The wrong way to score leads
Most lead-scoring is theoretical. Someone with a marketing degree picks signals that sound smart — "decision-maker title," "company size," "industry vertical" — without ever calling the prospects to verify what actually predicts buying.
We took the opposite approach. Built the score from what actually predicts whether a local business will engage in a sales conversation, derived from real outbound calls.
The 6 signals that actually matter
1. Reachability (10-20 points)
Has phone, has website, has verified email — these are baseline. A business you can't contact isn't a lead. We weight reachability low (it's table stakes) but missing it tanks the score.
2. Real-business indicator (10 points)
A Google rating means Google has verified the business exists, has customers, has a Maps presence. No rating = could be defunct, could be a duplicate listing, could be a scam. Real-business signal is binary and important.
3. Engagement / review volume (0-20 points)
Review count predicts activity. Banded:
- 100+ reviews = highly active, +20
- 30-99 = established, +17
- 10-30 = active, +13
- 3-10 = early-stage, +8
- 0-3 = unknown / new, +0
High review count doesn't mean "doesn't need help" — it means "has customers, has revenue, can afford services." That's a buying signal, not a disqualifier.
4. Quality (rating tier) (-5 to +15)
Rating predicts business health:
- 4.5+ stars = strong rep, well-run, ideal target, +15
- 4.0-4.5 = solid, +13
- 3.5-4.0 = mixed, +8
- 3.0-3.5 = struggling, +3
- Under 3.0 = likely dormant or bad rep, -5
5. Site health (varies by vertical)
Mobile page speed, SSL status, last update date — these signal whether the business has had any attention paid to its digital presence.
Important: what counts as "good" depends on what you sell. A web designer wants businesses with bad sites (opportunity). An SEO consultant wants businesses with sites that exist but aren't ranking. A POS reseller doesn't care about the site at all.
This is where one-size-fits-all scoring breaks down. Either pick a vertical and tune the weights, or use a tool (like ZipLead) that lets you re-rank by signal.
6. Activity trend (advanced)
Review velocity (growing, flat, declining), ad activity (started, stopped), recent listing updates — these tell you whether the business is in growth or maintenance mode.
Growing businesses have budget and curiosity. Declining businesses are receptive to new ideas because something isn't working. The middle (steady-state businesses) are usually the hardest to convert.
Putting it together: the score
Sum the signals, cap at 99, apply an orphan-listing penalty if everything is missing. You end up with a 0-99 score per prospect that genuinely predicts outreach quality.
In our internal testing on ~1,247 Costa Mesa businesses, the top 38 (score 90+) had a connect rate of 28% vs 12% for unscored cold outreach. Scoring more than doubled connect rate.
Why missing data should NOT penalize
A common mistake: penalizing businesses for missing data. A business with no rating shouldn't necessarily score lower than a business with a present (poor) rating — "missing" is unknown, not bad.
We give partial credit for missing data (10 points for "rating unknown" instead of 0) because in practice, unknown businesses can still be great prospects — they're often new, growing, or under-marketed.
How to use scoring tactically
- 80-99: Hot leads. Contact today. Personal email + phone. These convert.
- 60-79: Warm leads. Add to your 14-day sequence. Don't put energy here until hot leads are exhausted.
- 40-59: Cold leads. Bulk email only. Low effort, low expectation.
- Under 40: Skip. Or use for SEO content if you're trying to build local authority.
Score before contacting, always
The biggest leverage in outreach is doing it on the right list. Scoring takes the difference between "spray and pray" and "surgical outreach." It's worth the 60 seconds.