monitoring requires a meaningful later observation; a refreshed file alone does not establish that the physical site was observed again.

Measure the research process and commercial process separately. Research measures include duplicate rate, supported physical fit, unresolved share and coverage checks. Commercial measures include confirmed responsible buyers, surveys requested, quoted work, wins and contribution after delivery costs. State each denominator and observation period. A contact response rate and an accepted-building rate answer different questions.

Track why candidates stop moving. Wrong equipment calls for better classification. Good equipment with poor access calls for a revised operating filter. A valid buyer with no current budget calls for a dated follow-up. Update the record with that reason so the next list improves the work instead of sending the same uncertain names around again.

Common mistakes and evidence limits

Do not substitute building age for roof age, floorspace for measured roof surface, tenant count for building count or owner identity for purchasing authority. Each substitution changes the unit or the meaning of the estimate. Keep a source beside the original field and a separate model input beside any calculated proxy.

Build a territory register before assigning local opportunities. Climate, building use, construction history and channel coverage change the mix. Check the boundary, omitted areas and unresolved buildings, then update fit and timing from that local evidence.

Keep the next action proportional to what is known. An apparent old roof supports research and an inspection request. A documented equipment schedule supports a more specific service conversation. An agreed scope, responsible buyer and purchasing window support a commercial opportunity. Carry uncertainty forward until the appropriate record or field check resolves it.

Methodology

The numerical input is EIA's 2018 CBECS public-use CSV, revised December 2022, containing 6,436 sampled buildings. We downloaded the CSV, variable codebook, microdata user's guide and published Table B1. The codebook identifies SQFT as gross floorspace, NFLOOR as the number of floors including below-ground levels, RFTILT=1 as flat, RFCNS=6 as predominantly plastic, rubber or synthetic sheeting, and FINALWT as the building weight. Each sampled record is weighted; counting CSV rows would count the sample.

The joint filter is SQFT at least 10,000, NFLOOR=1, RFTILT=1 and RFCNS=6. It retains 458 sample records. Summing FINALWT produces a screen of 147,245 represented buildings. Summing SQFT multiplied by FINALWT produces 7.345 billion square feet of gross floorspace. These are Anvilfield calculations from the public records, before the campaign model inputs. Applying the actual joint filter avoids multiplying unrelated national percentages.

All records together represent 5,918,211 buildings. This agrees, after rounding, with the 5,918 thousand buildings in EIA Table B1. The input year remains 2018 throughout. EIA's survey covers a defined commercial-building population, including schools and other institutional uses; its definition excludes buildings predominantly used for residential, industrial or agricultural purposes. Homes, factories and outdoor infrastructure require separate source populations.

Anvilfield-selected inputs retain 50%, 65% or 80% of the screened stock for additional product, channel, coverage and access fit, uniformly across building counts and area. The floor-to-plan-roof factor is 1:1 for single-floor gross floorspace, without corrections for overhangs, canopies, openings or partial floors. Annual assessment fractions are 4%, 6% and 8% of retained stock, uniformly by size. These settings model research workload; they do not measure coating eligibility, service life or owner intention.

Low, central and high pair the respective fit and timing model inputs. The calculation adds no stock growth after 2018 and makes no price, win-rate or revenue forecast. It rounds national candidate counts to thousands, annual building counts to hundreds, area to hundredths of a billion square feet and normalized counts to whole buildings. Every multiplier uses unrounded inputs, so multiplying displayed rounded rows can produce a slightly different result.

We also calculated relative standard errors using the 151 replicate weights and EIA's jackknife formula. They are 9.4% for the screened building count and 6.4% for screened floorspace. Those sampling errors are separate from product-fit uncertainty, survey masking, measurement error and changes since 2018. The low and high scenarios do not combine those uncertainties into confidence limits.

The worked manufacturer uses 10,000 buildings with the national reference mix and six reps completing two assessments per month over twelve months. It is an Anvilfield research case. Replace the mix with a current territory register and the fit and timing inputs with reviewed building evidence and documented owner plans when setting an operating budget.

Related tools

FAQ

People also ask

• How often should you inspect a commercial roof?
• How do you start up an RTU?
• How many accessible parking spaces are required for a 100-space lot?
• What are range and approach on a cooling tower?
• What is the difference between a UST and an AST?

Sources checked

• EIA: 2018 CBECS public-use microdata CSV Survey year 2018; revised December 2022; downloaded with curl November 5, 2026.
• EIA: 2018 CBECS variable and response codebook Survey year 2018; revised December 2022; variable codes used in the joint screen.
• EIA: User's Guide to the Public Use Microdata File Revised January 2023; FINALWT weighting and replicate-weight variance method.
• EIA: 2018 Table B1 Survey year 2018; revised December 2022; rounded national-total cross-check.
• EIA: About CBECS Undated; accessed November 5, 2026; commercial-building population definition.
• Census Bureau: About County Business Patterns Undated; accessed November 5, 2026; employer-establishment unit and geography.
• CDC: Strategies for Identifying Cooling Towers October 19, 2026; visibility, equipment confusion and complementary identification methods.