What You Will End Up With
A project description in which each activity can be traced to a result. It has no more than three broad goals, objectives that say how much of what will be done by when, a list of activities with the product each one produces, and a timeline chart showing every activity with its projected date of completion. A flow chart of the sequence of events sits alongside the timeline, so a reviewer can follow the logic at a glance.
Before You Start
Settle the logic before you write the text. A logic model is a visual picture of the assumptions and theory of action behind your plan, and a few minutes spent on it will stop you from listing activities that go nowhere.
- Write down the current situation and the outcome you want.
- List what has to change to get there: behavior, knowledge and skills.
- List the activities and resources you will need, such as staff, money and equipment.
Then decide what your evaluation will cover: implementation, short-term outcomes, long-term outcomes, or all three. That choice shapes your objectives, because you can only promise what you will measure, and the budget needs line items for the evaluation work.
Steps
-
Set no more than three goals. A goal is a broad statement of a long-term, ideal outcome. Most proposals should carry no more than three. Keep them high-level; objectives come next and carry the detail.
-
Write objectives that can be checked. Goals are the result you expect from the whole project, and objectives are the specific steps that lead to it. Use the SMART test: specific, measurable, achievable, relevant and time-bound. Specific means the objective says what will be done and who will do it. Measurable means it says how the action will be measured. Time-bound means every objective has a timeline for completion. Put together, an objective says how much of what will be accomplished by when: a deadline, a percentage or quantity, and the result.
Keep the target realistic. A 2% decrease in smoking among adolescents over two years is a feasible objective. A 50% decrease is not. A reviewer will believe one and doubt the other.
-
List activities, starting each with an action verb. Developing training materials and conducting training sessions are typical activities. A leading verb shows that something is being done, and it keeps activities from turning into vague topics.
-
Name the direct product of each activity. Every activity needs a product that shows it happened. These are your outputs: typically counted in numbers, such as the number of people trained. If you cannot say what an activity will produce, reconsider the activity.
-
Separate outputs from outcomes. An output is a product of an activity. An outcome is a change that results in participants' knowledge, beliefs or behavior. Outcomes come in stages. One common framework puts short-term outcomes at within 1 to 3 years and long-term outcomes at 4 to 6 years. Ultimate outcomes, the impact, take 7 to 10 years and are changes at the level of an organization, community or system. If the call sets different horizons, use its numbers. Between short and long term sit intermediate outcomes: specific, measurable changes in behavior, decisions or policies.
A short illustration, using the smoking example. The goal is healthier adolescents. The objective is a 2% decrease in smoking among adolescents over two years. One activity is "Train peer educators," and its direct product is the number of peer educators trained. A short-term outcome is that adolescents know the risks; a longer-term one is that fewer of them smoke. Each link can be checked, and nothing floats free.
-
Build the timeline and the flow chart. The timeline chart, such as a Gantt chart, lists every activity with its projected date of completion, usually in sequence. The methods section should also hold a flow chart showing the sequence of events. The timeline shows when things happen, and the flow chart shows how one step leads to the next. Avoid timelines that are unrealistic for the work you describe.
-
Test the chain. Read the model from the start. Every element must link to something earlier, and every input, activity and output must lead to at least one short-term, intermediate or long-term outcome. An activity that leads nowhere does not belong in the proposal.
Using AI Safely
An assistant can draft wording for goals and objectives, but it does not know your local conditions. Edit what it gives you.
- Check that every objective is feasible. Assistants tend to propose ambitious targets; a modest, achievable change is a stronger promise than a dramatic one.
- Test the logic. Make sure every element of a generated logic model links to an earlier step and leads to an outcome. Lists of activities can look professional and still lead nowhere.
- Check that each activity opens with an action verb and has a direct product you can verify.
- Read any timeline it proposes date by date. It will not know about local holidays, hiring delays or slow purchasing.
Common Mistakes
Unrealistic timelines. Squeezing too much into too little time suggests you have not planned the work. Check each date against your context and resources.
Activities with no direct product. If an activity has nothing to show for it, a funder cannot see that it happened. Give each one a product, such as booklets produced or people trained.
Vague objectives. An objective without a quantity, a deadline or a specific result cannot be measured. Fill in all three.
Goals that do the work of objectives. A goal like "healthier adolescents" cannot be measured, and an objective like "reduce smoking by 2% over two years" is too narrow to be a goal. Keep each in its place so the reviewer can see how the small steps add up to the broad aim.
Broken chains. Activities that lead to no outcome waste the reviewer's time and weaken your case. Link each input, activity and output to at least one outcome, or take it out.