A customer profile that says “small and medium-sized businesses” cannot be searched. Here is how to turn your best customers into filters you can actually use.
Most sales teams have an ideal customer profile. The trouble is that it usually reads something like: “Small and medium-sized businesses that value quality and want to grow.” That works on a strategy slide, but you cannot search for it. Every company in Sweden wants to grow, and none of them describe themselves as indifferent to quality.
A useful customer profile is one you can type into a search box and get a list back. In Sweden you have an unusual advantage: the business register kept by Statistics Sweden (SCB) describes every company using the same set of variables. Once you learn to think in those variables, your profile becomes something you can test rather than something you debate.
Start with your twenty best customers, not a workshop
Pull out the customers that generated the most revenue over the past two years – or, better still, the ones that were easiest to close and have stayed. Twenty is enough. Look each one up by organisation number and record five things in a spreadsheet:
- SNI code (industry), ideally at the five-digit level
- Size class by number of employees
- Revenue class
- Municipality and county of the registered office
- Year founded
This takes about an hour, and the result is often surprising. Picture an IT consultancy convinced its customers are “manufacturing”. Once the list is in front of them, almost half turn out to be wholesalers in SNI 46 – businesses with warehouses, thousands of item numbers and an ERP system that has aged badly. The manufacturers are there, but they are not the core. That kind of discovery is the whole point of the exercise.
Look for patterns, not averages
With twenty rows in front of you, look for clusters rather than calculating a mean. An average of “35 employees” can hide the fact that half your customers have 10–19 staff and the other half have 100–199. Those are two different buying processes, two different decision makers and probably two different messages.
Questions that usually produce answers:
- Which two or three SNI codes cover more than half of your customers?
- Is there a size class where you almost never win, even though you keep trying?
- Are customers clustered geographically – and is that because of you (you have an office there) or because of them (the industry is concentrated there)?
- Is there a founding year before which no customer was started? Very young companies often lack budget; very old ones may have locked-in supplier relationships.
Understand what the classes really mean
SCB publishes employees and revenue as ranges. The class “20–49 employees” is not an exact figure, and it counts registered employees – not contractors, agency staff or owners who take dividends instead of salary. A holding company can show zero employees and still be the parent of a group with hundreds.
In practice that means two things. First, treat the classes as coarse filters rather than truths. Second, always combine size with industry. “0 employees” in SNI 64 (financial services) means something entirely different from “0 employees” in SNI 43 (specialised construction), where it is usually a lone tradesperson.
Write the profile as a filter
Once the patterns are clear, rewrite the profile so that it can be run. Compare:
Before: “Mid-sized trading companies in western Sweden that need to digitise.”
After: “SNI 46 (wholesale), 20–99 employees, revenue above SEK 50 million, registered in Västra Götaland or Halland county, founded before 2015.”
You can search for the second version today. It returns a list with a concrete number of companies, and that number tells you something on its own. If it is 40 companies, the market is too narrow to sustain a sales team. If it is 14,000, the profile is too broad for your message to be specific.
Test the profile before you believe it
A customer profile is a hypothesis. The simplest test is to take a hundred companies that match the filter and call or email them over two weeks with one consistent message. Then do the same with a hundred companies that don't match – for instance the same industry but a different size class.
If reply rates and meetings booked are not noticeably different, the profile has not captured what makes a good customer. Most likely a variable is missing that the register does not hold: a particular ERP system, a newly hired CFO, an expansion. That is not a failure; it is information. Register data gives you the right pond to fish in; you find the signals within that pond through conversations.
Keep the profile alive
Repeat the exercise every six months. The customers you won in the last half-year are the best evidence for adjusting the profile, and it commonly shifts as the product matures. A company that started out selling to one-person businesses often finds after a couple of years that the best deals come from companies with 10–49 employees, because the product has gained features only larger teams need.
Save your filters so you can run them again and see how the list changes. New companies are registered every day, and the ones that grow into your size class are often the most receptive – they have just acquired problems they did not have last year.
If you want to try it now, start in the company search and combine industry, size, revenue and location. It takes a few minutes to see whether your profile returns 50, 500 or 5,000 matches – and that number decides how you should plan the rest of your prospecting.