Manaz Afzal

Trades & Field Services · Enquiry sizing

Large jobs reach an estimator first, not last

A sizing step that reads each enquiry and returns a likely price band and range, tested on 80,327 alteration jobs filed after the model was built.

61%of job value in the first 20% of enquiries; rule 42%
62%of $250k+ jobs found in that first 20%; rule 41%
96%placed in the right price band or the next one

Executive summary

Sorting enquiries by size read from the description puts 61% of total job value in the first fifth of the queue, against 42% for the typical cost of that trade and borough, and finds 62% of $250k+ jobs there.

61%of job value in the first 20% of enquiries, against 42%
62%of $250k+ jobs in that first 20%, against 41%
96%of enquiries placed in the right price band or the next one

Situation

Estimators work enquiries in arrival order, so large jobs can wait behind small ones and go to the firm that called back first.

Approach

A text model on the job description, building type and location, trained on 59,920 earlier jobs and tested on 80,327 later ones, with calibrated 80% ranges.

Result

Rank correlation with declared cost rises from 0.30 to 0.66. Ranges are well calibrated at 79% coverage but wide, so the output orders the queue and gives a first band rather than a quote.

The problem

A builder, fit-out firm or building-products supplier gets new work from two places: enquiries through the website and phone, and project leads from building approvals through services such as Cordell Connect. An estimator can price only a few a day properly. Firms that filter and rank leads before they reach the estimator say the gain is in reaching the large, suitable jobs first.

A precise quote from a paragraph of text is not realistic. What a job description does carry is a strong signal of scale, and I set out to measure how well that signal sorts jobs by likely size.

Team
Estimators, sales coordinator, operations manager
Systems
Website enquiry form, Cordell Connect or council approval data, HubSpot, ServiceM8 or simPRO
Measure
Share of total job value reached by the first enquiries worked; price band placed right

What I built

A sizing step that reads the enquiry description with the type of building and location, and returns a likely price band and an 80% range. Enquiries are sorted by expected size, so estimators reach the large jobs first and the customer can be given a realistic first band on the day.

EnquiryDescription, site,building typeSizePrice band and80% rangeQueueEstimators worklargest firstReplySame-day first band,site visit if needed
The band sets the order of work and the first conversation. The estimator still prices the job.

Results

On 80,327 alteration jobs filed from July 2018 to December 2019, working the highest-sized 20% of enquiries first reaches 60.7% of all job value. Sorting by the typical cost for that trade and borough reaches 41.7%, and arrival order reaches 20%.

Taking enquiries in arrival order20%Typical cost for the trade and borough42%Size read from the description61%Perfect ordering (ceiling)83%
0%25%50%75%100%10%20%30%40%50%Description modelPerfect orderingTrade and borough ruleShare of enquiries worked
The description model stays well ahead of the trade-and-borough rule across the first half of the queue.
Arrival order20%Typical cost for trade and borough41%Size read from the description62%
Of the jobs declared at $250k or more, 62% are in the model’s first 20% against 41% for the rule.

Price bands and ranges

Trade and boroughDescription modelExact price band57%Within one band96%
The band is exactly right 57% of the time and within one band 96% of the time.
Description modelTrade and borough
80% ranges that contained the declared cost79.4%81.7%
Median range, top over bottom17.2×55.5×
Median error of the single figure59.5%71.9%
Single figure within 25%21.5%17.0%
Rank correlation with declared cost0.660.30

The 80% ranges hold their promise, containing the declared cost 79.4% of the time, but they are wide: the top is typically 17 times the bottom. That is a ranking and triage tool. Any figure given to a customer is a band, not a quote.

Where it goes wrong

Trade and boroughDescription modelMechanical and air conditioning62%Plumbing and gas61%Kitchen and bathroom55%Facade, roof and windows43%Interior fit-out66%Structural and other69%
The model beats the rule for every type of work. Structural and facade jobs have the widest spread of cost for similar wording.
Declared cost fifthTest jobsRangeModel median errorBaseline median error
Lowest fifth16,223$1,000 to $12,500157.5%438.2%
Second15,908$12,502 to $28,40651.1%45.8%
Middle16,065$28,413 to $55,10639.3%33.3%
Fourth16,448$55,110 to $150,00049.1%56.3%
Highest fifth15,683$150,001 to $20,000,00073.2%88.0%

The smallest jobs carry the largest percentage errors, because a description such as plumbing work could be $2,000 or $20,000. Those enquiries matter least for estimator time, which is why the sizing is used to order the queue rather than to price.

