Marketing, Advertising and Public Relations Case Studies
Professional5 Case StudiesUpdated Sep 2026
These five cases come from retail, healthcare, consumer goods, and public-facing services. Each begins with a decision under pressure: how to spend a fixed budget, how to reposition a brand that customers read differently, why conversions slipped while visits held steady, how to respond when a post outruns the facts, and which pages deserve attention in a library that has grown unevenly. Every case works through audience intent, channel performance, brand consistency, and conversion evidence. The organizations and figures are fictional and internal to each scenario, so read them as examples of reasoning, not as benchmarks.
Case Study 1
Splitting a fixed budget for a cold brew launch
Larkspur Coffee Roasters, a regional roaster with four cafes, a wholesale grocery line, and a subscription program of roughly 6,200 members.
The situation
Larkspur Coffee Roasters had 48,000 to spend on launching a bottled cold brew line over one season. The leadership team wanted the launch to pay back in subscription sign-ups, not just grocery shelf presence. The product would sit in two grocery chains and in its four cafes, and the existing audience was mostly loyal buyers of whole beans who had never bought a ready-to-drink product from the brand. Paid search had a strong record for bean purchases, paid social had shown weak returns in the last cycle, and a regional radio buy was on offer at a discount. The marketing lead, Marisol, had to recommend a split before the production order was placed.
The challenge
The budget could not cover every channel at a level that would show results. Marisol needed to decide which audiences to reach first, which channels to fund for measurable conversion, and how to keep the new product recognizably Larkspur. The measurement plan also had to work with the grocery partners, who would not share shelf-level sales with the brand in real time.
Analysis
Teams trained in Marketing, Advertising and Public Relations tend to begin with intent rather than channel lists. Bean buyers searching for a specific roast are already in a purchase frame, while a person scrolling past a video about iced coffee is still deciding whether they want the category. Paid search could capture the first group cheaply, but its volume for a new product name would be small and it could not create demand. Paid social could reach the second group at scale, yet the earlier returns suggested the creative had been aimed at existing followers. The split should follow where each audience shows intent, not where the largest number of impressions is available.
Brand consistency mattered here because the cafes had trained regulars to expect a particular voice: plain language, roast notes written by staff, and no discount banners. A radio spot built around a price cut would have been easy to measure but would have sent a different message to loyal buyers. For conversion evidence, the plan used three things that could be compared: promo codes unique to each channel, a holdout region where paid social stayed off for the first two weeks, and cafe receipts that asked for the code. None of these is perfect, but together they show whether a channel caused purchases or merely appeared before them.
- Read intent Match each channel to the audience showing interest
- Fund the channel Shift budget toward the channel with the lowest cost per sale
- Check the brand Confirm every message still sounds like the cafes and beans
- Count the codes Compare holdout regions and unique codes before changing course
Actions taken
- Fund paid search first with about a third of the budget, using pages that name the product and link to a tasting guide, so the channel catches people already looking for it.
- Use paid social for a lookalike audience built from bean buyers, with creative shot by the cafe staff, and keep one holdout region dark for two weeks.
- Accept the radio offer only if it drops the price message, and point listeners to a cafe tasting instead, so the message matches the beans.
- Set the production run against a conversion floor, meaning a cost per recorded sale that the team agrees on before launch, and review channel spend against that floor every week.
Results
By the end of the third week, paid search had the strongest cost per recorded sale, and promo codes showed that most of its buyers were existing customers trying a new format. Paid social produced more reach but thin code redemptions, so the team moved most remaining spend to search and cafe sampling. The holdout region bought at a rate close enough to the test regions to suggest paid social had a smaller effect than the dashboard had claimed. The launch finished within budget and added 410 subscription members.
Questions for discussion
- How would the split change if the grocery partners agreed to share weekly shelf sales with the brand?
- Which of the three measurement methods would you trust most if only one could be run, and why?
- What evidence would convince the leadership team to fund paid social again in the next season?
Case Study 2
Closing a gap between clinical intent and perception
Meridian Hearing Clinics, a network of six clinics offering audiology exams and hearing aid fittings, with about 3,800 active patients.
