Skip to content
Academy
Marketing Academy · Field Work●Marketing Fundamentals
CoreAudit· 45 minutes

The Switch Interview Audit: Finding Blue Bottle's Real Job

Blue Bottle Coffee

Objective: Given five Switch Interview transcript excerpts from recent subscribers, separate genuine job statements from feature requests and identify the real competing alternative the subscription beat, applying the lesson's interview framework without conducting new interviews.

You're a marketing analyst at Blue Bottle Coffee reviewing five Switch Interview transcript excerpts from customers who joined the Bottle Bank subscription (recurring coffee-bag delivery) in the last two months, before a messaging refresh.

Read the transcript excerpts, separate genuine job statements from feature requests, and identify the real competing alternative the subscription beat.

Before you start

What you'll need

Free path (everything below is enough to finish)

FreeHold and annotate the 5 transcript excerpts

Free, easy to tag and compare text side by side

The process

2 steps

Step 01 of 02

Distinguishing feature requests from the real job

The lesson's Mistake 1 warns that a feature request ('send better emails') is not a job. The job is the life outcome underneath it ('look credible to clients without hiring a designer').

Transcript excerpt: 'Honestly I just wanted the bag delivery to be on a schedule I could set myself, not whatever the default was.' Is 'a schedule I can set myself' the job, or a feature request sitting on top of a job? What's the underlying job?

Google Docs— The 5 transcript excerpts pasted into one doc, one per section.

Procedure

  1. Read all 5 excerpts once without taking notes.
  2. For each excerpt, ask: does this sentence describe a UI control, or a life outcome?
  3. Tag every feature-shaped sentence ('a schedule I can set') as FEATURE.
  4. For each FEATURE tag, write the one-sentence life outcome underneath it.
  5. Compare all 5 underlying outcomes for a repeated pattern.
Sample output
Excerpt 3, tagged FEATURE: 'a schedule I could set myself'
Underlying job: 'Stop wasting good coffee because a fixed delivery cadence doesn't match how fast my household actually drinks it.'

Excerpt 5, tagged FEATURE: 'the app remembering my grind setting'
Underlying job: 'Never have to think about ordering again, once it's set up right, it should just work.'

Healthy

3 to 4 of the 5 excerpts resolve to the same or a closely related underlying job.

Unhealthy

Treating the feature request itself as the job and writing messaging around 'customizable schedules.'

What this means

When multiple customers independently ask for the same feature, the shared underlying job is the real insight, not the feature they used to describe it.

So what do I do about it?

SymptomActionEffort
Messaging draft lists product features instead of outcomesReplace every feature-shaped headline with the one-sentence life outcome from this audit30 min
YouYou can do this yourself, no engineering access required.

Step 02 of 02

Identifying the losing alternative as the real competitor

The lesson's Switch Interview method treats 'what did you almost go with instead' as the single most valuable question, because the losing alternative reveals who you're actually competing against in the customer's mind.

Excerpt 2: 'I was this close to just setting a recurring reminder to buy bags at Whole Foods, way cheaper.' None of the 5 excerpts mention a competing coffee subscription brand. What does that tell you about who Blue Bottle's Bottle Bank is actually competing with?

Google Docs— The same transcript doc, this time scanning only for the words 'instead,' 'almost,' or 'before.'

Procedure

  1. Search all 5 excerpts for 'instead,' 'almost,' or 'before.'
  2. List every alternative named, no matter how informal (a store run, a habit, doing nothing).
  3. Group alternatives into: competing subscriptions, one-off retail purchases, and 'do nothing / status quo.'
  4. Note which group has the most mentions.
Sample output
Alternatives named across 5 excerpts:
  Competing coffee subscription: 0 mentions
  One-off retail purchase (grocery store, local cafe bag): 4 mentions
  'Just kept forgetting to reorder' (status quo / do nothing): 1 mention

Healthy

The real competitor is retail habit or inertia, not another subscription brand, so messaging should target 'stop forgetting' and 'stop overpaying at retail,' not feature comparisons against rival subscriptions.

Unhealthy

Assuming the competitor is whichever subscription brand shows up in a competitive analysis deck, because no customer actually mentioned one.

What this means

The named alternative in real customer language always outranks a company's internal assumption about its competitive set.

So what do I do about it?

SymptomActionEffort
Competitive messaging compares features against a rival subscription brand nobody mentionedRewrite the comparison against the real named alternative: retail habit and forgetting to reorder30 min
YouYou can do this yourself, no engineering access required.

Final deliverable

A one-page audit memo: the shared underlying job across the 5 interviews, and the real competing alternative (named in the customer's own words, not assumed).

See a reference example
Sample output
Intercom, Switch Interview audit memo excerpt (illustrative)

Shared underlying job (4 of 5 interviews): 'Answer routine customer questions without hiring another support agent.'

Real competing alternative: not Zendesk or Salesforce, in every excerpt customers named the alternative as 'hiring a part-time support rep' or 'just letting the email backlog grow.' Intercom's messaging shifted from feature comparisons against other chat tools to a direct cost comparison against a support hire.

Success criteria

You're done when you can:

  • Every excerpt is correctly tagged as feature vs. underlying job
  • The competing alternative is drawn directly from the transcripts, not assumed
  • The final memo states the job as a life-outcome sentence, not a product feature