$1k to $10k19%$10k to $100k52%$100k to $1m26%$1m and over3%
Declared costs span four orders of magnitude; the median is $39,761.

Data and method

Data
NYC Department of Buildings job application filings (NYC Open Data, public): Alteration Type 2 jobs, which cover multiple trades with no change of use, filed 2017 to 2019. The target is the estimated job cost declared at filing. Applicant and owner names are not downloaded. 179,270 filings; 170,701 kept after removing those with no cost or description
Split
By filing date. 2017 filings train (59,920), January to June 2018 calibrate the ranges (30,454), July 2018 onward test (80,327)
Model
Ridge regression on log cost, with TF-IDF word and two-word features of the description, one-hot building type, borough and work type, and numeric building details
Ranges
Split-conformal 80% ranges from the calibration filings, set separately by predicted price band
Baseline
Quantiles of declared cost for the same type of work and borough in training
Description length
median 20 words

Limitations

How I would take it live

Financial impact

Large jobs called back first: value from a higher win rate

The measured gain is in which enquiries reach an estimator first. The value depends on how much a faster call-back lifts the win rate on large jobs, which is an assumption.

InputValueSource
Enquiries a month120 mid-sized builder or fit-out firm
Share at $250k and over12.9%Measured test period
Large jobs in the first fifth, model62.0%Measured
Large jobs in the first fifth, rule40.6%Measured
Win-rate lift from a same-day call-back5%
Average large jobA$450,000
Gross margin12%

Calculation

  1. Large enquiries a month15.5
  2. More of them reached first each month3.3
  3. Extra large jobs won a year2.0

Annual gross margin from faster call-backs on large jobsA$107,453

Estimator time saved on small enquiries that receive a band by return is a second gain and is not counted here.

Implementation and rollout

Eight weeks from enquiry export to a sized queue

  1. Weeks 1 to 2

    Export and baseline

    Two years of enquiries with quoted or invoiced values from the job system; measure today’s call-back time by job size.

    Operations manager, Manaz

  2. Weeks 3 to 4

    Train and set bands

    Train on earlier jobs, test on the latest quarter, agree price bands and range wording with the estimators.

    Manaz, senior estimator

  3. Weeks 5 to 6

    Band beside each enquiry

    The band appears on each new enquiry; estimators work as usual and note where it is wrong.

    Estimators

  4. Week 7 onward

    Sized queue

    The queue is sorted by expected size; small enquiries get a first band by return. Quarterly retraining on won and lost jobs.

    Operations manager, Manaz on support

Tools

Website enquiry form or shared inbox, ServiceM8, simPRO, Tradify, Jobber or HubSpot as the job and enquiry record, Python (scikit-learn) model behind a small API, Make or n8n to write the band and range onto each enquiry

Team

  • Operations manager owns the queue order and signs off go-live
  • Senior estimator sets the bands and reviews misses
  • Manaz data work, model, integration and monitoring

The timeline is a plan for a typical engagement; it depends on data access and how many systems are involved.

Risks and mitigations

What could go wrong in production, and the control for each

RiskMitigation
A customer reads the band as a quoteEvery band is shown as a range with a line that a site visit sets the price.
Small jobs wait too longA maximum wait applies to every enquiry regardless of size.
Descriptions too short to sizeShort enquiries are flagged and the form asks two questions on scope and site.
Costs shift with pricesQuarterly retraining on the firm’s latest quoted and invoiced values.

Key takeaways

Key takeaways

  • The first fifth of a sized queue holds 61% of job value, against 42% for a trade-and-borough rule.
  • 62% of $250k+ jobs are in that first fifth.
  • The ranges are well calibrated but wide, so they order the queue and start the conversation; the estimator still prices.
  • The model is retrained on the firm’s own enquiries and job values before use.