The situation
Meridian Hearing Clinics had spent two years presenting itself as a clinical practice with audiologists on staff, longer appointments, and a follow-up schedule that most competitors did not offer. Its website, brochures, and clinic signage reflected that intent. Yet a recent round of interviews with current patients and people who had called but not booked described the clinics as affordable and convenient, with one interviewee calling them the cheap option in town. The gap mattered because the clinics were priced above big-box retailers and needed patients to accept that the premium was for care, not the device.
The challenge
The leadership team wanted to keep the price structure, since changing it would have meant rebuilding insurance contracts. The task was to reposition the clinics around clinical credibility without alienating the price-sensitive patients who already made up most of the appointments. The positioning also had to hold across the website, the phone script, and the front desk, where most first impressions were formed.
Analysis
Patients searching for a hearing test are in a different state from those searching for a hearing aid price. The first group is trying to find out whether something is wrong, and they respond to clinical language and proof of qualifications. The second group wants to compare costs and is drawn to listings that show a number. The clinics' search ads had been bidding on price terms, which brought clicks but few booked exams, while the referral channel from family doctors produced most of the new patients with a clinical reason to visit. Repositioning should follow the channel that brings intent, not ask paid search to fix a perception.
Brand consistency here meant the clinical story had to appear wherever a patient first met the clinics, not only on the homepage. The team wrote a short script for the phone line that opened with the audiologist's credentials and the length of the first appointment, and changed the signage so the word affordable no longer appeared above the desk. For conversion evidence, they tracked booked exams per 1,000 calls and the share of new patients who came back for a fitting, since a cheaper first visit that never led to a device would not show that the positioning worked.
Actions taken
- Rewrite the homepage and referral materials around clinical outcomes, naming the audiologists and listing what each first appointment includes, so the message starts where the referral patients already are.
- Move search budget away from broad price terms and toward clinical queries such as hearing test near me, with a landing page that answers insurance questions before quoting any device price.
- Train front desk staff on a short script and a list of words to avoid, including discount and cheap, and audit a sample of recorded calls each week.
- Review booked exams, fitting rates, and return visits at the end of the quarter, and treat fewer price-driven calls as progress rather than loss.
Results
After the first quarter, booked exams rose, and the share of calls that ended in a booking increased, although the total number of inquiries fell. The search campaign moved about 60 percent of its spend toward clinical terms, and the landing page's insurance section became the busiest area on the site. Return visits for fittings improved among patients who arrived through referrals, but the clinics saw a drop in cheap-device sales, which leadership expected. Patients who came in described the clinics as careful rather than cheap, although the sample was small and the team kept that finding provisional.
Questions for discussion
- Which patient group in the interviews would be hardest to reach with a clinical message, and what would you test first?
- If the price structure could change, how would the positioning argument shift?
- How should the clinics weigh a short-term fall in inquiries against a rise in booking quality?
Case Study 3
Diagnosing a conversion slide while traffic holds
Ostrander Home Furnishings, an online retailer of mid-priced sofas, dining sets, and storage, with one showroom in the home city.
The situation
Ostrander Home Furnishings had seen its site visits hold near the same level for three months, but orders had dropped by about a fifth over the same stretch. The drop was not spread evenly. Mobile visitors who reached a product page were buying at a lower rate, and desktop orders had also softened, though less sharply. The merchandising team had recently added a large set of new product descriptions and two promotional banners, and a shipping calculator had been moved from the cart to the product page. Nobody could say which of those changes, if any, had caused the decline, and the paid media team blamed attribution gaps for part of the picture.
The challenge
The company had a modest analytics budget and a small in-house team, so it could not run a full experiment on every suspected cause. The director of ecommerce, Tomasz, needed to decide which checks to finish first, which experiments to queue, and what to tell the finance lead about the likely cause of the drop before the next purchasing cycle started.
Analysis
Steady traffic can hide a shift in who is arriving. Paid campaigns had brought in more visitors from broad interest audiences during the same period, and those visitors were comparing prices rather than ready to buy a sofa. A flat visit count with a falling order rate means the mix changed underneath. The team split sessions by source and by intent signals such as product page depth, saved items, and search terms containing a brand name. Only then could they see that branded search was converting at its old rate while prospecting traffic had fallen off. That pattern pointed to a channel problem, not a site problem.
