How AI Is Rewiring Shopping Insights from Colin Jeavons of Nomix Group

(00:32) Hello everyone. I'm David Burkowitz. Welcome to another edition of AI Insiders with AI Marketers Guild. We have a fun conversation today with Colin Jvens, a founder and entrepreneur I've gotten to know over the past few months. Thanks also to Tony Winders of the Winders Group, a longtime industry friend, who connected me with Colin and the team. Colin is doing exciting work at the intersection of shopping, commerce, and the creator economy.

(01:13) Affiliate marketing is part of this, with a through line where AI is powering and empowering much of what he's building with the product suite. It's a topic we don't cover enough, and as I dug deeper, I was excited to have Colin share a perspective. Welcome.

(01:39) Thanks, David. I'm looking forward to today and thanks everyone for joining. I'll do my best to answer any questions. I gave a broader, more informal intro, but who are you anyway?

(02:02) For everyone who doesn't know me, I'm Colin Jeans. I've worked in search technology for 27 years. I'm the founder, CEO, and chairman of the Nomics group, which owns several companies designed to create a frictionless consumer shopping experience.

(02:35) Over 27 years in search, I've built many companies. I built the largest political publisher in Europe using search technology. I built a large adtech company licensed to every major publisher. I reversed and owned the largest defense and intelligence technology in search. We owned a contract that built the semantic web, which became a bedrock for AI, machine learning, semantic analysis, knowledge graphs, and entity extraction in 2010.

(03:09) I've worked in AI for a while. One of the most interesting, rapidly changing markets now is shopping. I'm happy to discuss how shopping is changing, how important commerce is to the global economy, and areas of AI that are misunderstood.

(03:42) Many focus on the wonders of ChatGPT, Perplexity, and other agentic systems, but there are many areas where AI is heavily involved and will drive significant change in shopping. I'm happy to discuss them in as much detail as you wish. Let's get a little more detailed.

(04:16) For shopping, there are many phases: awareness, discovery, research, the purchase process across channels, customer experience after, and advocacy. Shopping means different things in different contexts. Shopping for Halloween candy is different from shopping for a new Bentley. Where are the most immediate AI applications? What parts of the shopping process are you most excited about right now?

(04:53) First, significance. Over the past 20 years of digital content and the internet, shopping has been the backseat driver to advertising. When you talk about Google and Meta as the two 800-pound gorillas, advertising is a $600 billion bucket of money. It's a small portion of the conversion value of shopping, which is a multi-trillion dollar bucket of money.

(05:28) Channels of change in shopping include search. How do you discover a product, a price, or a holiday? A year ago, most people used a browser or app, likely Safari or Chrome, entered a query, saw blue links or PLAs (product listing ads), and clicked. Search is radically changing.

(06:39) Click volumes are changing. AI agents and LLMs are changing where consumers start the journey for answers. That has a huge impact on shopping and marketers, and a positive impact on consumers.

(07:03) Another area is UGC, user-generated content. What's the impact of AI-generated content? Photos, stills, editorial, videos—short and long form—that used to cost tens, hundreds, even millions of dollars can now be generated for a fraction of the price and hyper-targeted across the internet.

(07:38) All of these are areas of AI and shopping, and they warrant deep discussion. If there's one takeaway for an analyst, CMO, or CEO of a multinational brand, it's: where will we be 24 months from now?

(08:11) I think shopping will see a huge positive impact from AI. People will discover more, see more, have more opportunities to buy, and ultimately spend more in the next 24 months than in the previous 24 months.

(08:37) In the western shopping market, I expect above a 30% increase in total revenue spent. Those are big statements. I believe there's reason to back them up. There will also be significant losers in shopping. How marketers and publishers handle changes in SEO, SEM, content marketing, and affiliate marketing—and how they reach consumers at the right place and time to drive conversion—will determine outcomes.

(09:21) One question already coming up is where that 30% comes from, and how this happens without alienating people. There's an impression that AI for shopping, especially with social commerce and UGC, floods feeds with low-quality content. How do you reap benefits without going down a lowest-common-denominator path and making things worse?

(10:31) Not everything is better for all people. Whether Meta is a great product depends on perspective. Consumers will have more opportunities to see products than ever, and more opportunities to search for the exact product at the right place and time than ever.

(11:05) If you can cut past poorly created content, you'll be able to window shop when and where you want. You'll also be able to buy with less friction. Historically, the journey to purchase—especially via search—meant lots of links and time-consuming research to get answers.

(12:02) Now, even though it's early days, LLMs are experimenting with better display and experiences. If you want a Prada bag in a certain color, in a certain place, at the best in-store price available today, you no longer need to spend much time to find it. Answer engines make it quick, with less friction and more time saved. Overall, AI is good for consumers in shopping.

