Most lead scoring advice is written for companies with a sales team, a marketing team, and ten thousand leads a month. You have none of that. You have a list of forty people and a gut feeling about which three are worth a call. The good news is that the real lead scoring best practices boil down to something a solo founder can do in their head: figure out what your actual buyers had in common, and rank new leads by how many of those things they share. Everything else is dashboards you don’t need yet.
The point of scoring isn’t a tidy number. It’s deciding who gets your limited time today. When you only have an hour to spend on outreach, you want to spend it on the people most likely to say yes, not the ones who happened to reply first. Here’s how to do that without building a machine.
Lead scoring best practices start with your actual customers
Before you score anyone, look at the people who already paid you. What did they have in common? Same company size, same role, same trigger that made them look, same tool they were escaping. Those shared traits are your scoring criteria, and they’re worth more than any generic template because they come from real money changing hands. If your best three customers were all solo consultants who’d just lost a client, then “solo consultant who just lost a client” is a high score, full stop.
This keeps you honest. It’s tempting to score on who seems excited or who has a big logo. But excitement and logos don’t predict payment. The patterns in your existing customers do, and they’re sitting right there in your Stripe dashboard waiting to be noticed.
Weight behavior over attributes
Who someone is matters less than what they did. A perfect-fit lead who never opened your email is worth less than a slightly-off-fit lead who replied twice and asked about pricing. Actions are honest. They cost effort, and people don’t spend effort on things they don’t care about. Score a reply higher than a job title, a pricing-page visit higher than a follow, a “how much is it” higher than a “looks cool.”
For a tiny operation, this is freeing. You don’t need to track forty data points. You need to notice the two or three behaviors that have always come right before someone paid you, and pay attention when a new lead does them.
Keep it simple enough that you’ll actually use it
A scoring system you ignore is worse than none, because it gives you false confidence. Three tiers is plenty: people worth a personal message today, people worth a light touch, and people to leave alone for now. You can sort forty leads into those buckets in ten minutes. The fancy point systems exist to coordinate big teams. Alone, you just need to know who’s hot.
If you want the structure behind the buckets, the common lead scoring models show how the bigger versions work, and you can borrow the one idea that fits your stage and drop the rest of the apparatus.
Scoring only works if the leads were worth scoring
Here’s the catch nobody mentions: scoring sorts the list you have, it doesn’t fix a bad list. If every lead in your pipeline is a poor fit, the best scoring just tells you which bad fit to chase first. Garbage in, ranked garbage out. The real gain is upstream, in who lands on the list at all.
That’s the part we built Unbound Compute around: finding people already showing the exact problem you solve, so the list you’re scoring is full of real candidates instead of names you scraped. You still decide who’s hot. We just make sure the cold ones never crowd out the warm.