Customer trust vs algorithmic targeting

Only 32% of banks globally see significant returns from customer-facing AI investments, despite 99% prioritizing them. Meanwhile, 68% agree the real value comes from back-office efficiency. That gap? That's the sound of value being extracted, not created. The algorithm sees that you're a 32-year-old professional with consistent salary deposits and increasing discretionary spending. It knows you recently searched for property listings. It calculates—with 94% confidence—that you're a high-probability candidate for a personal loan plus credit card combo. What it doesn't know: You're already carrying debt from family obligations that doesn't show in your banking data. You're considering a career transition that might reduce your income. Your discretionary spending spike was wedding-related, not sustainable. The AI recommends. You accept. Twelve months later, you're financially stretched. When cross-sell recommendations flow through a journey orchestration layer—not just a recommendation engine—something fundamental changes. The system isn't just asking "what can we sell next?" It's asking "what does this customer's complete financial journey require to achieve their stated goals?" The difference isn't semantic. It's structural. A journey-first architecture forces banks to: ✔️ Make AI recommendations transparent and contestable ✔️ Connect product suggestions to explicit customer objectives ✔️ Surface the logic behind algorithmic decisions ✔️ Enable customers to override "optimal" recommendations ✔️ Measure success by financial wellbeing metrics, not just product adoption Consider the same scenario through this lens: The platform detects property search behavior and increasing spending. But before recommending credit products, it prompts a conversation about financial goals. It surfaces debt-to-income projections. It models scenarios. It suggests speaking with a human advisor for major decisions. The customer might still take the loan. But they do so informed, not algorithmically nudged. #banking #AI #Backbase #trust #transparency #valuecreation #algorithms

The Personalization-Privacy Paradox: AI in customer experience is most effective when it personalizes interactions based on vast amounts of data. It anticipates needs, tailors recommendations, and enhances satisfaction by learning individual preferences. The more data it has, the better it gets. But here’s the paradox: the same customers who crave personalized experiences can also be deeply concerned about their privacy. AI thrives on data, but customers resist sharing it. We want hyper-relevant interactions without feeling surveilled. As AI improves, this tension only increases. AI systems can offer deep personalization while simultaneously eroding the very trust needed for customers to willingly share their data. This paradox is particularly problematic because both extremes seem necessary: AI needs data for personalization, but excessive data collection can backfire, leading to customer distrust, dissatisfaction, or even churn. So how do we fix it? Be transparent. Tell people exactly what you’re using their data for—and why it benefits them. Let the customer choose. Give control over what’s personalized (and what’s not). Show the value. Make personalization a perk, not a tradeoff. Personalization shouldn’t feel like surveillance. It should feel like service. You can make this invisible too. Give the customer “nudges” to move them down the happy path through experience orchestration. Trust is the real unlock. Everything else is just prediction. #cx #ai #privacy #trust #personalization

Having dim sum and tea on a Sunday morning in Katong, Singapore where we talk about how you can't just AI your way into credibility and how that's actually great news if you're real. AI makes everything cheaper and faster to produce, which means the good, the bad, and the mediocre all scale up at once. Lazy people were already lazy, now they can be lazy at scale. Scammers were already scamming, now they can automate it. The problem is at the systemic level, because shortcuts and noise are easier to scale than quality, so we get marketing spam, auto-generated content, engagement bait, all the stuff that clogs everything up and makes it harder to find signal in the noise. When everyone can generate a polished pitch or a professional website in seconds, the differentiator becomes proof and personal credentials, not the output. Who are you, what have you done, will you put your name on it? This is why your "About Us" page with real profiles now matters more than your homepage with generic stock images. GenAI made it trivial to look legitimate, but it can't easily fake your track record or whether real humans with real expertise are willing to stake their names on your work. AI floods the zone with content, which means personal recommendations matter more than ads, known expertise matters more than polished marketing, and direct relationships matter more than algorithmic reach. Trust is becoming expensive again, not because AI made it harder to build trust, but because it made it easier to fake the appearance of it. If you're just amplifying noise, you're part of the problem everyone is learning to filter out. But if you're real, if you've done the work, if actual humans with actual track records stand behind what you're building, that matters now more than it has in years.

ABM without trust is just spam!!!! For years, ABM’s been seen as the silver bullet. But for me, the real silver bullet is trust. I was listening to Dan Sanchez’s latest podcast, AI Hits Its Tipping Point, and it really hit me how skeptical we’re all becoming. For me if something doesn’t look believable, I instantly scroll past. I just don’t trust what I read anymore. AI has democratised content, but it’s also commoditised credibility. The brands winning with ABM right now are doing it before the targeting and orchestration. They’re laying down trust foundations first. Forget the tech stack, in my view the new ABM stack should focus on reputation, relationships, and real people. 💫 Reputation before reach If your brand isn’t already known, liked, and trusted in your target accounts, no tech stack will save you. Build authority first. Borrow it if you need to. Get your thought leaders out there. 💫 People over platforms Buyers want humans. Dan talks about people gravitating towards personal brands because they represent lived experience and values as opposed to algorithmic content. Some of our most successful campaigns have been where employees became the brand ambassadors! 💫 Face-to-face is back There’s a stat that says 66% of ABM programs don’t include events. I’d argue it’s not an ABM program without one. Nothing builds trust faster than shared space, eye contact, and unfiltered conversation. 💫 Community and word-of-mouth We trust each other more than any campaign. Create spaces (online or in person) where your customers and prospects can connect. That’s your most scalable trust engine. 💫 Proof beats promises Testimonials are fine. But real-time evidence is better. Show how you work, share your thinking, your results, your failures. Transparency compounds trust. As Joel Harrison often says, “B2B is still P2P (person to person).” And in an AI world where sameness reigns, that human connection IS the competitive edge. So before you roll out your next ABM campaign, ask yourself: Have we earned the right to show up? Ps. Yes I did go out and specifically buy a can of SPAM to create this stock image, who the hell still eats this stuff???

