To present a recommendation at work, state the action you recommend before the background. Then give the decisive rationale, name the main tradeoff or uncertainty, and finish with the exact decision or next step you need. A useful spoken structure is: recommendation, rationale, tradeoff, next step.
This four-part sequence is a practical editorial framework, not a scientifically proven formula. Its underlying principles are better supported: put the bottom line early, compare options against relevant criteria, expose material risk, and make the decision request explicit. The goal is not to sound certain. It is to make your recommendation easy to evaluate.
Method
Make the decision easy to evaluate
Step 1
Recommendation
State the action you recommend.
Step 2
Rationale
Give the evidence that most changes the choice.
Step 3
Tradeoff
Name the main risk or uncertainty.
Step 4
Next step
Ask for the exact decision or action needed.
Why recommendations get buried
A recommendation often comes out in the order the work happened. You explain the problem, describe the research, list every option, and finally reveal what you think should happen.
That order makes the listener reconstruct the decision. Separate the recommendation from its background, then provide only the context needed to evaluate it. See related structure and organization techniques. If the broader problem is information order, start with how to get to the point when speaking.
What the evidence supports
US government plain-language guidance recommends stating the purpose and bottom line first, putting important information at the beginning, and moving background later when needed. It concerns government writing, not speech, but supports a useful convention: reveal the point early.
The UK government's Green Book is for formal public-sector proposals. It says advice should compare options, surface costs, benefits, risks, and tradeoffs, identify a preferred option, and communicate uncertainty. Routine workplace decisions need less analysis, but the logic applies.
Evidence on information order is indirect. DeAngelo and Yegiyan found that people spent less time reading inverted-pyramid news stories and recalled them better than narrative versions. That reading study does not prove a universal spoken order.
Uncertainty also need not be hidden. Fischhoff and Davis argue that uncertainty communication should fit the decision people face. In five experiments with 5,780 participants, van der Bles and colleagues found only small reductions in trust overall when uncertainty was stated, mostly for verbal rather than numeric uncertainty. These studies concern public facts, not internal meetings. Acknowledging a material limitation does not automatically destroy trust.
No source reviewed here validates the exact four-part sequence as a universal speech formula. It is a practical synthesis of better-supported principles.
The recommendation presentation structure
1. State the recommendation
Begin with the action you want taken.
I recommend moving the launch to 14 October.
A recommendation should contain a verb. “My recommendation is Vendor B” names an option. “I recommend approving Vendor B for the support-platform migration” makes the action clearer.
Add scope, timing, or ownership only when it changes the decision. Skip the long preamble.
2. Give the decisive rationale
Give the evidence that could change the decision, not every fact you collected.
One or two decisive reasons are a useful editorial default, not a scientific limit. Use criteria that separate the option from alternatives, such as security, cost, delivery time, impact, or reversibility.
Ask: if this reason disappeared, would I still make the same recommendation? If yes, it may be supporting detail.
3. Name the material tradeoff
State the cost, risk, lost benefit, or unresolved uncertainty a reasonable decision maker should know.
The main tradeoff is that this delays the reporting upgrade by two weeks.
Name the downside that matters and, when possible, its scale or the condition that could reverse your recommendation.
4. End with the next step
Say whether you are asking for input or approval.
Input: I am asking whether the integration risk changes this choice.
Approval: I am asking for approval today to begin contract review, with annual spend capped at €48,000.
A vague “What do you think?” can create an unfocused discussion. Name the decision, timing, and immediate action when they matter.
Use this template:
I recommend [action]. The decisive reasons are [reason one] and [reason two]. The main tradeoff is [material downside or uncertainty]. I am asking for [input or approval] by [time] so that [next action] can happen.
Before and after: recommending Vendor B
Before
We started with eight vendors and narrowed it to three. Vendor A is the cheapest, although security had questions. Vendor C has the best dashboards. Vendor B was stronger on implementation, and the security team preferred it. There are price differences, and we need to decide soon. I think Vendor B is probably the best option.
The facts may be sound, but the decision request is buried.
After
I recommend approving Vendor B for the customer-support migration. It is the only finalist that meets our security requirements without custom work, and its implementation plan fits the 30 September deadline. The main tradeoff is price: it costs €8,000 more per year than Vendor A. I am asking for approval today to begin contract review, with the final annual price capped at €48,000.