The content changes needed a separate check. New product descriptions had been written by a different team using a looser voice, with more adjectives and fewer measurements, and the promotional banners used a discount color that clashed with the rest of the store. Brand consistency was not a matter of taste here, since customers who bought from the showroom had described the brand as careful about sizing and materials. For conversion evidence, the team compared product pages before and after the description update within the same traffic segment, and checked whether shoppers who saw the shipping estimate earlier added more items to the cart.
Actions taken
- Repair the mobile product page first by moving the shipping estimate above the fold and checking the add-to-cart button on three common phone sizes, since this fix costs little and affects the largest drop.
- Pull the promotional banners from product pages for two weeks and split traffic so half the sessions see them, then compare order rates within each device group before deciding whether they stay.
- Rebuild attribution around three signals, branded search, direct visits, and saved items, and report prospecting traffic separately so the paid team stops judging its work by total sessions.
- Hold the description rewrite until the fitting-dimension fields are complete, and review the first thirty pages against the house style guide before publishing the rest of the catalog.
Results
Within the first three weeks, the mobile fix recovered part of the order rate, and the banner test showed that removing them lifted conversion on product pages by a modest margin. The attribution report made the prospecting decline visible to the paid team, which moved budget away from broad interest audiences. The description rewrite was paused, and the team kept the older copy on the catalog pages with the most sessions. Orders recovered most of their earlier level by the end of the quarter, and the finance lead accepted a forecast that assumed a slower return.
Questions for discussion
- What would you need to see in the traffic mix before accepting that a campaign change, rather than the site, caused a conversion drop?
- How would you sequence the four checks if the mobile fix had to wait for a developer?
- Which measure would you keep reporting to the paid team after attribution is rebuilt, and why?
Case Study 4
Responding to a viral claim about a snack bar
Bramblewood Snack Foods, a privately held maker of baked snack bars sold through grocery chains and a direct subscription service.
The situation
A video posted by a small health creator claimed that a Bramblewood bar contained a sweetener the company did not use. The clip had been viewed a few thousand times in its first day, then shared into parent-focused groups and picked up by two larger accounts. Within the same stretch, a retailer's social team asked whether the claim was true, and a customer began a thread saying she had stopped buying the product. The creator's claim was wrong about the ingredient but partly right about the labeling, which had used a vague term that could be read more than one way. The brand's comment team had responded with a single line that repeated the claim's terminology.
The challenge
The brand had about two days before the retailer's category review, and a national news outlet had emailed for comment. Ilse, the head of communications, had to decide whether to correct publicly, how fast to move, and who should speak. Her legal and quality colleagues wanted to hold statements until the ingredient records were checked, while the social team argued that silence was already being read as confirmation.
Analysis
Start with the stakeholder map, but treat it as a list of what each group can do next. Parents in the comment threads want a plain answer about what is in the bar. Retail buyers want to know whether the product will be pulled from shelves. The news outlet wants a named spokesperson and a timeline. The creator wants to be heard and may respond to a correction with a follow-up clip. Each audience reads the same statement through a different channel, so the team mapped each one to the place where its questions would arrive first: comments, buyer email, press inbox, and direct messages.
Verified facts had to come before tone. The team pulled the formulation sheet, the supplier documentation for the sweetener blend, and the label proof from the last print run, then confirmed in writing that the product contained none of the named ingredient. The labeling term was the harder part. The phrase had been approved before the current packaging, and the statement had to explain it without sounding defensive or contradicting the box. Conversion evidence in a crisis is mostly about trust: the team tracked subscription cancellations, retailer reorder notices, and comment sentiment against a baseline from the previous weeks.
- 1Map stakeholdersList what each group needs and where its questions will arrive
- 2Verify factsPull formulation, supplier documents, and label proof before any reply
- 3Coordinate messagesOne fact sheet, one spokesperson, and a holding script for social
- 4Contact the sourceCorrect privately first, and go public only if the private reply fails
- 5Track recoveryWatch cancellations, reorders, and comment sentiment against a baseline
Actions taken
- Publish one fact sheet with the formulation and supplier confirmation, and link it from every comment reply so the answer is the same everywhere.
- Name one spokesperson, the head of quality, to answer the press and buyers, and keep the social team on a holding script that acknowledges concern without repeating the disputed claim.
