Eat Your Own Dog Food
Use your product routinely to create a leader-owned flaw-fixing loop
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 98%
Leaders should use their own product frequently and complete the same journey expected of customers. Capture friction firsthand, route concrete notes to the relevant teams, fix the defects, and repeat the experience. At Tesla, leaders took cars from the end of the factory line, drove them home and back, and gave engineers notes. McNeill also buys Lululemon products and visits its stores to experience the purchase and product loop. The mechanism is direct exposure: defects that remain abstract in dashboards become intolerable when leaders experience them personally. Dogfooding does not replace customer research—leaders are not always representative—but it adds a fast, accountable quality loop that is otherwise forfeited.
Origin
McNeill presents dogfooding as one of the two secret ingredients around the algorithm and describes Tesla leaders driving fresh production cars every night to identify flaws.
Core principles
- 01Leaders who do not use the product abandon a valuable feedback loop
- 02Firsthand friction is harder to rationalize than a report
- 03Leaders should experience the same journey as customers
- 04Observed flaws need a direct path to the people who can fix them
How to run it
- 1
Set a usage cadence
Define how often each leader will use the product or complete the customer journey.
Pro tip Choose a cadence frequent enough to catch regressions.
- 2
Use the real path
Buy, onboard, operate, and seek support through the same interfaces and policies customers face.
Pro tip Avoid internal shortcuts that hide customer friction.
- 3
Capture specific flaws
Record what happened, where it happened, and why it harmed the experience.
Pro tip Give teams reproducible observations instead of broad dissatisfaction.
- 4
Route and prioritize
Send the evidence to the responsible team and prioritize flaws that customers also encounter.
Watch out One executive preference is not automatically representative customer evidence.
- 5
Retest the experience
Repeat the same journey after a fix to confirm the feedback loop actually closed.
In the wild
Tesla leaders took a newly produced car home each night, drove it back to the factory, and gave engineers notes about the vehicle they had just experienced.
→ Leaders became an immediate source of firsthand defect feedback.
McNeill asked banking CEOs whether they used their own apps weekly or monthly. Almost no hands went up, reinforcing his view that poor experiences persist when leaders stay outside the usage loop.
→ The exercise exposed a missing executive feedback mechanism.
Common mistakes
Using an executive-only shortcut
A privileged path prevents the leader from experiencing the friction imposed on ordinary customers.
Replacing customer research
Dogfooding adds evidence but cannot make a leader representative of every customer segment.
Is it for you?
Best for
Founders, CEOs, board members, and product leaders who can personally use the customer experience.
Not ideal for
Products that leaders cannot safely or legitimately use without specialist qualifications or belonging to the target population.
From the transcript
“The first is what we call it eat your own dog food, which is use your own product.”
“If you don't, then you're forgoing a whole feedback loop.”
“we used to take each of us used to take a car off of the end of the factory line every night to go home.”
From the episode
Episode 541: Jon McNeill: Why "Less" and "Simple" are the Smartest Growth Strategies
Jon McNeill