(12:41) Speaking of search and delivering what people want when they want it: looking 24 months out, how much comes from Google versus ChatGPT versus platforms like TikTok and Instagram?

(13:07) You have to segment the shopping experience, which isn't easy. For platforms with hundreds of millions of users—Snapchat, Pinterest, Meta, TikTok—ask: do people like looking at products? Do they like window shopping? I think they do. They like discovering things they didn't know they needed. Do they need them? Probably not.

(14:08) Trends already occurring in China will likely occur elsewhere. With video AI—positive and negative—current influencers will present many products to their followers, highlighting things they like. That will likely increase sales.

(14:56) ByteDance reportedly generates more money per quarter than Facebook at times, and roughly a quarter of revenue is now from shopping. Does the platform generate consumption or do consumers want what platforms surface? I believe it's the latter. Ask people if they like shopping: the answer is yes. In search, Google is the goliath, with Bing as secondary. LLMs like ChatGPT and Perplexity are disrupting consumer behavior there.

(16:10) Significant disruptions are already occurring. Search engines, answer engines, and AI engines serve consumers after they've decided to purchase. Intent is there: they need flights to Miami, new outdoor furniture, or a specific product at the best price, location, and time.

(16:33) Answer engines provide better solutions than traditional search engines, which give links and force discovery. If you say, "I want the best-priced automatic cat litter dispenser I can buy today," you'll use an engine that gives you the answer and saves time. Research suggests 15–30% disruption in click orientation from traditional search engines toward publishers or marketers.

(17:42) The impact is huge for marketers. You must manage traditional SEO and SEM and also figure out what to do on LLMs. As someone who's tested too many litter boxes, with the right prompt you might not even need ChatGPT. It's amazing.

(18:07) Shifting to the here and now: it's mid-November and the next four to six weeks are big for holiday shopping. For brands and consumers, what's different this year versus last year where AI is playing a role, visibly or invisibly? What's already changing?

(19:06) Starting with brands: a CMO today lives in a more complex world than 10 years ago, even compared to a year ago. There was above-the-line and below-the-line marketing: TV, newspapers, press. Then the internet era brought SEO, AdWords, and affiliates. About a year ago, ChatGPT and answer engines arrived, plus rapid change in the influencer world.

(19:45) The influencer model was dominated by sponsorships—hire a celebrity for millions and drive massive sales. Now there are hundreds of millions of creators across channels, all recommending products. How do you deal with this model? How do you communicate with this new channel? How do you optimize communication with consumers on LLMs?

(20:25) At Nomics, we're about the economics of shopping—ShopNomics—and making the journey easier for CMOs to reach consumers. We've verticalized our approach. From a consumer perspective: are you more irritated by better accuracy and less friction, or by seeing things you may not need? In both cases, consumers benefit: free, fast paths to products via LLMs and time spent with people they admire who highlight useful things.

(21:55) Consumers are happy with increased product awareness and faster ability to buy. For marketers, it's extra workload due to rapid, radical changes. There are big wins with new audiences, but dominant brands must get marketing right or risk losing share.

(23:22) On influencers: where is AI changing the flow? These markets are evolving, not mature. Companies are finding their feet. We invested in [creator.co](http://creator.co), led by Vinard (CEO) and Rob (chief strategy officer). They are building an AI agent that enables brands to specify the types of creators they want, automating discovery, guidance, and spend with influencers.

(24:26) You're effectively connecting influencers with brands to give consumers a better experience. This will mature over the next 18–24 months. Everyone is a publisher. There are hundreds of millions of influencers; tens of millions have audiences over 10,000 and can be very vertical—angling, travel, fashion, beauty. Brands within each vertical want to influence consumers.

(25:21) As consumers, we want to know about new products in our passion areas—photography, angling, fashion. That's been true for publishing for centuries. Platforms are emerging to simplify this world. We've made a couple of bets in the sector. We may not be 100% correct, but the tide will rise: consumers will discover more, influencers will show more, and consumers will buy more.

(26:20) Devil's advocate: what about those who loathe shopping? I'm not advocating that something is right or wrong. The majority of platform users today like shopping. Personally, I'm only on LinkedIn for work; I'm not the audience. We are in a changing world with new tech, positives and negatives. Over the next 24 months, regardless of personal preferences, behavior is changing. My kids like shopping.

(27:27) If you hate shopping, how does this help? Even if you dislike shopping, you still shop to sustain life—household goods, food, cars. We are creatures of consumption. Some love the experience; others buy out of necessity. If you don't like shopping, you want answers faster with less friction.

