All guides
10 min read·Beginner·6 steps·2 screenshotsData and targeting

How to choose a target group that replies

The best copy works no miracles on a bad list. The reply rate is decided mostly by the choice of companies, not by the wording. Here is how to narrow down your target group.

The best copy in the world will not save a bad target group. A campaign to "every company in the UK" gets 0.5% replies; a campaign to "independent garages in Greater Manchester with no online booking" gets 8%. The difference is not luck, it is who you write to.

This guide is about deriving the target from your own clients, narrowing it to one industry and area, and finding the signal that they need you right now.

1

Start from what you already have

Look at your three best clients. What do they have in common? Industry, size, region, the way they work? That is your target group, not the one you dreamed up.

2

Narrow it to one industry and region

A campaign for "all companies in the country" does not work because it cannot be written specifically. "Joineries in one region" does: you know what troubles them and you can mention a reference from nearby.

colldly.com/dashboard › Lead lists
Lead lists in Colldly split by industry and area
Lead lists in Colldly split by industry and area
3

Look for a signal that they need it now

The highest reply rates come from companies with a visible problem you solve. You build websites? Filter by website quality. You do social media? Use the profile audit. A signal turns you from a salesperson into someone who noticed.

4

Consider company size

A smaller company decides fast but has a smaller budget. A bigger one pays more, but more people decide and it takes months. Choose according to what you can handle and stick to it throughout the campaign.

5

Test on a small sample

The first 100 companies are a test, not a campaign. When five reply and two are interested, the segment fits and you can scale. When nobody replies, change the segment, not the subject line.

Tip: Measure replies, not opens. Open rate can be distorted, a reply cannot.
colldly.com/dashboard › Contacts
Contacts in Colldly: a mini-CRM with outreach status across brands
Contacts in Colldly: a mini-CRM with outreach status across brands
6

Write down what you learned

After every wave note which industry and which wording performed. After a few rounds you have your own map of the market that nobody can give you.

What to avoid
Targeting "everyone who might need it". Nobody is everyone.
Mixing industries in one list. Mixed numbers teach you nothing.
Skipping the 100-company test before scaling.
Checklist
  • The target group is derived from your three best clients, not from a feeling
  • One industry, one area, one lead list
  • You have a reason why now (old website, no socials, a new branch)
  • Company size fits your price (a two-person shop will not buy at 5,000, a corporate will not answer cold email)
  • You tested 100 to 200 companies before scaling
  • You write down which industry and phrasing performed
Questions about this guide
How many companies is enough?
For a test, 100 to 200. For a steady flow of leads you need a target group of at least 1,000 companies, otherwise you exhaust it in two months. Google Maps search in Colldly shows how many companies match before you start.
Should I contact my clients' competitors?
Yes, unless you have exclusivity. Companies in the same industry share the same problems and a reference from their industry is your strongest argument. Put current clients on the blocklist.
What if I sell something everyone needs?
Narrow it anyway. "Accounting for everyone" is for nobody. "Accounting for e-commerce shops that outgrew sole-trader status" is a specific person with a specific problem.
Continue with
Next guide in order: How to read campaign numbers
Start today

Colldly does all of this for you

It finds the companies, writes personalised emails, guards deliverability and puts the replies into one inbox. Start for free, no card.

Start for free
No credit card · First 200 credits free · Cancel anytime