
Strategy
Part of Before you shortlist branding agencies: decide whether to commission at all
Branding agencies ninety day plan before selection
Branding agencies ninety day plan: use three gated windows, option cards and bounded decisions to prepare for commissioning work before a pitch or selection.
What to take away
- This ninety-day plan for commissioning branding agencies runs as three 30-day windows, each closed by a gate that must pass before the next opens.
- Every window ends in a bounded verdictpilot, continue the baseline, gather missing evidence, or stop.
- It prepares an evidence-led shortlist before pitches; use this site's directory and reviews to identify candidates, then apply the scorecard below.
- Illustrative roles, internal planning costs and day counts appear below; they are not agency fee benchmarks.
What this plan is, and what it is not
This is a gated planning framework for a branding commission based in England. It prepares a shortlist; it does not replace the directory and review pages on this site. Some rules and guidance cited below are UK-wide rather than England-only.
It is not a pitch either. A pitch asks agencies to present, while this plan asks your own team to produce evidence first.
Days 1 to 30: define, scope and open evidence
The decision owner writes the objective, audience, England-based commissioning context, baseline, constraints and review date. A research lead lists unknowns and ties each one to a method.
GDS guidance on planning user research is written for UK public services, but its question-led approach stops a commercial team treating a persona as a finding.
Write at least four option cards:
Days 1 to 30
- Continue as now.
- Improve the manual arrangement in house.
- Commission external expertise.
- Build the capability internally.
The Green Book covers UK government appraisal, not private purchases, yet its discipline of documenting options, evidence and uncertainty still helps.
Gate 1 — scope ready: pass only when the decision question, baseline, research plan, option cards, evidence owners and unresolved-issue log are approved. If any are missing, gather the evidence or stop rather than opening the next window.
Days 31 to 60: test without launching
Run only approved discovery and controlled evaluation. That includes authorised interviews, a paper review of agency evidence, or synthetic checks of a proposed journey. Research participation never creates permission for sales follow-up.
For each option, log what was asked, what was observed, what was inferred and what stays unknown. Pairing that log with structured customer interviews keeps anecdotes out of the verdict, because the method is fixed before the first call.
The Magenta Book separates process, impact and value-for-money evaluation. Borrow that split: smooth delivery is not proof that brand work caused a business outcome.
Specialist reviewers answer in separate queues: trade marks, advertising claims, privacy and direct marketing, accessibility, security and contracts. The ICO's direct marketing guidance explains how the UK GDPR and PECR apply; it does not approve your contact route.
At Gate 2, score each option against identical option criteria and record the outcome of every applicable specialist review. Keep the evidence and any unresolved issues with the score so the decision owner can give a bounded verdict.
For the external-expertise option, agree an agency scorecard before comparing candidates. Check each candidate's relevant portfolio, case studies and their context, the named team proposed for your work, sector and audience fit, research and strategy approach, and the scope, deliverables and fee assumptions in its proposal. Ask for relevant references and verify any claimed membership of a trade body such as the Design Business Association (DBA); membership is a check, not proof of fit. Apply the same criteria to every candidate.
Example: one gate passes, one gate holds
Illustrative only, not a real commission. Two options reach Gate 2.
Option A, improve the manual arrangement, passes. Interviews ran under the approved consent wording and finance confirmed the cost band.
Option B, commission external expertise, holds. A trade mark search left a name conflict unresolved and no ceiling had been agreed, so one hold blocks progression.
Days 61 to 90: decide and prepare a reversible next step
The decision owner compares surviving options with the current arrangement. Define each measure by population, numerator, denominator, period, source version and uncertainty. Keeping the option comparison honest means listing unknowns as unknown rather than scoring them.
Write a bounded verdict: pilot, continue the baseline, gather missing evidence, or stop. A pilot needs a named owner, an exposure boundary, a rollback method and a decision date.
Check materials against WCAG 2.2, obtain the accessibility and equality review, and ask a security specialist to apply NCSC small-organisation guidance to the real setup.
Gate 3 is not publication approval. Nothing goes live until named authorities accept the evidence and a release owner signs the deployment record.
Internal time and roles across the ninety days
This table covers internal time only; it gives no branding-agency fee benchmark. Request written agency quotes against the same scope, and check whether VAT and expenses are included.
Cost, time and roles
| Workstream | Role | Time over 90 days | Illustrative cost band |
|---|---|---|---|
| Decision and sign-off | Senior marketing lead | 0.2 FTE | Internal time |
| Research and evidence | Research lead plus analyst | 0.4 FTE | GBP 6,000 to 12,000 |
| Specialist reviews | Legal, privacy, accessibility and security advisers | 5 to 8 adviser days | GBP 4,000 to 9,000 |
| Creative build or pilot | Not funded before Gate 3 | Held | Contingency line |
Open the blank budget template before requesting proposals, because it forces VAT and contingency into view. Ask each agency to price the same scope and state fees, expenses and VAT separately. ONS earnings data describes employee earnings, not agency rates, so it is not a like-for-like benchmark.
How this differs from a normal pitch or procurement process
A pitch rewards presentation. A gated plan rewards evidence you already hold. Run the gates first, then select an agency against the criteria the gates produced.
After Gate 3: turn the verdict into agency selection
If the verdict supports external commissioning, begin selection after Gate 3. Set dates for issuing the brief, receiving questions, closing proposals, interviews, reference and due-diligence checks, the decision and contract sign-off. Give every candidate the same brief, information and response window; set dates to suit your capacity rather than assuming a standard timetable.
Set a comparable scope
Depending on the brief, ask candidates to price the same requested work. It might include discovery and research, positioning, naming if needed, verbal and visual identity, a design system, brand guidelines, agreed final files and assets, rollout support and handover. State what is excluded, who approves work, the milestones, revision process and any post-launch support required.
Agree ownership and contract terms
Before signing, specify whether copyright in new work is assigned to you or licensed, and set the licence's territory, duration, media and exclusivity. Identify pre-existing agency materials and third-party fonts, images or software, and agree who obtains and pays for any required licences. Also record fees and payment stages, expenses, change control, cancellation or kill fees, termination and handover terms. Do not assume that paying for commissioned work transfers copyright; have the agreement reviewed.
The Consumer Rights Act 2015 governs contracts for services. The Trade Marks Act 1994 underpins name clearance and licence terms.
Common questions
What must be in place before Gate 1 passes?
An approved decision question, a traceable baseline, a research plan, at least four option cards, named evidence owners and an unresolved-issue log. If the audience has no defensible definition, stop.
How does the plan treat evaluation evidence?
The Magenta Book separates process, impact and value-for-money evaluation. This plan borrows that split, so evidence is logged against identical criteria and delivery quality is never treated as proof of effect.
What happens after the ninety days end?
One of four verdicts: pilot, continue the baseline, gather missing evidence, or stop. If the verdict supports external commissioning, start the agency-selection process after Gate 3: issue a common brief, assess proposals against the scorecard, complete due diligence and agree the contract. Any pilot needs a named owner, an exposure boundary, a rollback method and a decision date.
How does this relate to choosing an agency?
It runs before selection. The gates produce the audience, options and criteria, and any appointment is a later stage with its own contract and due diligence.



