Names, companies and quoted comments are altered to protect the people involved. The numbers, scores and method are exactly as they happened.
A Canadian business-process consultancy came to us with a clear problem. For most of its 30 years, new clients had come through referrals, word of mouth, and cold calling. It now wanted a repeatable way to find clients through LinkedIn because that was where it believed more of the right business owners were spending time.
The market was unusually narrow. The consultancy serves owner-led Canadian manufacturers and trades businesses with 5 to 50 employees. It did not need a global list of people talking about business problems. It needed local owners it could actually help.
The test was simple: could Signado find those Canadian owners and operators, and show why each person was worth contacting?
Signado finds the people already talking about the problem you solve on LinkedIn.
This LinkedIn lead generation case study shows what happened next: why three keyword searches found the wrong people, why we switched to a creator, which people were worth keeping, and which people Signado rejected. Nothing has been turned into a success story after the fact.
Visit the Signado homepage for the product overview. The full method is in how to find people discussing problems on LinkedIn. For the wider category, see the social selling tools comparison.
Why the keyword searches were noisy
The consultancy started with pain-language keywords. The phrases described the operational problems it solves for small manufacturers and trades businesses.
The first problem was location. LinkedIn keywords are global. A phrase used by a Canadian manufacturer can also be used by someone in the United States, Europe, Asia, or anywhere else. The words matched, but many of the people were outside the market the consultancy could serve.
The second problem was who used those words most often. Three batches brought back consultants, vendors, and other people who talk publicly about those problems because they sell services around them. They were peers or competitors, not manufacturing owners who had those problems.
The keywords were relevant. The people were not. A keyword can find the right conversation and still find the wrong people.
The answer was not to add more keywords. It was to change where the search started.
One creator source changed the audience
We switched to a Canadian manufacturing founder with an engaged operator audience.
Signado suggested this creator from the consultancy's business profile. The suggestion included a verified personal LinkedIn URL, so we could place that profile directly into the creator source field.
You do not have to use Signado's suggestions. You can ask Claude Code or Codex to research creators whose audiences contain the buyers you want, verify the person's public LinkedIn profile, and paste that personal LinkedIn URL into Signado.
What matters is the audience. Pick a person whose posts already attract the owners or operators you want to meet. In this case, the choice put the search around people running manufacturing, food, restaurant, and goods businesses. The creator's audience was closer to the consultancy's market before Signado scored a single person.
We never disclose which creators or competitors a workspace monitors. Not here, not anywhere.
The run happened in one afternoon during August 3 to 4, 2026, after the weak keyword batches. Signado checked the creator's recent public LinkedIn posts and found the people engaging with them.
The method stayed simple:
- Define the owners and companies the consultancy can help.
- Choose a Signado suggestion or research a suitable creator with Claude Code or Codex.
- Verify the person's public LinkedIn profile and add its URL as a creator source.
- Let Signado find the people engaging with recent posts.
- Check the role, company, location, comment, score, and buyer stage.
One afternoon produced 37 people
The creator run found 37 people.
Roughly half were outside the consultancy's target market. That is a normal cost of searching an audience instead of a fixed contact list. A creator can have the right core audience and still attract people from other countries, company sizes, and professions.
Country rules now handle one part of that waste automatically. When Signado finds that a new person is in an excluded country, the five discovery credits are refunded in full and the person is hidden.
The buyer stages were also modest:
- Most of the 37 were at Early interest.
- Two were Interested.
- None were Ready to buy.
That is the base rate of a real audience. Most people who engage with a relevant post are interested in the subject. Very few are publicly asking for a provider at that moment.
[SCREENSHOT: warm leads table, names/companies/avatars blurred, scores and stage badges visible]
Five Canadian operators were worth keeping
Five people were close enough to the consultancy's market to deserve a careful review.
| Role | Company size | Industry | Province | Score | Stage |
|---|---|---|---|---|---|
| Co-founder and COO | 5 people | Consumer products manufacturing | Ontario | 67 | Not supplied |
| Supply chain director | 4 people | Food manufacturing | Alberta | 64 | Interested |
| Founder | 11 to 50 people | Restaurant group | Not supplied | 63 | Not supplied |
| President | 5 people | Not supplied | Ontario | 59 | Not supplied |
| Owner | Small business | E-commerce goods | Not supplied, Atlantic Canada only | 53 | Not supplied |
Only one of these five records came with a named stage in the supplied lead-level pack. The other four are marked Not supplied rather than guessed.
