Outputs vs. Outcomes
By Millie O. Symns in Data Strategy
March 21, 2022
Infamous logic model confusion
Since there is no such thing as too much logic model chat, I thought, why not continue the series!
One of the questions that I get every time I work on a logic model with a client is, “What is the difference between outputs and outcomes?” So today, we will dig deeper into this battle of semantics. I’m sure it is not helpful in this context of a logic model to have two words never to each other that
- look similar,
- start with the same first three letters,
- and mean similar things in different contexts.
Let’s remind ourselves of the definitions of these terms.
Term | Simple definition |
---|---|
Output | The numeric description of the results of your activities. |
Outcome | The abstract or verbal description from the results of your activities (and outputs). |
Last time we used a cake analogy to describe a logic model, so this time to help us understand the difference between these two terms, let’s take some inspiration from the new spring season (for the northern hemisphere) and use a garden analogy.
Spring gardening is here! Photo by Jonathan Hanna on Unsplash
Building a garden
In this scenario, your goal is to have a beautiful multi-purpose garden that has a section of fragrant flowers and another section with herbs or veggies that you can harvest. Perhaps some of your goals include attracting as many pollinators as possible and having enough herbs and veggies to incorporate in your family meals over spring and the summer.
Your activities for this would include researching plants that can grow in your area that meet your needs, planting your desired seeds or starter plants, and having your friends and family help with the daily watering and tending of the garden.
Outputs = Numbers
Now, what could be the outputs? When you hear outputs, think numbers. If everything went as planned, what would be the quantifiable results of the hard work from your activities?
In your sizeable perfect world of a sizeable garden, you can list out your outputs (even break them out into categories) to be something like:
Flowers:
- 5 small bushes of lavender
- 5 small bushes of geraniums
- 15 stalks of sunflowers
- 3 bundles of lilies
- 8 bundles of bergamot flowers
Herbs and Veggies:
- 10 cups of basil
- 5 cups of thyme
- 20 carrots
- 5 cups of sage
- 5 heads of lettuce
Other:
- A 10% increase in pollinators (bees and butterflies) that come to visit your garden regularly
- A 50% increase of vegetable consumption in your household (carrots can be pretty tasty when freshly picked)
- A 30% decrease in social media consumption (everyone is into the garden and not doom scrolling all day)
As you can see in the pattern here, each output has a numeric value.
Outcomes = Words
What about outcomes? I like to think of these as descriptors that you would most likely hear in a casual conversation. These might be some of the things you tell your friends over some fresh lemonade as you casually talk about your life-changing garden results. You could put numeric values to these, but it would be harder to capture.
Some examples of outcomes could be (you could also break out into short and long term):
Short-term
- An increase in nutrients and minerals in your diet
- A better overall mood from family and friends who tended the garden
Long-term
- Your neighborhood has an increase in pollinators
- The local plant nursery survives another year with an increase in business
- Lead a healthier and less sedentary life over time
You could put numbers to some of these items in some cases. It is a little harder to measure mood and track nutrients and minerals or pollinators in the neighborhood. Your outcomes are often based on past literature or examples with similar activities to yours in a real-life program. You most likely will develop surveys and define metrics for these outcomes based on your activities and outputs.
Remember that your outcomes should be reasonable and realistic to your activities. You can shoot for the stars of wanting to save all pollinators from dying out, but it probably won’t be an outcome with just this intervention.
I hope you and your colleagues find this helpful as you work through creating your new logic model!