 |
|
|
|||||
The whole idea of advertisingusing pictures and words to get people to buy stuff, or to do somethingis old indeed, with the first known example dating back almost 5,000 years to the heady days of Ancient Egypt. The ads business changed a lot since we were writing notices on papyrus, but one thing thatuntil recentlyremained the same was that it was a deeply intentional business. The advertiser had to think about the language they used, the imagery they employed, the types of people they sought to reach, and how they would go about doing that. Whether the advertiser was touting a weaving shop on the banks of the Nile during the days of the Pharaohs, or selling detergent or cigarettes through new mass media innovations like the television or the radio, that same thoughtfulness was a constant. Advertisers had to thinkand, by virtue of the fact they were forced to make decisions, they were in control of everything. My biggest complaint with the digital ads ecosystem is it, by design, strips the ability of the advertiser to make some of those decisionsnot merely placement, but targeting, and with the emergence of dynamic creative and generative AI, messaging too. In the process, weve turned advertising from a very deliberately engineered systemwhere the architect knows what each part of the process should dointo one thats, essentially, a black box. And within this black box, theres little room for creativity. The Process Is Creative When we think about advertisingand, in particular, good advertisingwe always think about the messaging. Its true that some of the best campaigns in history have always used clever wordplay, or coy psychological tactics, to drive a point home. The Pepsi Challenge, for example, started off as a series of in-person taste tests and culminated in a campaign that could confidently say (though some have identified flaws in the test itself) that Pepsi was Americas preferred cola. Not only did this directly undercut Pepsis main adversaryCoca-Cola, which easily had the most powerful brand perceptionbut it also allowed people to differentiate between products that people might otherwise think of as identical. Messaging is important, but its far from the only creative part of the marketing process. Take Subaru, for example. In the 1990s, it was a struggling car brandeclipsed not only by its Japanese rivals like Nissan and Toyota, but also by fierce domestic competition in the U.S. market. Subaru hired a new advertising firm to turn its fortunes around, which ran a series of focus groups that asked why existing Subaru owners chose its vehicles, as opposed to those from one of its healthier rivals. That firm noticed that women dominated those focus groups, and many of those women identified as lesbian. The company then launched a campaign that targeted both women and lesbiansitself a brave choice, considering the climate of the 1990s, which saw the passage of both the Defense of Marriage Act and Dont Ask, Dont Tell. To help it reach lesbian audiences, it hired Mulryan/Nash to create content specifically for the gay press, and to handle ad buying. This campaign wasnt just pioneeringit also, arguably, helped revive Subarus fortunes, and the brand remains vibrant and relevant, especially in the U.S., where it sold over 667,000 cars in the 2024 calendar year. The Subaru example is a potent one, not simply because it was so successful, but because it illustrates how each step of the processfrom identifying the customer, to determining where to reach them, to crafting the messagingrequired human thought and human creativity. If were looking for a more contemporary example, Spotifys controversial Thanks 2016, its been weird springs to mind. Capitalizing on a year defined by seismic political shifts, celebrity deaths, and countless surreal moments to mention in the confines of this piece, Spotify tapped into its data, identifying equally surreal trends and turning them into highly relatable billboards positioned in prime urban locations. These billboards featured pithy one-liners (for example, Dear person who played ‘Sorry’ 42 times on Valentines Day, what did you do?), with the text localized for target markets (Dear 3,749 people who streamed Its the end of the world as we know it the day of the Brexit vote, hang in there). It served as a reminder of how music isnt simply a form of entertainment, but a way in which we process events in our personal lives, as well as those happening within politics and culture. Again, this process required creative thinking at every levelfrom identifying the patterns within the data that would lead to the funniest trends, to choosing the most valuable locations to place the billboards. I write all of this not because I believe that all digital advertisingwhere these decisions are outsourced, particularly to third-partiesis bad, but because I believe that the most effective and memorable campaigns are thoughtful ones. The reason why I believe digital advertising is the enemy to advertising is because, by design, it strips us of the ability to use that creativity across all stages of the advertising process, from conceptualization to creating the final product. Battling the Black Boxes Last year was the 30th birthday of digital advertising. Its interesting to see how, as the internet grew and an adtech ecosystem emerged, the very nature of how this segment actually works changed. Whereas at one point advertising deals were inked between companies, with money changing hands in exchange for prime placement for a set number of days, those manual transactions are now a thing of the past. Todays digital advertising mechanics are based on systems which the advertiser doesnt control or even understandand in the case of those which heavily rely on AI, even the developers dont have full insight into the factors behind each targeting and placement decision. This opacity also allows the adtech provider or advertising network to act in ways that are contrary to the interests of the advertisereither by obfuscating data that could allow them to make more effective decisions, or by failing to protect said advertiser from, for example, click fraud. Although digital ads allow a company to target and market at scaleand, arguably, with the economies of scale that wouldnt be otherwise possiblethe downside is, arguably, a degradation of the online experience for end users, profound concerns about user privacy, and an absence of transparency for those actually purchasing the ads. Arguably, the biggest downsidefrom someone who cares profoundly about the intellectual and creative brilliance of the ads industryis that digital ads havent really produced something thats memorable, or has had any meaningful cultural impact. Coca-Cola gave us Santa Clauss red outfit and the iconic flashing delivery trucks. Decades after tey commercials first aired, we still remember the Budweiser frogs croaking bud-wise-er, or its later ads that turned wazzup into a legitimate pop culture phenomenon (albeit a really irritating one). And thats because creativity is like a muscle, and if you dont exercise itor dont have to exercise ititll wither away.
