---
name: b2b-case-study-writer
version: 1.0.0
description: |
  AUTO-TRIGGER: Apply this skill when the user wants to write, structure,
  or improve a B2B case study or customer success story. Trigger phrases
  include: "case study," "customer success story," "write up our results
  with," "client story," "proof point document," "I want to document what
  we did for," "we got great results with a client," "turn this into a
  case study," or any request to document a program result or client
  outcome in a format that supports sales conversations or credibility
  building.

  Also trigger when the user has raw data or interview notes from a client
  or customer and needs to turn them into a structured asset, or when an
  existing case study needs to be made more effective.

  Do NOT trigger for general content writing, blog posts, or LinkedIn
  posts. This skill is specifically for the structured proof document that
  supports late-stage sales conversations and demonstrates measurable
  outcomes to prospects who are evaluating whether to buy.
allowed-tools:
  - Read
  - Write
  - Edit
---

# B2B Case Study Writer: Proof That Works in a Sales Conversation

This skill writes B2B case studies that function as sales assets, not
content marketing decoration. The difference matters. Most case studies
fail as sales tools because they are written to satisfy the content
calendar rather than to answer the question a prospect has at the moment
a deal is at risk.

A case study earns its place in a sales conversation when it answers
a specific objection, demonstrates a specific outcome that the prospect
cares about, and uses a customer voice that is credible to the reader.
A case study that is too long, too vague, or too focused on the vendor's
process rather than the customer's outcome does not move deals forward.

This skill also covers the decisions that most practitioners get wrong:
which result to lead with, how to get customer approval on specific
numbers, how to write a case study when the customer will not approve
quantified results, and how to repurpose one case study across formats
without reproducing it from scratch.

---

## HOW TO SET UP THIS SKILL

Provide:

- The customer or client and what they do (anonymized if required)
- The problem they had before engaging with the program or product
- What you or the product did specifically
- The measurable results with numbers where available
- Any direct quotes from the customer or key contact
- Where the case study will be used: website, sales outreach, proposal,
  presentation, or all of the above
- Whether customer approval is required and whether numbers can be
  attributed by name

If results are not yet confirmed or approved, the skill will help
structure the interview and approval process before writing.

---

## The Strategic Decisions Before Writing

---

### Decision 1: Which result to lead with

Most case studies lead with the result the company is most proud of.
The right case study leads with the result the prospect cares most
about at the moment they are reading it.

These are different things. A company might be proud of a 10x organic
traffic increase. A prospect evaluating a demand gen program cares
about pipeline and revenue attribution. Leading with traffic when the
prospect needs revenue proof loses the case study before the second
paragraph.

Before choosing the lead result, answer: what objection or question
does this case study need to address? The answer determines which
result goes first.

Common lead results by use case:

Revenue and pipeline: use when the prospect's primary concern is
ROI and whether marketing spend produces measurable pipeline.
Best for late-stage deals where procurement and finance are involved.

Speed and efficiency: use when the prospect's primary concern is
time to value or resource constraints. Best for prospects who have
tried the category before and failed due to implementation complexity.

Risk reduction: use when the prospect is risk-averse or has been
burned by a prior vendor. A case study that demonstrates how you
handled a difficult situation can be more effective than one that
shows only the upside.

---

### Decision 2: How to structure the proof

The challenge-solution-result structure is correct but insufficient
on its own. Prospects skip to the results section. If the results
section does not deliver immediately, the case study fails.

The structure that works in sales conversations:

**Headline:** The result, stated in specific terms. Not "How Company
X Grew Their Pipeline" but "How [Company] Grew Pipeline 4x in 18
Months Without Adding Headcount." Numbers in headlines earn more
reads than generic descriptions.

**The situation in one paragraph:** Who the customer is, what their
situation was, and what was at stake. This should take 60 to 80 words.
Not the customer's full company history. The specific context that
makes the problem relatable.

**The problem that could not be solved internally:** What they had
already tried and why it was not working. This is the paragraph most
case studies skip. It matters because it preempts the objection "why
couldn't you just do this yourself?" and it makes the customer's
decision to bring in outside help feel rational rather than weak.

**What changed:** What you or the product did, described in terms
of the customer's outcome, not your process. Not "we implemented a
multi-touch ABM program" but "we rebuilt the way [Company] identified
and prioritized target accounts, moving from spray-and-pray to a
focused list of 200 accounts that matched their best existing
customers."

**The results:** Specific, numbered, attributed to a named contact
where possible. Three to five results maximum. More than five dilutes
the impact. Lead with the most important one.

**The quote:** One quote from the primary customer contact that
addresses the emotional dimension of the result: not just what
happened but what it meant. "We finally had something we could show
the board" is more useful in a sales conversation than a quote that
repeats the numbers in different words.

---

### Decision 3: Getting customer approval on specific numbers

This is where most case studies die. The program produced great
results. The customer liked working with you. But when you ask to
publish a case study with specific numbers, legal or the CMO says
no.

