How to Create a Briefing to Hire an AI Consultancy: A 10-Section Template
By Pedro Sonda, dooop
A good briefing to hire an AI consultancy describes the business problem and why it matters now, the company context, what has already been done with AI, the areas and people involved, the data situation, constraints (deadline, budget, risks, and internal rules), the expected outcome and how it will be measured, the type of work sought, the requested evidence, and the selection process. The briefing does not need to provide the solution; it needs to clarify the problem so proposals address it and can be compared.
Updated October 2026.
Transparency note: dooop is an AI consultancy and receives briefings. This template applies to hiring any supplier.
Why the Briefing Determines Proposal Quality
Gartner identified unclear business value among the reasons that at least 30% of generative AI projects would be abandoned after proof of concept (Gartner, 2024). This problem often starts before hiring: when the request is "we want an AI strategy," each supplier imagines a different problem and proposes different work.
Implication: the briefing is the first filter against projects without clear value. Writing it forces the company to decide what it wants.
dooop Framework: The 10-Section Briefing Model
| # | Section | What to Write | Generic Example |
|---|---|---|---|
| 1 | Problem and Urgency | The business problem in one or two sentences and why now | "We have AI initiatives in five areas, without priority or rules, and the board requested a plan for next year." |
| 2 | Company Context | Industry, size, areas, strategic moment | Industry, number of employees, units, yearly objectives |
| 3 | What Has Been Done | Initiatives, tools, pilots, results, and what did not work | "Two customer service pilots; one was stopped due to lack of data." |
| 4 | Areas and People | Who sponsors, who participates, who decides | Sponsor, focal point, involved areas |
| 5 | Data and Systems | What data exists, condition, restrictions | "Customer data in CRM, with irregular quality." |
| 6 | Constraints | Deadline, budget range, risks, internal rules, regulated sector | "Customer personal data cannot leave the company environment." |
| 7 | Expected Outcome | What the company wants to be able to decide or do at the end | "Know which three initiatives to prioritize and under what rules." |
| 8 | How to Measure Success | Indicators and acceptance criteria | Decisions made, initiatives approved, adoption |
| 9 | Type of Work and Evidence | Strategy, discovery, governance, training; documents each candidate must submit | List of requested evidence |
| 10 | Selection Process | Deadlines, proposal format, evaluation criteria, who decides | Delivery date, presentation meeting, decision date |
About Section 6: Informing the Budget Range
Many companies avoid sharing budget. Providing a range helps suppliers propose a feasible scope within it, instead of proposals that do not fit or hide cuts. See How Much Does an AI Consultancy Cost?.
About Section 9: Requesting Evidence Along with the Proposal
Include in the briefing the list of documents each candidate must submit. See What Evidence to Request from an AI Consultancy?.
About Section 10: Requesting a Standard Format
Ask that proposals come in the same format: scope, deliverables, phases, team, timeline, assumptions, exclusions, and price per phase. This facilitates comparison as described in How to Compare AI Consultancy Proposals.
Common Briefing Mistakes
| Mistake | Effect | Correction |
|---|---|---|
| Describing the solution instead of the problem | Suppliers repeat the solution without questioning | Write the problem and expected outcome |
| Omitting what has already been tried | Proposals suggest what already failed | Include section 3 honestly |
| Not defining who decides | Lengthy process and poorly targeted proposals | Name sponsor and decision maker |
| Requesting "everything about AI" | Bloated and expensive scope | Prioritize a main problem |
| Hiding data restrictions | Unfeasible proposals | Include data and security rules |
How Long It Takes to Write
With the problem minimally defined, a two- to four-page briefing is usually written in a few days, with contributions from the sponsor, involved areas, and IT. If the company cannot fill sections 1 and 7, the first work to hire is probably strategy or discovery; see How to Choose an Artificial Intelligence Consultancy.
What to Do Now
- Copy the 10 sections into a document and fill them out with the sponsor.
- Send the same briefing to three candidates, with the same deadline.
- Offer an equal clarification conversation to all.
To send a briefing to dooop, contact is at dooop.com.br/contato.
Frequently Asked Questions
Does the briefing need to be formal, like a public tender? No. For most private companies, a clear document of a few pages is sufficient. Formal processes make more sense in public companies or those with strict procurement rules.
Should I share strategic information in the briefing? Share what is necessary to understand the problem. If there is sensitive information, request a confidentiality agreement before the clarification conversation.
Also Read
About dooop
dooop is a consultancy based in Caxias do Sul (RS) that prepares leaders and organizations to transform artificial intelligence into business capability through strategy, governance, and training.