Marketing · 7 min read

What makes a case study sell? Structure, honest numbers and the client's own words

Accendoz·
Case studies sell on proof, not praise. Flaws and all.

A case study sells when a buyer can see at a glance that the client was like them and what measurably changed, then finds enough process detail and honest limits to believe it. Lead with a snapshot and numbers, show your decisions, and quote the client with written approval.

A case study sells when a buyer can tell at a glance that the client was a business like theirs and can see what measurably changed, then finds enough process detail and honest limits to believe it would work for them. The order that achieves this is a snapshot with the numbers, the problem and its constraint, the decisions you made, the outcome against a stated baseline, and the client in their own words.

Most service businesses write the opposite: a celebration of the agency, the result buried in paragraph six, adjectives where figures should be. Research on how people screen web pages and weigh evidence, India's rules on endorsements and Google's guidance all point the other way.

of B2B marketers named case studies and customer stories among the content types that produced their best results
53%
of B2B marketers named case studies and customer stories among the content types that produced their best results
conversion lift when reviews appeared on a higher-priced product, against 190% for a cheaper one
380%
conversion lift when reviews appeared on a higher-priced product, against 190% for a cheaper one
average star rating where purchase likelihood typically peaked; near-perfect scores sold less
4.0 to 4.7
average star rating where purchase likelihood typically peaked; near-perfect scores sold less
maximum CCPA penalty for a misleading advertisement, rising to ₹50 lakh for repeats
₹10 lakh
maximum CCPA penalty for a misleading advertisement, rising to ₹50 lakh for repeats

01

What a buyer checks first

Nobody reads a case study in the order you wrote it. Microsoft researchers who modelled how long visitors stay on web pages described a pattern they called 'screen-and-glean': a page has to pass an initial screening before anyone examines it in detail. For a case study, the screening comes down to three questions. Is this a business like mine? Did something measurable improve? Could it work for us?

Evidence also carries more weight as the stakes rise. Tracking about 15.5 million page views at a high-end gift retailer, Northwestern University's Spiegel Research Center found that displaying reviews lifted conversion by 190% for a lower-priced product and 380% for a higher-priced one, because more information offsets a riskier decision. A service engagement worth several lakh sits at the expensive end of that comparison.

In the Content Marketing Institute's survey of 980 B2B marketers, fielded in mid-2024, 75% had used case studies or customer stories in the previous year, and 53% named them among the content types that produced the best results, second only to video.

02

A six-part structure, in the order buyers ask

Each part answers one question in the buyer's head. Give the snapshot the most care: a reader who stops there should still know who the client was, what changed and by how much.

Case study structure: what each part must answer
PartBuyer's questionWhat to write
SnapshotIs this someone like me?Client type, city and size, the problem, two or three result figures
SituationWas their problem my problem?The starting numbers, and the constraint that made it hard
ApproachDid they know what they were doing?Three to five decisions, including one you changed midway
OutcomeDid it really work?Each figure with its baseline, time window and how it was measured
Client's wordsWould they hire them again?A short, approved quote about the experience
FitCould this work for us?What the client provided, and where it does not apply

03

Numbers that survive a sceptical reader

Every figure needs a baseline, a time window and a method. 'Enquiries up 40%' invites doubt. A hypothetical better version: 'website enquiries rose from 22 to 31 a month, March to May against the same months last year, counted in our CRM'. It answers the doubt before it forms.

Resist the perfect story. In the Spiegel data, purchase likelihood typically peaked at average ratings of 4.0 to 4.7 and fell as ratings approached 5.0, which the researchers read as perfection seeming too good to be true. A case study that admits what took longer than planned, or what did not move, gives the reader a reason to trust everything else.

Google's guidance for writing reviews sets a similar standard: provide evidence of your own experience, explain what sets something apart from its competitors, and discuss benefits and drawbacks based on your own original research. Swap 'product' for 'project' and it reads like a brief for a case study.

Share quantitative measurements about how something measures up in various categories of performance.

Google Search Central, guidance on reviews

04

The client's words, with permission

The quote is where many case studies go wrong: lines drafted by the agency, polished until no human would say them, or approved by someone who never used the service. India sets a useful bar. Guidelines on misleading advertisements and endorsements, notified on 9 June 2022, cover advertising by manufacturers, service providers and traders, and say an endorsement should reflect the genuine, reasonably current opinion of whoever makes it, based on adequate information or experience.

Where an endorser has a connection with the advertiser that could materially affect the endorsement's credibility, and that the audience would not reasonably expect, it must be disclosed. Whether a website testimonial counts as an endorsement in law is for your lawyer, but the test is worth adopting: a discount given in return for a quote is exactly what a reader would want to know. In practice:

  • Interview the client and use their words, lightly trimmed. Do not write the quote for them.
  • Quote the person who lived the change, such as the clinic manager or sales head, not only the owner who signed the cheque.
  • Get written approval of the final quote, their name and title, and every figure you publish.
  • If they cannot be named, say so plainly. 'A 40-bed hospital in Thrissur' is honest; a stock photo with an invented name is not.
  • Date every case study and refresh quotes more than a couple of years old.

05

Make it findable, and quotable by AI

Google's guidance on helpful content asks whether a page provides original information, reporting, research or analysis, and whether it clearly demonstrates first-hand expertise, such as from having actually used a product or service. Google says E-E-A-T is not itself a specific ranking factor, but a well-documented project is exactly the kind of page those questions describe.

Give each case study its own page with a descriptive title, such as the hypothetical 'How a Thrissur clinic halved its evening enquiry backlog'. Put the figures in the text, not only in images, and open with a two-sentence summary an assistant could lift whole. A gated PDF is invisible to anyone who has not handed over an email address, and to assistants.

Do not expect star ratings from testimonials. Since September 2019 Google has not shown review rich results for businesses and organisations when the reviews sit on their own website, which it calls self-serving. Link each case study from the matching service page and, where buyers ask about cost, from an honest price page.

What to do with your next case study

Start with a handful. In the Spiegel data, the first five reviews drove the bulk of the lift in purchase likelihood. Reviews are not case studies, but the pattern is a useful guide: five strong, specific stories covering your main kinds of buyer will do more than a wall of logos.

  1. Pick the right client, one whose situation matches the buyers you want more of, not simply your biggest logo.
  2. Agree the metrics first: the baseline, the time window and where the numbers come from.
  3. Interview the client for 30 minutes: what was hard, what they expected, what changed and what surprised them.
  4. Write in the six-part order, snapshot first, and include one limit: something that took longer, cost more or did not move.
  5. Get written sign-off on every figure, the quote and the name.
  6. Publish it on its own indexable page, dated, and link it from the matching service page.

If you would like a second opinion on a draft, ask us for a free audit.

Sources

All insightsMarketing
Chat with us