Category:
E-Commerce
To announce its entrance into 5G home internet service, Mint Mobile found the real-life version of a new AI-generated actress, even if only in (nick)name. Tilly Norwood is the name of a so-called AI actress launched by AI talent studio Xicoia. It also happens to be the name of a woman who stars alongside Ryan Reynolds in Mint Mobile’s new ad for its home internet service, which its branding “Minternet.” “It’s hard to believe that Mint is launching 5G home internet. It’s also hard to believe that a real version of an AI actress is out there,” a Maximum Effort representative tells Fast Company. “And thanks to the incredible and somewhat disturbing stalking detective abilities of our team, we found her. Just outside of Dallas, Texas, just one day before filming the commercial. Luckily she responded to our random DMs and was happy to assure the world that both she and the internet are very real.” The fake Norwood has inspired backlash and a Wikipedia page, and the labor union SAG-AFTRA refused to say in a statement that the AI character is an actual actor, instead stating its “a character generated by a computer program that was trained on the work of countless professional performerswithout permission or compensation.” As it happens, the real NorwoodNatalie “Tilly” Norwoodis a real Mint Mobile customer. In the commercial, Reynolds, a former Mint Mobile co-owner who still makes ads for the wireless service provider through Maximum Effort, a production company he cofounded, asks if Norwood is real and “not an AI-generated combination of actors.” “I’m a combination of my parents,” the real Norwood says. Mint Mobile’s parent company was acquired by T-Mobile in 2023 in a deal worth up to $1.35 billion, and its 5G home internet service shows the brand is broadening its ambitions beyond mobile. The brand says its home internet service will use T-Mobile’s 5G network, and Mint Mobile is offering it for as low as $30 a month for customers with a Mint Mobile phone plan who prepay for three months. In an advertising landscape that could increasingly see more AI-generated ads, sticking to real people is a smart strategy. A 2024 YouGov poll of respondents from 17 markets around the world found 51% were uncomfortable with a brand creating a virtual ambassador (34% were comfortable with it; 15% didn’t know how they felt about it). In other words, using a fake Tilly Norwood in your ad could turn away half your audience. Meanwhile, the real living, breathing Tilly Norwood appears to be anything but polarizing.
Category:
E-Commerce
The J.M. Smucker Co. says it doesn’t have a problem with other companies selling their own prepackaged, crustless sandwiches like its own popular Smuckers Uncrustables. They just have to get their own design. Uncrustables is on its way to becoming a $1 billion brand, so of course there will be knockoffs, but according to Smucker, a recent Trader Joe’s version of Crustless Peanut Butter & Strawberry Jam Sandwiches is a bit too blatant. The company is using the design of the Trader Joe’s product and packaging to prove its point in a new lawsuit. Smucker accused the grocery store chain of “an obvious attempt to trade off of the fame and recognition” of Uncrustable’s protected design marks in a suit filed Monday in the U.S. District Court for the Northern District of Ohio. The round shape and crimped edges of Trader Joe’s crustless sandwiches, which it released in late summer, look too similar to Uncrustables, Smucker says. [Photos: Smucker’s, Trader Joe’s] “Smucker does not take issue with others in the marketplace selling prepackaged, frozen, thaw-and-eat crustless sandwiches,” attorneys for the Orrville, Ohio-based food and beverage manufacturer wrote in the suit. “But it cannot allow others to use Smucker’s valuable intellectual property to make such sales.” Smucker, which reported annual net sales of $8.7 billion in the most recent quarter, says it has invested nearly $1 billion over 20 years to grow its brand of crustless sandwiches into the No. 1 frozen handheld brand in its category. It’s paid off even as Smucker’s snack brands like Hostess Twinkies and Ding Dongs, which have recently rebranded, struggled in a shifting snack food landscape. CFO Tucker Marshall said on Smucker’s August earnings call that Uncrustables is a “growth brand” for the company, along with the pet food brands Meow Mix and Milk-Bone. Marshall said that “people who are consuming Uncrustables for the most part are athletes, families with kids,” and that the brand performs strongly at universities and convenience stores. “We really haven’t seen any impact at all from the GLP-1,” Marshall added, referring to weight-loss medications that are driving a trend toward healthier, high-protein snacks. The importance of Smuckers Uncrustables in the companys portfolio helps underline the urgency of the lawsuit. In the suit, Smucker argues its trademarks for images like a “pie-like shape with distinct peripheral