Somewhere along the way, lead scoring models turned into a spreadsheet with eleven columns that nobody fills in. Demographic points, behavioral points, a decay function, a threshold that triggers a handoff to a sales team you don’t have. It’s machinery built for a company with a marketing department and a quarterly target. You have neither. You have twenty minutes a day and a list of names, and you need to know which ones to actually message.
So forget the enterprise version for a minute. The point of any lead scoring model is one thing: spend your limited attention on the people most likely to say yes. That’s it. Everything else is decoration you can’t afford yet.
What the big lead scoring models get wrong for you
The classic models score two things and add them up. Fit, meaning how much someone looks like your ideal customer on paper. Engagement, meaning how many emails they opened and pages they viewed. Sounds reasonable. The problem is that both halves lean on data you don’t have at your size, and both miss the thing that actually predicts a reply.
Opening three emails isn’t intent. It’s a habit, or boredom, or an inbox left open in a tab. A founder who matches your profile perfectly but is perfectly happy with what they use is not a hot lead. The enterprise model would score both of those people high and waste your week on them.
The two-factor version that fits a tiny team
Score two things, and weight them honestly. First, fit: does this person match the kind of customer your thing is genuinely built for? Be strict. “Close enough” is a no. Second, and this is the one people skip, timing: is there a recent, real signal that they need you right now? Something they said or did in the last few weeks, not a vague interest from last year.
- Fit, no signal. Right person, no reason to talk today. Park them. Watch for a trigger.
- Signal, no fit. Loud need, wrong person. Tempting, and a trap. Skip it.
- Both. Right person, real recent reason. This is the only bucket worth your twenty minutes.
Run a list through that and most of it falls away, which is the point. The names that survive are the ones you can plausibly start a real conversation with today.
Timing is the factor everyone underweights
Most lead scoring models treat timing as an afterthought, a small engagement bump if someone visited your pricing page. But timing is half the whole thing. The same person is a cold lead this month and a hot one the week their current tool breaks, or the week they post “anyone know a better way to do X?” Nothing about them changed except the moment. A model that can’t see the moment is scoring the wrong variable.
If you want the deeper version of why this beats the firmographic checklist, it lives next door in what lead scoring actually is for a small team. The short version: a perfect-fit name with no live reason to talk is not a lead this quarter, and pretending otherwise is how founders burn through a list feeling busy and getting nothing back.
Keep the model in your head, not a CRM
You don’t need software for this yet. The whole model is two questions you can ask in the time it takes to read someone’s last post. Right person? Real reason, right now? Two yeses and you write to them. One no and you move on without guilt. That discipline is worth more than any scoring field, because it keeps you out of the trap of half-personalizing a message to someone who was never going to care.
The hard part isn’t the scoring. It’s spotting the timing signal without reading the whole internet every morning. That’s the job we built Unbound Compute to do: watch for the recent, real reasons your buyers show they need you, and hand you the names that already pass both questions. You still decide who’s worth it. We just stop you from scoring a thousand people by hand.