Avoid this by setting expectations before the engagement ends, not
after. When a program is wrapping up and results are strong, say
explicitly: "I would like to write up what we achieved together as
a case study. Would you be open to that if we run the final version
by you?" Getting verbal agreement before results are published is
easier than getting legal signoff on a document they are seeing for
the first time.

When you do go through approval, send the case study in draft form
with a specific deadline: "I would love your feedback by [date].
If I do not hear back, I will assume the draft is approved as
written." This is a professional and commonly accepted approach.
It changes the default from "needs approval" to "approved unless
you object."

**When the customer will not approve specific numbers:**

Use ranges instead of exact figures: "pipeline increased by more
than 3x" instead of "pipeline grew from $1.2M to $4.7M."

Use relative metrics: "sales cycle shortened by more than half"
instead of "sales cycle dropped from 8 months to 3."

Use the customer's words in a quote without the numbers: let the
quote carry the emotion while the body copy carries the ranges.

Use an anonymous version: "A mid-market B2B SaaS company in the
supply chain space" identifies the customer type without naming
them. This is less powerful but still useful for prospects in
the same segment.

---

### Decision 4: Making one case study work across multiple formats

Most case studies are written as long-form documents and then
rarely used. The solution is to plan the formats before writing
so the core content serves all of them.

From one well-written case study, produce:

**The full version (600 to 900 words):** For the website, proposals,
and sales decks where the prospect has time to read.

**The one-paragraph version (100 words):** For sales emails,
LinkedIn outreach, and situations where the rep needs a quick
proof point they can paste into a message. Lead with the headline
result, one sentence on the situation, one sentence on the
outcome, and a link to the full version.

**The headline stat:** A single number or percentage for use in
presentations, one-pagers, and sales decks. "4x pipeline in 18
months" in large type on a slide.

**The quote card:** The customer quote formatted for LinkedIn,
email signatures, and social proof sections on landing pages.

Write the full version first. Extract the others from it. Do not
write each format separately.

---

## When You Do Not Have Enough to Write a Full Case Study

Sometimes the result is real but the documentation is thin: no
formal interview, no approved numbers, only informal feedback and
internal data.

In this situation, write a structured customer interview guide
before the case study. Ask the customer five questions:

1. What was the specific problem you were trying to solve before
   this program?
2. What had you tried before that was not working?
3. What changed after the program was in place?
4. If you had to put a number on the improvement, what would you say?
5. What would you tell a colleague at a similar company who is
   evaluating this type of program?

The answers to these questions provide everything needed for a
case study that the customer will also be willing to approve,
because the words came from them.

---

## Common Case Study Failures to Avoid

**The vendor-centric case study:** Describes in detail what the
vendor did without explaining what the customer experienced.
Prospects do not care about your methodology. They care about
what happened to the customer.

**The unverifiable claim:** "Significant improvement" and "dramatic
results" without numbers. These signal that the results were either
not measured or not good enough to state directly. Either is
damaging.

**The overlong case study:** More than 900 words for a web-format
case study is too long. If the result cannot be communicated in
900 words, the case study needs editing, not padding.

**The buried result:** The result appears in paragraph four after
three paragraphs of context and company background. Prospects
are reading to find out if this worked. Put the result first.

**The quote that restates the results:** "We increased pipeline by
4x" as a customer quote adds nothing. The quote should add
emotional context that the numbers cannot: "For the first time
in two years, I could walk into a board meeting with a pipeline
number I actually believed in."

---

## Deliver the Case Study

Output in this format:

```
B2B CASE STUDY
Customer: [name or anonymized description]
Primary use case: [website / sales outreach / proposal / all]
Built: [today's date]

---

HEADLINE
[The result, specific, with numbers where approved. 10 to 15 words.]

FULL VERSION (600-900 words)

[Situation paragraph]

[The problem that could not be solved internally]

[What changed]

[Results: 3 to 5 specific outcomes, lead result first]

[Customer quote: emotional dimension, not a repeat of the numbers]

---

ONE-PARAGRAPH VERSION (100 words)
[For use in sales emails and outreach]

---

HEADLINE STAT
[Single number or result for presentation use]

---

QUOTE CARD
[Customer quote formatted for standalone use]

---

APPROVAL NOTES
[What needs customer sign-off before publishing, and the recommended
approval language to use when sending the draft]

OBJECTION THIS CASE STUDY ADDRESSES
[The specific prospect objection or question this case study is best
positioned to answer. If it does not clearly answer a specific
objection, flag that as a structural problem to fix.]
```

---

## Output Rules

- Lead with the result that matters most to the prospect, not the
  result the company is most proud of. Ask which objection this
  case study is meant to address before choosing the lead.
- If the user does not have approved numbers, do not fabricate them.
  Produce ranges or relative metrics and note clearly which figures
  need customer approval before publishing.
- If the user does not have a customer quote, produce the five-question
  interview guide instead of a placeholder quote.
- The full version should be 600 to 900 words. Flag if the content
  provided is insufficient to reach 600 words and identify what
  additional information is needed.
- Every case study must close with the objection it addresses.
  A case study that does not clearly answer a specific prospect
  question is a content asset, not a sales asset.
- No em dashes. Use commas or periods.
