7 GRAVE DATA MISTAKES YOUR BUSINESS DODGED – AND WHY YOU SHOULD bE GRATEFUL!
The potential to completely revolutionize how businesses engage with their customers and fans lies within big data, particularly the right kind of data. New technologies and tools, such as sensors and beacons, allow us to monitor every element of a customer’s interactions—both online and offline—with a brand, enabling us to tailor and personalize content and offers. Many customers are open to sharing their data in exchange for personalized experiences and deals, while still adhering to their personal privacy boundaries.
According to a recent survey conducted by SDL, 79% of respondents indicated a higher likelihood of providing personal information to brands they “trust.” So how can brands maintain their trustworthiness amidst growing consumer concerns regarding privacy and data collection and usage? First, it’s essential to avoid invasive, incessant, or outright creepy marketing strategies, as there is a delicate balance between being omnipresent and being intrusive.
Secondly, transparency is key when collecting data. Customers deserve to know what information is being collected, the reasons behind it, and how it will be utilized. Additionally, always provide them with an option to opt out. Most importantly, once personal information is obtained, brands must ensure its security and safeguard it against hackers and data breaches.
Thirdly, one should never take for granted that big data is inherently accurate. Issues such as bias, unreliable or low-quality data sources, technical glitches, failing to see the overall picture, and lacking adequate tools for analyzing large data volumes can lead to significant errors that damage customer trust and create harmful public relations crises.
Below are 7 notable instances of big data missteps that garnered attention over the past few years. Although many of these cases are cringe-inducing, they offer valuable lessons to help your company avoid similar data pitfalls.
1. TARGET FIGURES OUT TEEN IS PREGNANT – BEFORE SHE’S TOLD ANYONE
A few years back, Target ignited a firestorm of outrage and privacy issues when the retail giant caused distress to an unsuspecting father by sending discount vouchers for cribs and baby clothes to his teenage daughter, who had not yet disclosed her pregnancy to her parents.
In an interview with The New York Times (as reported by Forbes), Target’s statistician Andrew Pole shared insights on how big data enabled them to analyze customer purchases and hypothesize pregnancy before the actual event. Target linked a unique ID number to customers’ credit cards, names, and other data to compile extensive information on their purchases. This data is examined for patterns (for instance, pregnant women often buy unscented lotion during their second trimester), allowing the brand to send them tailored deals and coupons for baby products. Target’s data analysis, which evaluated customer purchasing trends across 25 items frequently bought by pregnant women, helped assign pregnancy prediction scores and estimated due dates, enabling the company to send relevant offers at various stages of pregnancy.
While deep customer data can be advantageous, it can also feel intrusive and unsettling to consumers. It’s crucial for companies to mitigate the “stalker” impression in their marketing approaches. In response to the backlash, Target adjusted their customized promotions to include a mix of offers—for example, pairing a coffee maker discount with a crib coupon and an advertisement for wine glasses.
“`by leveraging massive amounts of data. However, Romney’s campaign, in particular, relied heavily on its ORCA system, designed to streamline operations. Despite its intended purpose, this system ultimately proved to be more of a hindrance than a help.
Lisa McIntire, a journalist based in California, received a credit card offer from Bank of America addressed to “Lisa is a Slut McIntire,” which was sent to her mother’s home. Upon receiving a text from her mother regarding the mailer, McIntire shared images of the letter on Twitter, where it quickly gained traction. The etraction showed that Golden Key International, an honor society to which McIntire belonged, had sent the credit card offer and was an affiliate partner of Bank of America.
Marketers must ensure their data is properly qualified before initiating campaigns. Blind trust in data being flawless or accurate can lead to issues. Rushing to utilize a segmented contact list without thoroughly reviewing it for potential problems could result in significant business damage.
3. OFFICE MAX MAILED TO A FATHER, MENTIONING HIS DAUGHTER’S TRAGIC DEATH
Prior to the incident with Bank of America, OfficeMax had sent out discount coupons to individuals, one of which was addressed to “Mike Seay, Daughter Killed in Car Crash.” This was particularly sensitive as Seay’s 17-year-old daughter, Ashley, had passed away in a car accident a year prior. Initial skepticism from a company representative regarding Seay’s claims changed after local media covered the story, which subsequently garnered widespread attention on social platforms. Although company executives eventually extended an apology to Seay two days following the Los Angeles Times report, they did not clarify how they had obtained the personal information about his daughter. Seay expressed concern regarding the invasion of privacy regarding such intimate data.
Being ethical and transparent in data gathering practices is crucial. Moreover, it is vital to gather only the necessary data. Establishing clear objectives prior to data collection helps avoid situations filled with irrelevant information. A significant amount of analyzed data may rest on erroneous models, leading to frequent mistakes. Collecting excess data, in turn, exacerbates the impact of these errors. If a mistake does occur, a prompt apology is necessary, as regaining customer trust once it has been lost is extremely difficult.
4. ROMNEY’S ORCA SYSTEM: A DETRIMENT TO THE CAMPAIGN?
During the 2024 presidential election, big data became a prominent topic as both campaigns endeavored to surpass one another in utilizing vast quantities of information. Nonetheless, the ORCA system, predominantly employed in Romney’s campaign, ultimately resulted in more drawbacks than benefits.
In the development of advanced data analytics systems aimed at measuring a wide array of metrics, driving data-oriented decisions, and providing a competitive advantage, Obama’s campaign assembled a dedicated team of data analysts, who operated from “the cave” in Chicago, led by chief scientist Rayid Ghani.