undulated crimping” as well as “a round crustless sandwich with a bite taken out showing filling on the inside” are being duped by Trader Joe’s without authorization. The suit extends to packaging concerns, as Smucker believes even the blue used for Trader Joe’s box of crustless PB&J sandwiches is strikingly similar to the blue used in the Uncrustables logo. Smucker is seeking damages and demanding that Trader Joe’s destroy all the products, packaging, and promotional materials that use the current designs. Trader Joe’s did not respond to a request for comment. There are other crustless sandwich brands that don’t use the Uncrustables-style circular shape and crimped design, like the square-shaped Jams and Walmart’s Great Value No Crust Sandwich. Chubby Snacks originally launched with circular sandwiches before getting hit with a cease-and-desist from Smucker. It pivoted in 2021 to a cloud-shaped sandwich. Smucker’s suit follows a May lawsuit filed by Mondelez International against Aldi accusing the grocery store chain of duping the packaging of popular snack brands like Oreo and Nutter Butter. Aldi unveiled redesigned private-label packaging in September amid a wider industry trend toward upgrading generic branding that has spanned from Amazon to Walmart. As lawsuits like those from Smucker and Mondelez show, with a rise in private-label competition, the big industry players are ready to protect their own branding, and with teeth if necessary.
Category:
E-Commerce
I was interviewing for a job as a customer service agent with Anna. She had a low, pleasant voice and shed nailed the pronunciation of my namesomething few people do. I wanted to make a good impression except I had no idea what Anna was thinking because Anna couldn’t think. Anna wasnt technically a person. She was AI. Not only is AI changing how we do our jobs, its also changing how we get jobs. This ranges from using AI to screen resumes, schedule interviews, even conduct them. According to a 2025 report, 20% of companies are using AI to interview candidates. Even so, nothing can replace human recruiters, the folks whove deployed Anna into the wild stressed to me. After I spoke with her, I quickly understood why. In this story, paid subscribers will learn: What its like to actually go through a job interview with an AI agentand how to speak to them Where companies should deploy AI interviewers that would benefit them and job seekers AI Anna clocks in Even though I wasnt really interviewing for a jobthis was all an exercise for this story, of courseI was still nervous. I asked the team behind Anna to provide a job description so I could prepare, but outside of this experiment, I was sadly lacking in actual customer service experience. I also didnt know how AI Anna was going to react to awkward silences, panicked misdirection, or if shed be able to tell if I was lying. These worries are bad enough with a human. How would a computer program react? I got on the phone and connected with Anna. She was pleasant, and frankly, sounded way more human than I was expecting. We exchanged greetings, and before long, I was in full-on job interview mode with an AI. First up, she asked me to describe a time when I had to explain something complex over the phone clearly. I blanked. Finally, I described how journalism involves explaining complex ideas because youre asking questions. It sounded weak even to my own ears. Sure enough, she was not impressed. Id like to explore a scenario thats more specific to the role were discussing, she replied firmly. Fair point. I managed to dredge something up from a high school job. Mercifully, AI Anna accepted the answer and moved on. Next, AI Anna wanted me to talk about a time when I had to problem solve for a customer. This, I could answer. I dove into my brief stint organizing a literary conference where writers paid to meet with agents. Occasionally agents went astray because they were hungover or running marathons and Id be left to find alternatives like rescheduling Anna cut me off. That sounds like a high-pressure situation. . . . Its great that you were able to come up with alternatives. Now Id like to switch topics for a moment. Yikes. I wasnt ready to switch topics, but AI Anna was, and I couldnt tell why. Was my example off topic? Was I taking too long to answer her question? Before I could ask, Anna had already swept on to background checks. I invented a criminal background and told AI Anna I had done some time in prison. She thanked me for being honest, and told me that she could not make any decisions. She said candidates with a criminal record would be considered on a case-by-case basis (something that would have to be verified by a human). Then I wanted to know if Id be required to work overtime. She let me know Id be required to do overtime the first six months, but only one or two times a month. Needed, accurate information that couldnt just be googledgreat. Honestly? While I found her transitions a bit jarring