The highest score was 67. That is useful, but it is not a claim that the person is ready to buy. The score says the role, company, location, and public activity are close enough to the consultancy's market to justify a human review.
The rejected people show why the score matters
A list of 37 names would have hidden the most important part of the result: who should not be contacted.
Signado scored a direct competitor consultancy at 22 because it sold similar services. It understood the problem, but it was a peer, not a buyer.
An employee at a company with more than 1,000 people scored 19. The person had joined a relevant conversation, but the company was far outside the consultancy's target of 5 to 50 employees.
The creator's own reply was excluded. A source should not discover itself and present that person as a lead.
Those rejections are part of the proof. Relevant words alone are not enough. The person and company still have to match.
One comment supplied the opener
The clearest result was the supply chain director at the four-person food manufacturer in Alberta. She scored 64 and reached Interested.
Her comment, paraphrased, said that she believed in the product but the business was not profitable enough yet.
That is already something real to say. The consultancy does not need to invent a reason based on her job title or send a vague compliment about the post.
A safe paraphrased opener could start like this:
You mentioned that you believe in the product, but the business is not profitable enough yet. When that happens in a small manufacturer, where do you feel the pressure most today: purchasing, production, or the handoff between them?
The message reflects the problem she described and asks one relevant question. It does not claim she asked for consulting help. It does not pretend she is ready to buy.
That is the product's whole argument. Find the right person, keep what they said beside their profile, and start with something real.
[SCREENSHOT: one lead drawer, identity blurred, stage badge and score reason visible]
Keep the limits beside the numbers
This batch had no high scores. The highest was 67.
The consultancy serves owners in businesses where most work happens offline. Those owners do not publish detailed operational problems on LinkedIn every day. Signado therefore has less public evidence to score, and the numbers tend to top out lower than they would in a market that talks about its work online all day.
The other limits are just as important:
- Signado found people from public LinkedIn activity only. Quiet buyers who did not post, comment, or react could not appear.
- This run did not include email sending. It measured who might be worth contacting and why, not replies or meetings.
- Results depended on the source. Three keyword batches found peers and vendors. One creator source found the five strongest operators.
- Two Interested people and zero Ready to buy people means this was a discovery result, not a list of people asking for a proposal.
- Five promising operators from one afternoon does not prove that every creator will produce the same result.
The next proof is contact. The consultancy can reach out to the Interested lead using the point she already made, then record whether that conversation becomes a reply, meeting, or client. Until that happens, this page claims discovery only.
Use the same test in another market
The useful lesson is not that creator sources always beat keywords. It is that the source determines who Signado finds.
Pain-language keywords can attract people who sell around the problem. A creator with the right operator audience can bring the search closer to the people who live with it. Signado then checks the role, company, location, activity, score, and stage so the wrong people do not look equal to the right ones.
Start with one source. Read the weak matches. Check the rejected people. Then decide whether the keepers give you something relevant to say.
Signado finds the people already talking about the problem you solve on LinkedIn.
Start Free and test one creator, competitor, keyword, or public LinkedIn post against the market you serve.
Three-minute video outline
| Time | What to say | What to show |
|---|---|---|
| 0:00 to 0:25 | "The first three batches failed. Our keywords found consultants and vendors who talked about the problem, not owners who had it." | The problem and the anonymous consultancy description |
| 0:25 to 0:50 | "We changed one thing. We added a Canadian manufacturing founder whose audience includes operators." | Anonymous creator-source card with every identity hidden |
| 0:50 to 1:15 | "One afternoon produced 37 people. Most were Early interest, two were Interested, and none were Ready to buy." | Blurred warm leads table with scores and stages visible |
| 1:15 to 1:50 | "Five Canadian operators were worth a closer look. The strongest score was 67. A food manufacturing supply chain director scored 64 and reached Interested." | Five anonymous role, company-size, industry, location, score, and stage rows |
| 1:50 to 2:15 | "The rejects matter too. A competitor scored 22. An enterprise employee scored 19. The creator's own reply was removed." | Rejection reasons with no identities |
| 2:15 to 2:40 | "The Interested lead said, in paraphrase, that she believed in the product but profitability was still a problem. That gives the consultancy something real to say." | Blurred lead drawer with stage and score reason visible |
| 2:40 to 3:00 | "No high scores. LinkedIn only. No email sending in this run. The source determined the result. Next, we measure whether the relevant opener earns a reply." | Limits, then Start Free |
Land your next client with something relevant to say
Signado finds the people already talking about what you do on LinkedIn, and shows you what they actually said. You reach out with a reason they'll recognize.