Data-Driven PR Campaigns: 5 Cases From Dove to Spotify

Data-driven PR works when one verifiable number becomes the story. Dove turned a survey of 3,200 women into a global campaign, Old Spice turned live social comments into 186 video replies, and Spotify turns listening history into personal recaps that 200 million people opened in about a day. This guide covers five documented cases, the figures each one reported and a method you can copy.

Updated November 2, 2026. Figures come from the sources named in each section, and several are reported by the brands or their agencies, so treat them as claims, not audited results.

On this page

• What makes a PR campaign data-driven?

• Which five campaigns are worth studying?

• How did Dove turn a survey into a campaign?

• How did Coca-Cola use name data?

• How did Old Spice use real-time comments?

• How did Red Bull make a record the story?

• How does Spotify Wrapped earn its coverage?

• What do the five campaigns have in common?

• How do you measure a data-driven PR campaign?

• What mistakes weaken data-driven PR?

• Frequently asked questions

What makes a PR campaign data-driven?

A PR campaign is data-driven when data decides the idea, not just the report that follows it. The data can be a survey the brand commissions, a list of popular names, a stream of live comments or a customer's own listening history. In each case the number gives reporters something specific to check and quote.

That is why data-led campaigns tend to earn coverage that opinion alone does not. A reporter can write "only 2% of women call themselves beautiful" or "200 million people opened their recap" and point to a source. A claim such as "we care about beauty" gives them nothing to print.

Which five data-driven PR campaigns are worth studying?

The five campaigns below each used a different kind of data and each published a figure that other outlets repeated.

Campaign Data used Headline figure reported Year

Dove, Campaign for Real Beauty Commissioned survey of 3,200 women in 10 countries 2% of women described themselves as beautiful 2004

Coca-Cola, Share a Coke (U.S.) Lists of popular teen and millennial names 250 names on bottles, U.S. sales up about 2.5% 2014

Old Spice, response videos Live comments from social platforms 186 custom videos, body wash sales up 107% in a month 2010

Red Bull Stratos Mission records for altitude and speed More than 8 million concurrent YouTube viewers 2012

Spotify Wrapped Each user's listening history 200 million engaged users in about 24 hours 2025

How did Dove turn a survey into a campaign?

Dove built the Campaign for Real Beauty on its own research. The brand commissioned a global study, The Real Truth About Beauty, which collected data from 3,200 women aged 18 to 64 between February 27 and March 26, 2004. The study's best-known finding was that only 2% of the women described themselves as beautiful, and about three quarters rated their beauty as average.

The survey did two jobs. It gave reporters a headline number, and it gave the creative team a problem to answer. Dove's advertising launched in the fall of 2004 around real women instead of models. The 2007 Ogilvy case study filed with the Advertising Research Foundation describes the research as the foundation for the campaign, with the 2% figure at its center.

One caution: the study was commissioned by the brand, and it measured self-description, not behavior. Quote it as Dove's research, and name the sample and the dates.

How did Coca-Cola use name data?

In 2014 Coca-Cola brought Share a Coke to the United States by printing 250 of the most popular first names among teens and millennials on 20-ounce bottles. The most-stocked names, according to the Wall Street Journal, were Chris, Alex and Jess. The idea had started in Australia in 2011.

Marketing magazine, citing the Wall Street Journal, reported that U.S. sales rose about 2.5% after the rollout, after 11 years of declines. Beverage Digest reported that Coke grew sales volume in 2014 for the first time since 2000. The campaign's website also let visitors build virtual bottles, and one trade report counted more than six million of them.

The lesson is that data does not have to be exotic. A ranked list of names was enough, because the product itself carried the story. Note that the sales figures come from press reports of company data, and other factors can affect a sales trend. Read the original report in Marketing magazine.

How did Old Spice use real-time comments?

After the 2010 Super Bowl spot "The Man Your Man Could Smell Like," agency Wieden+Kennedy asked a harder question: could the character answer fans directly? In July 2010 actor Isaiah Mustafa recorded 186 customized video responses to comments from Twitter, Reddit, Facebook and other platforms. The agency has described shooting them in about three days.

The campaign's own Cannes Lions entry reports that Old Spice body wash sales rose 27% after the February launch, 55% after the second spot and 107% after the response videos, compared with a year earlier. Adweek, citing Nielsen data supplied by Old Spice, reported the same 107% for the last month and 11% growth over 12 months. A separate Effie Awards entry cites a 125% unit sales increase by the end of July 2010, so the exact number depends on the measurement window.