sometimes, she handled most questions with aplomb. How we got here AI Anna is the product of PSG Global Solutions, a staffing firm. Before deploying AI Anna in the market, the firm asked Brian Jabarian, a researcher at the University of Chicago Booth School of Business with doctorates in economics and philosophy, to study the AI Annas effectiveness. (Jabarian received no funding from PSG). In a study released in September, Jabarian conducted an experiment where 70,000 applicants for a customer service job were randomly assigned a human interviewer, an interview with AI Anna, or the ability to choose between the two. The results are surprising, and surprisingly promising for the candidates. AI interviews resulted in a 12% increase in job offers, and a 17% increase in 30-day retention on the job. Moreover, when offered a choice, 78% of applicants chose to be interviewed by AI. Jabarian theorized this was because the AI was easier to schedule with: job applicants who needed a job quickly could book a call immediately. Why the positive outcome data? Jabarian pointed out that, on average, an AI interviewer got through more required topics than human recruiters since they couldnt be distracted. (I mean, Anna did move at a brisk clip.) AI leads to a more consistent interview experience, he said. It lets the candidate talk more, and has a 50% chance of covering 10 of 14 required topics compared to 25% for human interviewers. AI Anna clocks out Afterwards, I debriefed the interview with David Koch, PSGs chief transformation and innovation officer. First, he showed me AI Annas backend: The platform had generated a recording of our conversation, a transcript, a summary of the call (including suggestions for next steps, like a follow up to discuss my criminal background), and an overall recommendation: AI Anna thought I was qualified (yay!) but merited human follow-up because of my criminal background. AI Anna also recommended a follow-up because shed cut me off when I was talking about the literary conference. Koch explained my speaking cadence is a touch slower than average, and AI Anna is programmed to respond after a certain amount of time or else the flow of conversation can become jerky. Koch noted that AI interviewing was better suited for some situations and not others. He recommended AI interviewing for high-volume hiring where theres a need to source candidates quickly for jobs that are seasonal and high turnover, like customer service agents or travel nurses. Koch also said AI interviewing is best suited for cases where theres less complexity, in which you dont need to sell a candidate on a role. From my standpoint as a lay person, the technology behind AI Anna struck me as marvelous. She corralled me into staying on topic, and was capable of social niceties. She provided detailed answers to all my questions. For recruiters, this could be life changing. Its not that AI Anna might replace them, per se (there were already things from the interview that a human colleague would have to address or follow up on). But recruiters could farm out tasks like screening calls to AI while they worked on more hih-level tasks. However, this made me worry. If AI Anna existed to save companies time, what happens to candidates who get flagged for follow-up, even for something as simple as speaking slowlylet alone a criminal background? If there are more than enough qualified candidates to fill roles, I imagine a harried hiring manager would make offers to people who dont require follow-up. Exception cases that require more time, like me, might fall to the wayside. The future: cold, but competent After my conversation with AI Anna, I felt hollow. Typically, if an interview goes well, I have the high of having connected with someone who might make me feel valued, desired, and possibly in the mix for a new job. If it doesnt go well, I spend the next couple of days wallowing in self-pity and dissecting potential red flags. AI Annas preprogrammed human-like intonation left me nothing to go on. Did she like me? Or was meh on me, but still think I was qualified? I couldnt tell probably because AI Anna does not have emotions and did not care about me. But how much does this matter? A Gallup study found that 44% of respondents said their interviews drove them to accept an offer or not. Ideally, candidates would be able to interview with their direct supervisor before getting a job in order to suss out personality matchbut for a screening interview, AI Annas value was undeniable. She raised the floor for interview quality: Shes personable and she offers a consistent experience. Theres no need to worry about the mysterious intangible of chemistry. Jabarian also pointed out that AI interviews reduce gender discrimination by half. Done right, AI interviewers could reduce bias and help qualified job candidates who may not perform well during interviews because they lack intangibles such as charisma. Still, I missed talking to a human.
