What Is Dogfooding? | Meaning & Examples

Dogfooding refers to a practice where a company tests and evaluates its own products or product updates in real-life settings to collect feedback from its employees before public release. It comes from the phrase eating your own dog food.

Dogfooding can help businesses ensure the quality, usability, or reliability of their products and is a common practice in the tech industry. Dogfooding has two variants that are often combined:

  • Many companies use dogfooding before a product reaches its customers. This allows businesses to collect user experiences and identify bugs without harming their reputation. They process the feedback from the research process before the official release to actual customers.
  • Most companies also promote the internal use of their own software products after their release in order to collect more feedback on real-life issues other users might also face.
Dogfooding example
Meta, Microsoft, and Google are examples of companies that use dogfooding during the development of new products or product updates. For example, Microsoft offers “Outlook Dogfood,” which helps them test the newest versions of Outlook before they release it to the public.

QuillBot also uses dogfooding to continuously improve tools such as the Grammar Checker, Paraphrasing Tool, or AI Detector.

It’s essential to recruit employees with characteristics that mimic those of your end users to participate in dogfooding.

Get our writing assistant wherever you go

QuillBot works across all of your favorite apps and websites, so you can write flawlessly everywhere.

Add to Chrome. It's free!

How does dogfooding work?

There are multiple ways companies can make use of dogfooding:

  • Pilot tests. You can create a pilot test program for a specific (small) group of employees to test new products or updates. It’s best to gather a sample of employees from different ages, genders, and departments to mimic your audience. This can become costly if you’re conducting mixed methods research or qualitative research to collect feedback (e.g., interviews, elaborate surveys with open-ended questions).
  • Beta testing. You can also invite employees to test your product before a public release (voluntary sign-up). By collecting bug reports and experiences, you can improve the product before it reaches its audience.
  • Forced adoption. You can roll out a mandatory release for all employees before going public to force them to use the product. This provides you with a large group of test users. The method is particularly useful if employees are already using the product and are testing a new update.

In most cases, it’s advisable to combine open feedback or bug forms with other methods of data collection, such as surveys with Likert-scale questions about product satisfaction.

Dogfooding example: Collecting demographic data
Surveys can help you collect different types of data. You decide on the demographic variables of interest (e.g., age, gender, technical knowledge) to get some background information on the employee that provides feedback. Then, you present them with a few statements on satisfaction with the product.

For example:

  • How would you describe your technical knowledge?
    • Very poor
    • Poor
    • Average
    • Good
    • Very good

This question provides you with ordinal data on someone’s technical background. It can help you put someone’s feedback into perspective so you can decide whether your end users are likely to encounter the same issue.

By using Likert-scale type questions for your survey, you can collect additional quantitative data, which is easier to analyze than qualitative data.

What are the benefits of dogfooding?

Dogfooding offers several benefits:

  • Increased credibility. Dogfooding can lead to increased credibility in two ways. By using your own products, you show that you have confidence in them. It also allows you to resolve bugs and process feedback before the public sees your product. This protects the company’s reputation.
  • Increased product understanding. Dogfooding can help employees understand the products better. This is especially relevant for marketing and sales employees, as well as for customer support employees, who can provide better assistance to customers.
  • Better product quality. Companies can identify and resolve bugs and issues before the product is released to a wider audience. This leads to a product of higher quality and reliability.
  • Decreased costs. Dogfooding can help a company detect issues early, before they escalate. It can also help avoid customer complaints and refund requests.
  • Increased innovation. Employees from other departments might have fresh ideas after using the product in a real-life scenario. These ideas might lead to new use cases or enhancements that hadn’t been considered before.

What are the risks of dogfooding?

While dogfooding has many benefits, there are also potential risks that come with this practice:

  • Limited scope. The face validity of feedback collected through dogfooding often seems high, but employees might actually use the product in different ways than the intended end user. This means some real-life scenarios might not be encountered, which leads to a low ecological validity and low generalizability to a wider audience.
  • Bias. Employees might have different characteristics than the intended end users. For example, they might have different technical expertise. This can result in feedback that is not representative of the broader user base. Some employees might also be hesitant to offer honest feedback. You can enhance the validity of the dogfooding research by adhering to research ethics (e.g., anonymization of feedback).
  • Resistance to change (with forced adoption). Some employees might not be willing to use unstable or unfinished versions of a product, especially if it slows down their workflow. This can lead to decreased motivation to deliver proper feedback in cases of forced adoption. If possible, it’s advisable to recruit a volunteer sample for internal beta testing.
  • Confidentiality risks. Pre-release versions of software can contain exciting new features that haven’t been announced yet. When releasing a new product or feature to a large employee base, you might risk the new features getting leaked.

To mitigate the risks, dogfooding should be accompanied by other testing methods, such as external beta testing and user research. These methods often have a higher internal validity because the experimental design is more thought-out.

Frequently asked questions about dogfooding

What is dogfooding in tech?

In the tech industry, dogfooding refers to the practice of companies using their own products or software internally before releasing them to a wider audience.

The goal of dogfooding is to collect feedback from employees in order to fix bugs and enhance the product pre-release.

Why is it called dogfooding?

The term dogfooding comes from the phrase eating your own dogfood, which refers to the practice of testing your own products before you release them to the public.

The term dogfooding is said to originate from a 1970s Alpo dog food commercial. The face of the commercial, actor Lorne Greene, claimed that he fed his own dogs Alpo. He tried to show confidence in the product by telling the audience it was good enough for him to use for his beloved pets.

The term was allegedly coined by Microsoft manager Paul Maritz when he sent an email with the heading “Eating your own dogfood” in which he motivated his team to use the company’s product.

Is this article helpful?
Julia Merkus, MA

Julia has a bachelor in Dutch language and culture and two masters in Linguistics and Language and speech pathology. After a few years as an editor, researcher, and teacher, she now writes articles about her specialist topics: grammar, linguistics, methodology, and statistics.