Why a flat response rule wastes your best hours
Most local service businesses treat every inbound lead the same: first in, first called, everyone gets the same follow-up sequence. That feels fair. It is also expensive. A same-day install request from a homeowner inside your service area sits in the same queue as a quote shopper three towns over who filled out a form at midnight and never answered the phone again.
Your estimators and office staff have finite hours. When those hours spread evenly across leads of wildly different quality, the hot leads wait behind the cold ones. Speed-to-lead matters, but speed only pays when it is aimed at the right lead. Scoring is how you aim it.
The three scoring dimensions
A useful score answers three separate questions. Keep them distinct in your spreadsheet or the numbers blur together.
1. Intent signals: how badly they want this now
Intent shows up in behavior. A visitor who completes a detailed form, project description, photos attached, preferred time windows, is further along than one who dropped an email into a newsletter box. Search keywords tell the same story: an urgent repair query carries more buying intent than a general cost question. Repeat visits, pricing page views, and a short gap between first visit and form submission all push intent up.
2. Fit signals: whether you can actually serve the lead
Fit is about alignment, not urgency. Is the job address inside your real service area, not just your stated one? Is the job type something your crew handles every week, or a one-off you would subcontract? Does the timeline match your calendar? A perfect-fit lead with low intent belongs in nurture; a high-intent lead outside your territory is a referral you give away, not a job you chase.
3. Value signals: what the job is worth if it closes
Not every booked job deserves the same effort. A full roof replacement, a multi-unit commercial cleaning contract, and a single-room paint touch-up are different businesses inside the same company. Estimate ticket size from the job type and the details given. Many firms see the bulk of their margin come from a minority of job categories, and those categories deserve faster, more personal handling.
A five-point model you can run in a spreadsheet this week
You do not need scoring software to start. A shared spreadsheet with honest rules beats an automated tool nobody tunes. Here is the model we hand owners before any platform conversation:
- Export your last 90 days of leads into a sheet with columns for source, form fields, job address, job type, and outcome.
- Score intent from 0 to 2: 2 for urgent keywords or a detailed, complete form; 1 for a real form with thin detail; 0 for newsletter signups and content downloads.
- Score fit from 0 to 1: 1 for in-area, core service work; 0 for anything else.
- Score value from 0 to 2: 2 for high-ticket or recurring work, 1 for mid-range jobs, 0 for small one-off tasks.
- Total the three columns so every lead lands between 0 and 5, then label tiers: 4 to 5 is hot, 2 to 3 is warm, 0 to 1 is nurture.
- Agree on routing rules with whoever answers the phone, and write a response standard next to each tier so it is a commitment, not a vibe.
Routing: hot leads to closers, low scores to nurture
Score without routing is trivia. Each tier needs a different owner, channel, and cadence.
Tier 1, hot (4 to 5 points): these go to your best closer immediately, by phone, with a five-minute first-attempt standard where staffing allows. If the call does not connect, a text follows within minutes. The goal is a booked estimate while intent is live.
Tier 2, warm (2 to 3 points): these enter a same-day callback queue plus a short follow-up sequence: a call attempt, an email with proof and pricing guidance, a second call. Warm leads convert on persistence and reassurance, not on raw speed.
Tier 3, nurture (0 to 1 points): no human calls. These leads go into an automated email or text sequence that educates and stays present, because a low score today is often a mid-score next quarter when the project is funded or the season turns. Skipping nurture is how companies pay for the same lead twice.
Re-score when new signals arrive
A score is a snapshot, not a verdict. Leads change behavior, and your routing should change with them. Set rules that adjust the score when fresh signals show up:
- Promote a lead one tier when a nurture contact opens three emails in a week or returns to the pricing page repeatedly, and move them into the warm queue.
- Add a point when someone submits a second form or requests a callback, because repeat action is the loudest intent signal a lead can send.
- Add a point when a reply names a timeline or a budget, and route that conversation straight to a closer.
- Apply a decay rule instead of deleting silent leads: no engagement after 60 to 90 days drops the contact into a low-frequency newsletter tier, so your list stays clean without discarding future buyers.
Close the loop: feed booked and won outcomes back into the score
The model gets smarter only if outcomes flow back in. Every month, compare what each score tier actually produced: booked estimates, won jobs, and revenue. Two patterns usually appear. Some signals you assumed meant buying intent, a certain form field or a certain source, correlate with tire-kickers, and their weight should drop. Other signals quietly predict your best jobs and deserve more weight or an automatic promotion rule.
This feedback loop is the difference between a score and a system. Treat the model like a pricing sheet: revisit it quarterly, adjust one weight, and watch close rate by tier for a month before changing anything else.
The measurement mistakes that make scoring look like it fails
Scoring fails on paper more often than it fails in practice. Three measurement errors cause most of the false alarms.
- Judging on volume instead of close rate by tier. Scoring almost always reduces the number of calls your team makes. That is the point. The metric that matters is whether Tier 1 closes at a higher rate than your old average, not whether total touches went up.
- Comparing against a cherry-picked past. Owners remember the one cold lead that became a flagship job. Score the whole lead population over a full quarter before drawing conclusions.
- Changing the model every week. Small samples make every tweak look like genius or disaster. Adjust quarterly, one variable at a time, and let each version run long enough to produce real data.
Start with the spreadsheet. If close rate by tier climbs for two consecutive quarters, you have earned the right to buy software, and you will know exactly what to configure.