Foundations
Part of Setting evidence rules for a branding-agency directory
Four public signals that may point to branding demand, and their limits
Four unranked signals that may justify branding research in England, with official sources, selection rules and limits on interpreting demand.
Branding agencies demand signals can justify a closer buyer investigation, but they cannot prove that work has been commissioned. Four current records show activity around new firms, intellectual property, creative businesses and public briefs. Each measures a different population, mostly across the UK, so the entries are deliberately unranked.
What to take away
- Demand signals can justify a closer buyer investigation but cannot prove that branding work has been commissioned.
- Each of the four records measures a different population, so the entries are deliberately unranked.
- A signal becomes useful only after a prospective buyer confirms its problem, authority, timetable and budget.
- Public notices provide the most concrete signal because they record an actual procurement process.
- If enquiries do not meet the service definition, stop describing the indicator as demand.
How the signal set was chosen
Research date: Records were checked on 5 September 2026.
Signal inclusion test
- UK statistics producer, department or authority
- Dated event creating naming question
- Exclude surveys, trends, directories, awards
- No funding or commercial influence
- Sequence is practical, not a score
Scope: records from a UK statistics producer, government department or contracting authority that an England-focused agency can retrieve again.
Inclusion test: the record must describe a dated event that could create or reveal a naming, identity or brand-governance question.
Exclusions: agency surveys, search trends, directories, award entries, testimonials, forecasts and supplier claims. No organisation funded the review, and no commercial relationship influenced inclusion.
The sequence below is practical, not a score. A signal becomes useful only after a prospective buyer confirms its problem, authority, timetable and budget.
1. Registered business births
The ONS Business demography 2024 bulletin, released on 20 November 2025, reports births and deaths among enterprises registered for VAT or PAYE, or at Companies House. Regional tables include the nine English regions, with counts rounded to the nearest thousand. The underlying population omits many unregistered businesses.
A birth may prompt naming or identity work, yet it may equally be a dormant vehicle, a continuation under a new entity or a firm with no external launch. Use local births to choose where to conduct interviews, not to calculate agency demand.
2. Trade mark applications
The Intellectual Property Office published its 2025 patents, trade marks and designs facts on 30 July 2026. The statistics come from UK registered-rights administration and carry the publisher's warning that IP data should not be used alone as a measure of innovation.
An application proves that somebody sought a registered right for specified goods or services. It does not identify an England-based buyer, an external agency appointment or the amount spent on brand development. A useful follow-up is to ask applicants how the name and identity were developed, using a properly designed sample.
3. Creative business locations
DCMS counted business sites as local units in its Creative Industries business demographics. The March 2025 snapshot found the sector concentrated in London, but the definition spans many creative industries and the register mainly covers incorporated, VAT-registered or PAYE-registered organisations.
This can guide regional prospecting or event research. It cannot show where branding clients are located, because a supplier site and a buyer's purchasing need are different units.
4. Published brand briefs
Public notices provide the most concrete signal because they record an actual procurement process. A February 2026 Jisc brand and production tender described externally supplied brand graphic-design work within a wider UK framework. It is one buyer's notice, not a market sample, and its estimated value must not be generalised to private assignments.
The Cabinet Office's central digital platform guidance, updated on 17 August 2026, explains that Find a Tender publishes notices from contracting authorities in one place. Search results can form a dated brief log, provided duplicates, amendments, cancellations and out-of-scope lots are recorded.
Turn a signal into evidence
Assign an owner to each hypothesis. Record the source date, geography, buyer type and the event that would disprove it. Proceed only when interviews, live briefs or signed transactions confirm a recurring problem. If enquiries do not meet the service definition, stop describing the indicator as demand.
Before you act
- Check the record was published by a UK statistics producer, department or contracting authority.
- Confirm the record describes a dated event that could raise a naming or identity question.
- Exclude agency surveys, search trends, directories, award entries and supplier claims.
- Ask the prospective buyer to confirm problem, authority, timetable and budget.
- Record source date, geography, buyer type and a disproof event.
- Proceed only when interviews, live briefs or signed transactions confirm a recurring problem.
Common questions
What makes a demand signal useful rather than just interesting?
The article says a signal becomes useful only after a prospective buyer confirms its problem, authority, timetable and budget. Each record measures a different population, mostly across the UK, so the entries are deliberately unranked. The sequence is practical, not a score, and cannot prove that work has been commissioned.
Why are public brand briefs the most concrete signal?
Public notices record an actual procurement process. The February 2026 Jisc brand and production tender described externally supplied brand graphic-design work within a wider UK framework. However, it is one buyer's notice, not a market sample, and its estimated value must not be generalised to private assignments.
How should a reader turn a signal into evidence?
Give each hypothesis an owner and note its source date, geography, buyer type and a disproof event. Only move ahead once interviews, live briefs or signed deals show a recurring problem.