Category:
E-Commerce
Generative AI is evolving along two distinct tracks: on one side, savvy users are building their own retrieval-augmented generation (RAG) pipelines, personal agents, or even small language models (SLMs) tailored to their contexts and data. On the other, the majority are content with LLMs out of the box: Open a page, type a query, copy the output, paste it elsewhere. That dividebetween builders and consumersis shaping not only how AI is used but also whether it delivers value at all. The difference is not just individual skill. Its also organizational. Companies are discovering that there are two categories of AI use: the administrative (summarize a report, draft a memo, produce boilerplate code) and the strategic (deploy agentic systems to automate functions, replace SaaS applications, and transform workflows). The first is incremental. The second is disruptive. But right now, the second is mostly failing. Why 95% of pilots fail The Massachusetts Institute of Technology recently found that 95% of corporate GenAI pilots fail. The reason? Most organizations are avoiding friction: They want drop-in replacements that work seamlessly, without confronting the hard questions of data governance, integration, and control. This pattern is consistent with the Gartner Hype Cycle: an initial frenzy of expectations followed by disillusionment as the technology proves more complex, messy, and political than promised. Why are so many projects failing? Because large language models from the big platforms are black boxes. Their training data is opaque, their biases unexplained, their outputs increasingly influenced by hidden incentives. Already, there are companies advertising SEO for GenAI algorithms or even Answer Engine Optimization, or AEO: optimizing content not for truth, but to game the invisible criteria of a models output. The natural endpoint is hallucinations and sponsored answers disguised as objectivity. How will you know if an LLM recommends a product because its correct, or because someone paid for it to be recommended? For organizations, that lack of transparency is fatal. You cannot build mission-critical processes on systems whose reasoning is unknowable and whose answers may be monetized without disclosure. From out of the box to personal assistant The trajectory for savvy users is clear. They are moving from using LLMs as is toward building personal assistants: systems that know their context, remember their preferences, and integrate with their tools. That shift introduces a corporate headache known as shadow AI: employees bringing their own models and agents into the workplace, outside of ITs control. I argued in a recent piece, BYOAI is a serious threat to your company, that shadow AI is the new shadow IT. What happens when a brilliant hire insists on working with her own model, fine-tuned to her workflow? Do you ban it (and risk losing talent) or do you integrate it (and lose control)? What happens when she leaves and takes her personal agent, trained on your companys data, with her? Who owns that knowledge? Corporate governance was designed for shared software and centralized systems. It was not designed for employees walking around with semiautonomous digital companions trained on proprietary data. SaaS under siege At the same time, companies are beginning to glimpse what comes next: agents that do not just sit alongside software as a service (SaaS); they replace it. With enterprise resource planning systems, you work for the software. With agents, the software works for you. Some companies are already testing the waters. Salesforce is reinventing itself through its Einstein 1 platform, effectively repositioning customer relationship management, or CRM, around agentic workflows. Klarna has announced it will shut down many SaaS providers and replace them with AI. Their first attempt may not succeed, but the direction is unmistakable: Agents are on a collision course with the subscription SaaS model. The key question is whether companies will build these platforms on black boxes they cannot control, or on open, auditable systems. Because the more strategic the use case, the higher the cost of opacity. Open source as the real answer This is why open source matters. If your future platform is an agent that automates workflows, manages sensitive data, and substitutes for your SaaS stack, can you really afford to outsource it to a system you cannot inspect? China provides a telling example. Despite being restricted from importing the most advanced chips, Chinese AI companies, under government pressure, have moved aggressively toward open-source models. The results are striking: They are catching up faster than many expected, precisely because the ecosystem is transparent, collaborative, and auditable. Open source has become their work-around for hardware limits, and also their engine of progress. For Western companies, the lesson is clear. Open source is not just about philosophy. Its about sovereignty, reliability, and trust. The role of hybrid clouds Of course, there is still the question of where the data lives. Are companies comfortable uploading their proprietary knowledge into someone elses black-box cloud? For many, the answer will increasingly be no. This is where hybrid cloud architectures become essential: They allow organizations to balance scale with governance, keeping sensitive workloads in environments they control while still accessing broader compute resources when needed. Hybrid approaches are not a panacea, but they are a pragmatic middle ground. They make it possible to experiment with agents, RAGs, and SLMs without surrendering your crown jewels to a black box. The way forward Generative AI is splitting in two directions. For the unsophisticated, it will remain a copy-and-paste tool: useful, incremental, but hardly transformative. For the sophisticated, its becoming a personal assistant. And for organizations, potentially, a full substitute for traditional software. But if companies want to make that leap from administrative uses to strategic ones, they must abandon the fantasy that black-box LLMs will carry them there. They wont. The future of corporate AI belongs to those who insist on transparency, auditability, and sovereignty, which means building on open-source, not proprietary, opacity. Anything else is just renting intelligence you dont control while your competitors are busy building agents that work for them, not for someone elses business model.
Category:
E-Commerce
Sites : [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] next »