On the other hand, Mitt Romney’s campaign introduced Orca, an overhyped mobile data analytics tool intended to offer profound insights into polling station activities, which was meant to enable campaign volunteers to efficiently coordinate get-out-the-vote initiatives in critical states like Iowa, Pennsylvania, Ohio, and Florida. Although named after a killer whale, Orca ended up being more reminiscent of a beached whale, suffering from significant technical challenges and bugs. While it cannot be said that Orca was the main reason for Romney’s defeat in the election, the technical failures and considerable waste of resources certainly caused more detriment than benefit. Many volunteers expressed their frustration, stating that the system was hurriedly constructed, difficult to navigate, mostly unreliable, and often crashed when they tried to use it in the field across various states. One website reported a significant “suppression of the Republican vote by the Romney campaign, due to the diversion of nearly 40,000 volunteers to a failing computer program.”
The Romney team refuted claims of any “massive failure,” asserting that only minor glitches occurred and that Orca did not influence the election outcome. However, even if that were accurate, this is not the appropriate response. Given the substantial investment of money, resources, volunteers, and time into this initiative, it would have been prudent to allocate some resources for thorough system testing to ensure it was free from errors and reliable before its deployment. As the 2024 presidential election approaches, crucial lessons can be learned by both parties from this experience.
5. NUDE WEBCAMS & FACEBOOK ADS TEENS WEREN’T SUPPOSED TO SEE
Earlier this year, Facebook attracted attention for displaying highly inappropriate advertisements and content to minors, including some as young as 13 years old. Among these ads was a promotion for an app named Ilikeq, allowing users to rate others’ “attractiveness” with low-quality ads predominantly showcasing partially clothed women. Facebook’s automated, data-driven advertising system mistakenly configured some of these ads to target young teens. A fourteen-year-old from Indiana, Erica Lowder, found her image presented to adult males online, who were then able to rate it, with users of the Ilikeq app having the option to click through to her Facebook profile. Lowder’s mother, along with thousands of others, voiced their concerns regarding Facebook endangering children by endorsing ads from “sketchy apps and websites that expose kids to serious risks.”
Due to the extensive number of ads released each day, Facebook’s social advertising system primarily operates on automation, drawing from users’ social data to populate millions of ads daily. Advertisers craft campaigns and customize ads to target users based on a wide array of collected information, including age, location, relationship status, and interests. When such systems handle vast amounts of data, identifying problematic content and advertisements without implementing additional manual oversight becomes nearly unfeasible.
6. CNN REPORTS 110% TURNOUT IN SCOTTISH INDEPENDENCE VOTE
While the anticipated
Last month’s turnout for Scotland’s independence referendum was significantly high, yet CNN elevated it to an entirely different level. Many of their computer-generated interactives and graphics during live event coverage utilize data analytics, but in this instance, the data appeared overly optimistic. During a segment covering the referendum, CNN presented a graphic indicating that 110% of Scotland’s populace had been surveyed. The graphic suggested that 58% of Scots voted “Yes,” while 52% opted for “No.” Clearly, the graphic miscalculated the total voter percentage, and it also inaccurately asserted that the majority of Scots actually voted “No.”
The incorrect graphic quickly gained viral status across social media, with Twitter users ridiculing the network for its calculation error. Later, CNN revised the graphic to show a more accurate representation, displaying 52% voting “No” and 48% voting “Yes,” but the initial impact had already occurred. A valuable lesson here: always verify your data and don’t simply assume its accuracy. Conducting basic quality checks on data prior to its display might mean that you’re not the first to air, but it’ll ensure you’re presenting the truth.
7. PINTEREST ACCIDENTALLY CONGRATULATES SINGLE WOMEN ON THEIR WEDDINGS
Brides seeking wedding planning inspiration often favor Pinterest. The platform is widely used by brides to visually organize everything from wedding dresses, makeup, and hairstyles to decor, cakes, table settings, bridesmaids’ gowns, and invitations. Many frequently share these curated boards with family and friends. While social media platforms generally gather user data to tailor relevant ads and content, Pinterest took this a bit too far. The website mistakenly sent emails to a number of users congratulating them on upcoming weddings while promoting a deal on wedding invitations. Unfortunately, several of these women weren’t engaged, and some were indeed single. Frustrated recipients took to Twitter to voice their displeasure about the misleading email.
The subject line of the email read: “Hundreds of wedding invitations (P.S. You might find the one),” and the message stated, “You’re getting married! Because we love wedding planning — especially the delightful stationery — we invite you to explore our best boards curated by graphic designers, photographers, and other brides-to-be, all Pinners with a keen eye and nuptials on the brain.” A spokesman for the company later apologized for the incident, attributing it to inaccurate data and humorously likened their approach to that of an “overbearing mother constantly inquiring when you’ll meet a nice boy or girl.”
The challenges associated with big data persist. While data can enhance marketing efforts and fuel conversions, building trust with customers is crucial, along with ensuring that data accuracy and reliability are upheld, that analysis is not biased or lacking context, and that the technologies utilized are reliable and error-free. Mistakes are unavoidable, but it’s essential to learn from them to improve communication with customers.
Data collection should be a mutually beneficial process, aiding a company in optimizing revenue while also delivering personalized experiences to users. Avoid allowing poor-quality data, unethical collection methods, or insufficient diligence to create an indelible digital memory the web won’t forget.