The revision makes the recommendation, two reasons, tradeoff, and approval request visible.
A project-priority example
I recommend pausing the dashboard redesign for one sprint and moving two engineers to the onboarding defects. The defects are blocking account setup and generating most of this month's support escalations. The tradeoff is that the redesign moves from 18 September to 2 October. I am asking for approval to change the sprint plan today. If approved, I will update the owners and support team by 16:00.
The listener can see the action, rationale, cost, and next step.
Should you present every alternative?
Name the serious alternatives a reasonable decision maker would expect you to have considered, then explain briefly why they are not preferred. Keep the full comparison available for questions or an appendix.
There is no evidence-based rule that every recommendation should show exactly three options. A meta-analysis of choice overload covered 63 conditions and 5,036 participants. Its average effect was close to zero, with substantial variation. “Fewer options is always better” is too simple.
For procurement, legal, medical, safety-critical, or regulated decisions, follow the required review and documentation process rather than compressing the evidence to fit a speaking template.
Try it aloud, then retry
Choose a real, low-stakes recommendation about a meeting, tool, deadline, or project priority.
For your first attempt, speak for no more than 45 seconds. Include one recommendation, one rationale, one material tradeoff, and one specific next step. Record it without scripting every sentence. Then mark the moment when a listener could first tell what action you wanted.
Retry the same recommendation, but put the proposed action in the first sentence. Keep the rationale, tradeoff, and ask. Remove one detail that does not affect the decision.
Compare the attempts on one question: How many seconds passed before the requested action became clear? One exercise will not solve every communication problem, but it gives you one controlled change to observe.
When this structure does not fit
Do not force a recommendation when the purpose is discovery. Ask for input first when you genuinely lack the evidence or authority to choose.
When the audience disputes the problem itself, give the minimum shared context needed, then state the proposed action. High-stakes decisions may require full analysis, written records, specialist review, or prescribed formats. This structure can introduce the recommendation, but it cannot replace due diligence.
A conditional recommendation is also valid:
I recommend Vendor B, provided the security review confirms data residency by Friday. If it does not, I recommend extending the current contract for one quarter.
That is clearer than pretending the uncertainty does not exist.
Frequently asked questions
How much evidence should I include?
Include enough to show why the recommendation wins on the criteria that matter. One or two decisive reasons often work in a short meeting answer, with supporting detail ready for questions. Increase the depth when the decision is expensive, irreversible, contested, or governed by a formal process.
Should I present all alternatives?
Present the serious alternatives, not every discarded idea. Name the strongest competitor and the criterion on which it loses. Keep the full analysis in an appendix. Do not hide an option that would materially change the decision.
How can I sound confident while acknowledging uncertainty?
State what is known, what is uncertain, and whether the uncertainty could reverse the recommendation. Replace vague hedging with a condition or range when evidence supports one. Confidence comes from clear reasoning and honest boundaries, not invented certainty.
What should the final ask be?
Ask for the smallest specific decision that moves the work forward. State whether you need input, approval, budget, an owner, or permission to investigate. The listener should know what to do and what happens after they agree.
Should the recommendation come before the background?
Usually, yes, when the audience knows the context and the purpose is to decide. Lead with the action, then provide the rationale and background needed to evaluate it. Put brief context first only when the recommendation would otherwise be unintelligible or agreement on the problem is the first decision.
Practise the skill
Open impromptu speaking practice, select the Workplace Scenarios category, and record a short recommendation. The structure review maps the order of your answer, identifies the closest framework when one fits, and provides a refined version to compare with what you said. Retry the same question with the recommendation in the first sentence. For a framework reference before you practise, see speaking practice with clear frameworks.
The practice can help you examine your transcript, pacing, filler words, and answer structure. It cannot determine whether the underlying business recommendation is correct.
Sources
- Digital.gov: Organize the information
- HM Treasury: The Green Book 2026
- Fischhoff and Davis: Communicating scientific uncertainty
- van der Bles and colleagues: The effects of communicating uncertainty on public trust in facts and numbers
- DeAngelo and Yegiyan: Looking for Efficiency
- Scheibehenne, Greifeneder, and Todd: Can There Ever Be Too Many Options?