- Message the creator privately first with the documents, and decide on a public reply only if the private one fails, since a public correction can extend the thread.
- Set recovery indicators for the next two weeks, including cancellations per 1,000 subscribers, retailer reorders, and the share of negative comments, and report them to leadership at a fixed time each day.
Results
The fact sheet was live before the retailer review, and the buyer accepted the explanation without pulling the product. The creator posted a correction within two days, although the original clip stayed up and continued to draw comments. Subscription cancellations rose for about a week and then returned close to the baseline, while retailer reorder notices held steady. Negative comments fell as a share of total replies, but the team recorded that the phrase on the label would be revised at the next print run, which was a decision the company had planned anyway.
Questions for discussion
- Who in the stakeholder map had the most influence over the outcome, and how would you test that assumption early?
- When is a private correction to a creator the wrong first step?
- Which recovery indicator would you trust least, and what would you add to offset it?
Case Study 5
Auditing a content library with uneven results
Northfield Credit Union, a member-owned lender with four branches and about 62,000 members.
The situation
Northfield Credit Union had built up a library of 1,240 articles, guides, and calculators over several years, but no one owned it as a whole. Each branch manager wrote pages for their local members, the lending team added rate explainers, and the compliance group approved most pages in a queue that often took weeks. A sample of recent traffic showed that a small group of pages earned most of the search visits, while many older guides had no visits in the last period. Quality also varied. Some pages were clear and current, others repeated rates that no longer applied, and few said who had last reviewed them.
The challenge
The content lead, Dara, had been asked to cut the review queue in half without lowering compliance standards. She needed to know which topics and formats were worth refreshing, which pages could be merged or retired, and how to assign ownership so the next audit would not start from zero. Search intent varied widely, and a page that ranked for a rate question might be the wrong answer for someone who needed help with a payment problem.
Analysis
Search intent should decide the shape of each page before anyone edits a sentence. A member searching for a current savings rate wants a number and a short note about how it changes. A member searching for how to dispute a fee wants steps and a phone number. A guide that tries to do both usually does neither well. The audit grouped pages by the intent they served, then compared their search visits, time on page, and calls to branches that cited the page. Channel performance mattered, because a few high-traffic pages fed the mobile app and the branch kiosks, so a change there reached members who never searched.
Brand consistency in a credit union is partly a matter of tone and partly a matter of accuracy. Members expect the same voice from a branch page as from a rate explainer, and an outdated rate on a page they found through search damages trust faster than a clumsy sentence does. The team set a rule that every page showing a rate or fee had to name its reviewer and a review interval, and it checked those fields against the compliance log. For conversion evidence, the audit linked pages to applications and call outcomes where the member had clicked through, which was imperfect but far better than page views alone.
- First weekPull all pages with traffic, owners, and last review notes
- Second weekGroup pages by search intent and mark duplicates for merging
- Third weekStart the fast lane for rate and fee corrections
- By end of quarterReport overdue reviews and retire pages that no longer serve intent
Actions taken
- Merge the rate pages into one maintained table with a single owner, and retire the duplicate explainers that repeat old figures, redirecting their search visits to the current version.
- Group the remaining pages by search intent into three types, how-to, rate or fee lookup, and eligibility, and write a template for each type so reviewers check the same things every time.
- Set a fast lane for rate and fee corrections that compliance approves within two business days, and keep the full queue for new guides, so urgent fixes stop waiting behind new content.
- Assign each page an owner and a review interval in the content system, and report the share of pages past their interval to leadership at the end of each quarter.
Results
The first refresh pass covered 210 pages, which was the set of rate and fee guides that drew most of the search visits. Corrections went through the fast lane in a median of one business day, and the review queue for new guides shortened by about a third. Search visits to the merged rate table rose after the redirects settled, and calls that cited rate pages fell slightly, which the team read as members finding answers faster. Eleven older guides on minor topics were retired, and the owners agreed to review the retirement list each quarter.
Questions for discussion
- How would you decide which rate pages to merge when two versions have different traffic and different accuracy?
- What should count as an urgent correction, and who should be allowed to trigger the fast lane?
- Which measure would show most clearly that a refreshed guide is serving its search intent better?
Learning Resources
Professional CASE STUDIES
China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.