You don’t need an advanced platform to prioritise leads. A simple points model in a spreadsheet goes a long way, provided it’s built on the right signals.
Lead scoring sounds like something that needs a marketing automation platform, a data analyst and six months of implementation. For large organisations with thousands of inbound leads a month, that may be true. For a sales team of two to ten people doing outbound prospecting, it's overkill.
But the need is still there. You have more companies on the list than you have time to contact, and you want to start with the ones most likely to become customers. A simple points model helps with exactly that.
Two kinds of signal
There are two basic types of information that say something about how good a lead is:
- Fit – how closely the company matches your best customers. Industry, size, revenue, geography. You know this before you make contact.
- Interest – how engaged the company is. Have they replied to an email, booked a meeting, asked about pricing? You only learn this once you've started talking.
A good model keeps the two apart. A company with perfect fit but no interest is a good lead to contact. A company with lots of interest but poor fit can be a trap – they may want something you don't really deliver.
Build the fit score from your customers
Start with the same exercise as in the article on customer profiles: look at your best customers and see what they have in common. Then award points for each characteristic.
An example for a company selling to wholesalers:
- Industry: SNI 46 (wholesale) = 3 points, adjacent industry = 1 point, other = 0
- Size: 20–99 employees = 3 points, 10–19 or 100–199 = 1 point, other = 0
- Revenue: above SEK 50 million = 2 points, SEK 20–50 million = 1 point
- Geography: within your area = 2 points, neighbouring county = 1 point
The maximum is 10 points. Companies scoring 8–10 are top priority, 5–7 are worth contacting, and below 5 go last.
What matters isn't the exact numbers but that they reflect what you actually know about which customers work. Don't guess – start from your history.
Add interest points once you start talking
Once contact is established you can add interest signals:
- Replied to an email or call: +2
- Asked for more information about something specific: +2
- Booked a meeting: +3
- Asked about price or terms: +3
- Referred you to a colleague with decision authority: +2
- Said “not now, but get back to me in [month]”: +1 and a follow-up in the calendar
And negative points for clear signals the other way, such as “we've just signed a three-year contract with a competitor”.
Keep the model simple enough to use
The most common mistake is building a model that's too complicated. If it takes five minutes to work out a lead's score, nobody will do it. If the model has twenty variables, nobody will understand why a given company got its score.
Three to five fit variables and half a dozen interest signals are enough. You can often calculate the fit score automatically in Excel with IF formulas, straight after exporting the list. Interest points are set by hand in the CRM or as tags.
Check that the model holds
A scoring model is a hypothesis. After a quarter you should be able to answer a simple question: did high-scoring companies become customers more often than low-scoring ones?
If yes, good. If no, or if the difference is small, there are two possibilities. Either the model weights the wrong things, or the characteristics that really decide it aren't visible in the data you use – such as whether the company recently changed leadership or is about to expand.
Adjust the weights, but don't change the model every week. You need enough deals to see a pattern, and that takes time.
The point is to rule things out
The biggest gain from lead scoring isn't finding the best leads but daring to set the worst aside. Almost every salesperson spends too much time on companies that will never buy, because they happen to be pleasant on the phone or because giving up feels wrong.
A clear model makes it easier to say: this company scores 3 out of 10, we'll park it and come back in six months. That time can go to the companies that score 9 instead.