Every trades owner has a lead source they complain about, and they are often wrong about which one it is. The memorable bad call skews judgment. The scorecard replaces memory with a few seconds of scoring per enquiry and, after about 60 days, a clear table of which sources pay. This guide shows how to set it up in a spreadsheet and how to act on what it tells you.
Why cost per lead misleads
Answer first: the number that matters is cost per won job, not cost per lead, because cheap leads that never become work are not cheap.
Take two sources over the same two months:
| Source A | Source B | |
|---|---|---|
| Spend | $1,500 | $1,500 |
| Leads | 60 | 25 |
| Cost per lead | $25 | $60 |
| Jobs won | 4 | 7 |
| Cost per won job | $375 | about $214 |
Source A looks like the bargain on a cost-per-lead basis. On cost per won job, Source B is far cheaper, and it also used much less office time, because there were 35 fewer conversations. Most owners who switch to judging by won jobs end up moving money between sources they were already using.
The three marks
Answer first: score each enquiry 0, 1 or 2 on fit, timing and budget signal, for a total from 0 to 6.
Fit (is this our work?)
- 0: a job you don't do, or outside your service area
- 1: borderline, such as a small job below your usual minimum or an area at the edge of your range
- 2: exactly the work you want more of, in your area
Timing (when do they want it?)
- 0: more than six months away, or "just getting ideas"
- 1: within one to six months
- 2: now or within a few weeks
Budget signal (what did they ask about?)
- 0: only asked the price and nothing else
- 1: asked about price and something else
- 2: asked about quality, warranty, timing, experience or process
The budget mark is the most subjective, so keep the definitions pinned near the phone. It is not about judging people; it is about noticing whether they are buying a job or shopping a number.
Setting up the spreadsheet
Answer first: one row per new-customer enquiry and about ten columns are enough. You don't need special software.
Columns, left to right:
- Date
- Source (pick from a fixed list: Google Business Profile, search ads, Local Services Ads, website organic, referral, lead platform, social, other)
- Job type (from a short fixed list of your services)
- Area (town or postal-code prefix)
- Fit (0–2)
- Timing (0–2)
- Budget signal (0–2)
- Total (the sum of the three)
- Outcome (no quote, quoted, won, lost)
- Job value (if won)
Keep the source and job-type lists fixed. Free typing produces "google", "Google", "GMB" and "maps" as four different sources, and the totals stop adding up.
At the end of each month, build a summary by source: number of leads, average total score, number won, total spend for that source, cost per lead, cost per won job and average job value. Most spreadsheet programs can do this with a pivot table or a few SUMIF and COUNTIF formulas.
A note on privacy: the scorecard doesn't need names or phone numbers, so leave them in your job-management system and keep this sheet anonymous. Canadian private-sector privacy law, summarised by the Office of the Privacy Commissioner, expects businesses to collect only what they need, and an anonymous sheet is simpler to look after.
Reading the results after 60 days
Answer first: look at three numbers per source, in this order: average score, cost per won job and average job value.
- Low score, high cost per job: the source is sending the wrong people. Fix targeting or reduce spend.
- High score, high cost per job: the right people are coming, but you're losing them later. Check answer speed, quote follow-up and reviews.
- High score, low cost per job: your best source. Find out whether it can take more budget without losing quality.
- Low score, low cost per job: cheap but noisy. Often a referral or organic source with mixed enquiries; usually leave it alone and filter at the phone.
Also compare job value by source. A source that wins fewer jobs but much larger ones can still be the most profitable.
Fixes for a badly scoring source
Answer first: match the fix to which mark is dragging the score down.
Low fit scores
- In search ads: review the search terms report, add negative keywords for jobs and areas you don't serve, and tighten the location settings.
- On your website and Google profile: list your services clearly, list what you don't do, state your service area, and show a starting price or minimum if your trade allows.
Low timing scores
- Remove "ideas" and "inspiration" wording from ads for urgent trades.
- For project trades, keep early-stage leads but move them to a slower follow-up path rather than sending a crew to quote.
Low budget-signal scores
- Add proof to the pages those leads land on: photos of real jobs, a short process explanation, warranty details and recent reviews.
- Avoid "cheapest in town" wording; it attracts exactly the leads it promises.
After a change, keep scoring. Judge the fix on the next 30 to 60 days, not the next week.
Common mistakes
Answer first: most scorecards fail because they stop being filled in, not because the method is wrong.
- Scoring only some calls. If only the bad ones get logged, every source looks bad. Score every new-customer enquiry.
- Changing the definitions mid-way. Keep them fixed for at least two months.
- Too many sources. Ten categories at most. Put rare ones under "other".
- Cutting a source on one bad week. Small numbers swing wildly. Wait for 20 to 30 leads per source.
- Forgetting office time. A source that wastes five hours a week of a coordinator's time has a real cost even if the ad spend is low.
What to do next
Set up the sheet today and start scoring tomorrow. If you're not sure lead quality is your main problem, run the self-diagnosis guide first. If the scorecard confirms a bad source, the bad-leads fix is the done-for-you version of everything above. Results from our HVAC work, including HVACDeal.ca and its 12:1 ROI, are on the results page. If you would like a second pair of eyes on your first 60 days of data, call, text or email it to us.
Frequently asked questions
How many leads do I need before the scorecard is reliable?
Aim for at least 20 to 30 enquiries per source. For smaller sources that can take longer than 60 days; keep scoring until you reach that number before drawing conclusions.
Who should score the leads?
Whoever takes the first call or message, at the time. Scoring later from memory is less accurate and tends to get skipped.
Should I score repeat customers?
Keep them in a separate column or leave them out. The scorecard is for judging how well marketing attracts new customers, and repeat work would flatter every source.
Is a high-score lead that didn't buy a failure?
No. It means the source is sending the right people and the loss happened later: in the quote, the follow-up or on price. That points to a different fix.