AI Consulting: What to Evaluate Before Hiring (with an 8-Criteria Scorecard)

By Pedro Sonda, dooop

Before hiring an AI consulting firm, evaluate eight points: understanding of your business, proven experience with similar problems, the team that will actually work on the project, methodology, how results are measured, approach to risks and data, capacity transfer to your team, and commercial clarity. Assign weights to these criteria before meetings and score each provider the same way; this prevents choosing the one who presents best instead of the one who delivers best.

Updated October 2026.

Transparency note: dooop is an AI consulting firm. The scorecard in this article is designed to evaluate any provider, including dooop.

Why Evaluate Business and People Before Technology

Evidence about AI results points to organizational factors. BCG observed that companies generating the most value from AI dedicate about 70% of effort to people and processes, 20% to technology and data, and 10% to algorithms (BCG, 2024). The MIT NANDA project report concluded that 95% of corporate generative AI pilots did not generate measurable financial return and identified organizational barriers, more than technological ones, as the main obstacle (The National CIO Review on MIT NANDA, 2025).

Implication: a consulting firm that only demonstrates technical mastery covers the smallest part of the problem.

dooop Framework: The 8-Criteria Scorecard

#	Criterion	What to Observe	Warning Sign	Suggested Weight
1	Business Understanding	Asks questions about goals, processes, and constraints before proposing	Ready-made proposal before understanding the problem	15%
2	Comparable Experience	Cases with problem, approach, and results in a similar context	Only logos, no explanation	15%
3	Project Team	Names, roles, and dedication of those who will work	Seniors in sales, unknown in execution	15%
4	Methodology	Explicit steps, deliverables, and decision criteria	Vague methodology or only tool-based	10%
5	Result Measurement	Business indicators defined with you	Success measured by "deliverables completed"	15%
6	Risks, Data, and Governance	Data handling, LGPD compliance, security, use of AI by the provider itself	Generic answers about confidentiality	10%
7	Capacity Transfer	How your team learns and takes over the work	Permanent dependency on the provider	10%
8	Commercial Clarity	Clear scope, assumptions, pricing model, and exit conditions	Open scope with fixed price, or vice versa	10%

Adjust weights to your case. A governance project may give more weight to criterion 6; a training project, to 7.

How to score: give 1 to 5 points for each criterion, multiply by the weight, and sum. Each scorer should do this individually before group discussion to reduce the influence of the strongest opinion in the room.

How to Verify Each Criterion in Practice

Criteria 1 and 5: ask the provider to describe, in their own words, the problem you presented and how they would measure success. If they cannot, the proposal is not anchored in your business.
Criteria 2 and 3: request detailed cases and speak with references. See How to Evaluate AI Consulting Experience.
Criterion 4: ask for an example of a real, anonymized deliverable.
Criterion 6: ask how your company’s data will be handled and if the provider uses AI tools in their own deliveries. The LGPD defines responsibilities of controller and operator in personal data processing (Law 13.709/2018); the contract should reflect this.
Criterion 7: ask what your team will be able to do at the end that they cannot do today.
Criterion 8: compare proposals in a standardized format; see How to Compare AI Consulting Proposals.

The complete list of documents to request is in What Evidence to Request from an AI Consulting Firm?.

What to Do Now

Before the first meeting: define the weights with decision-makers.
During meetings: ask the same questions to all candidates.
After: score individually, compare, and discuss differences, not impressions.
Before signing: review the contract with AI Consulting: 20 Questions Before Signing a Contract.

To evaluate dooop with this scorecard, contact is at dooop.com.br/contact.

Frequently Asked Questions

Are technical certifications a good criterion? They help confirm technical competence on specific platforms but do not replace business, team, and results criteria.

Should I request a paid diagnostic proposal before choosing? For larger projects, a short paid diagnostic is an efficient way to evaluate the provider working before a long-term commitment.

About dooop

dooop is a consulting firm based in Caxias do Sul (RS) that prepares leaders and organizations to transform artificial intelligence into business capability through strategy, governance, and training.