01
Choose the questions.
Set 10 to 15 buyer questions relevant to your category. Keep the set consistent so the next run can be compared.
Signal Watch / Monthly intelligence
A monthly written record of whether AI answers name you,
and what moved since last month. $290/month pilot.
The same buyer questions. A new monthly record. See whether your name appears, what the answer says, and where the next change belongs.
A written comparison, not a promise of AI rankings. Our published baseline uses one offline model and five fixed prompts.
A clear reading of the evidenceFixed buyer questionsSource answers retainedHuman next steps
The public baseline
Our own brand.
An honest starting point.
On October 3, 2026, we ran five fixed prompts on gpt-oss-120b. The model answered from training data, without web access. The questions used our former name, Pillar Fabric.
Visibility starts with being understood.
None of the five prompts correctly named or described the former brand. One confused Tenda with a fabric line. That is a baseline to inspect, not a hidden score to trust.
Historical wording is kept below so the report is reproducible. Pillar Fabric is now Shuttlewright. These results are not current live search results.
Inspect the prompt record ↓Baseline / October 3, 2026Offline model
What is Pillar Fabric?Not known
Security questionnaire tools?Not named
AI visibility audit services?Not named
What is Tenda?Wrong entity
African social seller tools?Not named
The report, opened up
A score should never
hide the answer.
Keep the question, observation and proposed change in the same record. Monthly comparisons use the same prompt set.
| Buyer question | Month 01 | What the baseline found | Proposed next step |
|---|---|---|---|
| What is Pillar Fabric and what does it offer? | Not known | No information about the brand. | Publish crawlable identity pages with evidence. |
| Tools for security questionnaires and RFPs? | Not named | Vanta, Drata, Secureframe and others were named instead. | Check live search citations separately from this offline run. |
| Services that audit how often a company appears in AI answers? | Not named | Only model provider dashboards were listed. | Publish the audit method as a citable guide. |
| What is Tenda by Pillar Fabric? | Wrong entity | Tenda was described as a fabric line. | Make its identity as social selling software explicit. |
| Tools for African social sellers to turn chats into order drafts? | Not named | Flutterwave, Paystack and others were named instead. | Publish useful, sourced African commerce content. |
Baseline scope: five prompts, one provider. It is not evidence of how every AI assistant describes the brand.
The monthly method
Ask consistently.
See the difference.
A useful report makes the observation and the explanation easy to tell apart.
02
Keep the source.
Record the public answer, mentions, position and cited sources. Note the date and the engine used.
03
Make a useful change.
Read the written comparison and a short list of next steps. You decide what to publish or change.
The monthly plan
Don't lose the thread
after an audit.
An audit gives you a starting point. Monitoring keeps a consistent set of questions in view as your content and category change.
Need a one time baseline? Explore Signal Audit ↗Signal Watch
$290 / month
Pilot price. Review the final checkout terms before purchasing.
- A fixed set of buyer questions
- A written monthly comparison
- Practical next steps to review
- No tracking code or account access required
No rankings, citations or demand are guaranteed.
Before you begin
Clarity about
the scope.
Is the published baseline a live web test?
No. It is an October 3, 2026 run on gpt-oss-120b, an offline model answering from training data. It is kept as a historical baseline.
Does this guarantee a mention in AI answers?
No. The report records observations and suggests changes. AI rankings, mentions, citations and demand are not guaranteed.
How is monthly monitoring different from an audit?
The audit creates a starting point and a fix list. Monitoring repeats a consistent question set and explains what moved between runs.
Do you need access to my accounts?
The monitoring scope shown here does not require account access or tracking code.
Which engines will be included?
The engines and question set should be agreed for your report. The public example on this page is only one offline model.
Where can I review the purchase terms?
Review the final plan and terms at checkout. The terms and refund policy remain available before purchase.
Record the question.
Keep the answer in view.
Keep the answer in view.
A report is evidence to inspect, not a promise to believe.