(28:57) New AI, particularly answer engines, makes life easier: quicker answers, faster purchases, then back to what you prefer doing. For those who love browsing, they already spend hours daily in social apps. Some will embrace more shopping content, some won't, but the net effect is more purchasing.

(29:59) A good point from the chat: people may not trust AI tools to recommend products, but they may trust people they already follow. Exactly. Influencers with integrity who recommend products they actually like will drive higher conversions than those who promote everything.

(30:41) One big tool for influencers is AI-assisted content creation. Content creation is time-consuming; AI's role here isn't discussed enough. The ability to create content—even video—at scale and speed will increase radically over the next 24 months.

(31:43) There will be good and bad actors. Well-intentioned influencers who test products they recommend will drive higher sales than those who promote indiscriminately. Bottom line: consumers already like much of this. They will consume more and buy more. Fundamentals of advertising are reach and frequency. Familiarity drives consideration.

(33:18) Some brands will avoid AI-generated UGC for fit reasons. For example, a brand like Dove that campaigns around real beauty might avoid AI in visible content but still use AI behind the scenes—for GEO, finding influencers, or other processes. You can embrace some parts of AI and not all of it.

(33:52) Agreed. Not all models will be AI. Certain components will. There's no avoiding new technology. The smart path is how you use it to achieve your objectives. You might use AI to find the right influencers and use LLMs to communicate policies like "we don't use AI imagery"—both can coexist.

(35:34) Let's take audience questions. Adam asks about dynamic landing pages and personalization at scale. Do you see systems built end to end in-house versus best-in-class components? How will this settle out? Think at a high level, not just agentic workflows.

(37:08) Change is opportunity. Will there be needs for thousands of landing page variants? A landing page is content creation on one page; AI will help automate that and bring it to life with sight, sound, and motion. How you get traffic will also change. You may have mastered SEO/SEM and arbitrage to drive traffic and revenue via CPM, CPC, or CPA.

(38:16) For PE firms: in media, there are huge opportunities for companies with rapid growth. For brands, PE might acquire underperformers and apply a proven AI and influencer playbook. Mastering the new world of content, LLMs, AI generation, and influencer marketing will be a competitive advantage, much like Coca-Cola's early Olympic sponsorship.

(39:19) In any changing landscape, there are winners and losers. We've grown quickly by helping brands verticalize traditional publishing, new-world publishing, programmatic, and LLM/AI publishing to drive conversions and sales.

(40:20) Alexander asks about zero-marginal-cost content and avoiding "AI slop." If AI can make infinite content, how do brands deliver value and survive in a world of infinite content?

(41:58) Eric Schmidt once discussed jobs in an AI world and emphasized creativity. Content still requires an editor, much like journalism. Someone must decide what is seen. The role of creative judgment is unlikely to be fully automated.

(43:37) Example: a brand spent $500,000 on a one-day shoot with a famous runner. They wanted to produce thousands of assets from the footage. Now they can, at low cost, without disrupting talent or schedules. But talented people still need to decide what resonates. AI data and surveys will inform decisions, but someone still chooses the smiling cat over the gorilla. The artist in the machine is still required.

(44:43) Daniel asks about interfaces. Do you see invisible interfaces—WhatsApp, voice, email—growing versus direct LLM usage?

(45:17) Invisible interfaces will be huge. Conversational commerce will be huge. A product to try is [wprflow.ai](http://wprflow.ai), a strong speech-to-text engine you can plug into devices and phones. The speech-to-text world has been discussed for a long time; it’s now arriving.

(46:22) Invisible engines and ways to communicate with AI are real. Three-year-olds will be taught by voices from speakers, able to ask any question and get answers. In 10–15 years, children will use these tools in positive ways. As parents and grandparents, it's natural to be concerned. We are living in one of the largest experiments in human history, with AI in the living room, helping create content and access knowledge.

(47:32) It's going to be change, scary and exciting. Let's hope we make it a positive.

(48:01) Change is scary, and AI brings radical challenges. But there can be positive outcomes from incredible technology. Yes, it's disruptive. If embraced, it can be positive. Social media has negatives and positives. This is an opportunity for brands to embrace new tech and for consumers to discover more information at their fingertips. It's up to responsible actors and governments to ensure good management, policing, and regulation so this becomes a net positive for humanity.

(49:20) It's important to focus on elements we can control. Many in this room are building, analyzing data, creating content, or learning. How do we harness this and steer work in the right direction to add value? There are things we control, even if not everything. We'll watch how the holidays play out and continue the discussion. We'll share links in the AMG Slack and on YouTube. Thanks to Colin for being the main event. Have a great rest of the week.