The data here was attention: which comments were rising, who was asking and what could be answered quickly. Mind the caveat that the sales figures are company-supplied and overlap with the Super Bowl advertising, so the videos cannot take all the credit.

How did Red Bull make a record the story?

Red Bull Stratos is a data story in a different sense. The numbers were the event: Felix Baumgartner jumped from about 128,100 feet on October 14, 2012 and reached a top speed reported as 833.9 miles per hour, or Mach 1.24. Because each figure could be checked against an existing record, the media had a clear reason to cover it.

YouTube confirmed that the live stream drew more than 8 million concurrent views at its peak, the most for a live event on the site at that time, a record that Guinness World Records lists. Organizers said more than 130 digital outlets carried the feed.

For a smaller brand the takeaway is not the budget. It is the habit of choosing an outcome that can be measured by someone other than you, so the claim holds up when a reporter checks it. See our guide to PR stunts that worked for more examples of that pattern.

How does Spotify Wrapped earn its coverage?

Spotify Wrapped turns each user's own listening data into a shareable recap. The first version in 2016 was a playlist of top songs, and it grew into a December media event. Spotify said Wrapped 2025 reached 200 million engaged users in about 24 hours, a 19% increase on the year before, and that users shared it 500 million times in the same period. The company counts an engaged user as someone who viewed at least one story inside Wrapped.

For 2024, reaching the same 200 million took about 62 hours, and Spotify said the recap ultimately engaged 245 million users. The 2024 edition was also criticized for its AI-generated podcast hosts, and reports said Spotify removed that feature for 2025. The figures are Spotify's own, reported by outlets such as Music Business Worldwide.

The point for other brands: customers' data, returned to them as a story about themselves, generates its own earned media. It also shows the risk, because a recap that misses the mark gets criticized in public.

What do the five campaigns have in common?

Each campaign followed the same four steps, whatever the budget.

• Start with a number a reporter can check. Dove had a sample size and dates. Red Bull had records.

• Find the human insight inside it. Two percent said they were beautiful, so the campaign was about self-image, not soap.

• Build the creative around the insight. Names on bottles, replies to commenters, a recap about you.

• Publish the result with the method. Name the sample, the source and the date so the figure can be repeated accurately.

The data rarely was the story. The story was what the data said about people.

How do you measure a data-driven PR campaign?

You measure it against the claim you set out to make, using sources outside your own dashboard where possible.

• Pick one headline number before launch. For example, the share of respondents who agree with a finding, or the number of users who open a recap.

• Record the baseline. Sales, search interest or mentions in the four weeks before launch.

• Track earned coverage separately from owned reach. Count placements that repeat your figure and name your source.

• Compare against a control where you can. Old Spice's own sales window overlapped with TV, which shows why.

• Report the method with the result. Note the sample, dates and who supplied each figure.

Our guide to PR KPIs and measurement covers the metrics in more detail, and the earned, owned and paid media framework shows where each figure belongs. For teams adding research to a PR program, digital PR services is the practical next step.

What mistakes weaken data-driven PR?

Five mistakes show up repeatedly.

• Quoting a number without a source. A figure with no sample, date or origin is easy for a reporter to drop.

• Treating a commissioned survey as neutral. Say who paid for it.

• Claiming all the credit. Old Spice and Coca-Cola ran other marketing at the same time as the campaigns above.

• Ignoring how it can backfire. Spotify's 2024 AI features drew criticism despite strong totals.

• Using data for decoration. If the creative would be the same without the number, the number is not driving anything.

Frequently asked questions

What is a data-driven PR campaign?

It is a campaign where research, platform data or customer data decides the idea and supplies the figure that reporters quote. The data comes first, and the creative follows from what it shows.

Do you need a big budget for data-driven PR?

No. Coca-Cola's data was a ranked list of popular names, and Dove's was a single survey. The cost lies mostly in running the research carefully and publishing the method.

How many people should a PR survey include?

There is no fixed number, but the sample must be large enough to defend. Dove's study surveyed 3,200 women in 10 countries. For a smaller brand, state the sample size and dates clearly, and avoid claims the sample cannot support.

How do you prove a PR campaign caused a sales increase?

Usually you cannot prove it alone. Compare against a baseline, note overlapping advertising and report the result as an association, as the Old Spice and Coca-Cola figures should be read.

Which metrics matter most for data-driven PR?

Placements that repeat your figure and name your source, the reach of those placements, and one business metric you chose before launch, such as sales, sign-ups or search interest.

How long should a data-driven campaign run?

It depends on the data. Dove's research fed a campaign that ran for years, while Old Spice's response videos were produced in days. Plan the timeline around when your figure stays newsworthy.