“Show me the incentive and I’ll show you the outcome.” For years, the incentive in algorithm-based social media has been pretty evident: maximize engagement, get more clicks, keep people scrolling. Essentially, manufacture screen addiction. That led to … “outcomes,” but it’s proven to be pretty bleak for the consumer. Meanwhile, platforms that prioritize trust AND positive consumer experiences (shoutout Pinterest and of course tvScientific) prove that you don’t have to sacrifice performance to get there. When consumers trust the environment, they’re more open, and as a result, more likely to take action. Crucially, that trust carries through to the brands that show up there. At tvScientific, we think a lot about what we’re actually monetizing. We’re focused on high-quality media and environments that are additive for the consumer (not addictive). Over time, trust compounds. When it does, it drives better outcomes for everyone, including consumers, advertisers, and platforms. The idea that there’s a tradeoff between trust and performance simply doesn’t (and won’t) hold up.

CES content has dominated my LinkedIn feed this week. Here's what it reveals about the next revenue challenge for companies in growth mode. CES always showcases what tech companies think the future looks like. This year: AI agents, spatial computing, personalized everything. But here's what the tech demos consistently miss: The future isn't just about what technology can do. It's about who trusts it enough to use it. And trust doesn't distribute evenly across demographics. Research consistently shows that multicultural audiences, particularly Black and Hispanic communities, adopt and interact with new technology differently than white audiences. Not slower. Not less sophisticated. Differently. Multicultural consumers are more likely to: Trust recommendations from community voices over algorithms. Adopt mobile-first solutions before desktop. Value human connection even in automated experiences. Prioritize cultural relevance over technical novelty. I believe that this creates a strategic dilemma for companies. If you are investing heavily in AI-driven personalization, automated content creation, and algorithmic recommendations, but your fastest-growing consumer groups are in markets like Houston, Dallas, Atlanta, and Philadelphia, it is time to reevaluate your marketing strategy. Consumers trust local personalities and community voices more than they trust machines. You're optimizing for yesterday's audience, not tomorrow's revenue. The companies that will win in the next five years will have more than just the best AI infrastructure. They'll be the ones that understand where technology amplifies human trust, and where it erodes it. This is not a question CES answers. It's cultural intelligence translated into competitive strategy. And right now, most companies pouring millions into AI aren't asking it. The tech is impressive. But without cultural strategy, it's optimization in the wrong direction.

𝐀𝐈 𝐢𝐧 𝐌𝐚𝐫𝐤𝐞𝐭𝐢𝐧𝐠: 𝐅𝐫𝐨𝐦 𝐇𝐲𝐩𝐞 𝐭𝐨 𝐑𝐞𝐚𝐥𝐢𝐭𝐲, 𝐁𝐮𝐭 𝐓𝐫𝐮𝐬𝐭 𝐈𝐬 𝐋𝐚𝐠𝐠𝐢𝐧𝐠 𝟗𝟐% 𝐨𝐟 𝐦𝐚𝐫𝐤𝐞𝐭𝐞𝐫𝐬 are now using AI in their daily work. What used to be theory is now practice. Campaigns are launched faster, teams spend less time on repetitive tasks, and customer engagement is improving. But consumers are telling a different story. The latest 𝘚𝘈𝘗 𝘌𝘮𝘢𝘳𝘴𝘺𝘴 𝘳𝘦𝘴𝘦𝘢𝘳𝘤𝘩 highlights a growing personalisation gap: ⭕ 40% of shoppers say brands don’t understand them as individuals, up from 25% last year. ⭕ 60% feel marketing emails remain irrelevant. ⭕ 63% globally don’t trust AI with their data, with UK shoppers even more sceptical at 76%. This lack of trust threatens to unravel the personalised experiences brands are working so hard to deliver. Regulations like the EU AI Act are pushing businesses to act more responsibly, but the challenge is to do so without stifling innovation. Some companies are showing what’s possible. 🔹Gibson uses AI to free up staff for more creative thinking in a highly artistic business. 🔹Australian retailer City Beach applied AI to predict churn at an individual level, winning back nearly half of customers who were about to leave. The difference is clear: they use AI to solve real problems for people, not just to automate. 𝐌𝐲 𝐭𝐚𝐤𝐞𝐚𝐰𝐚𝐲𝐬: ➡️ AI can accelerate campaigns and bring efficiency, but it cannot replace trust. ➡️ Relevance must go beyond surface-level personalisation. Consumers want value they can feel. ➡️ Transparency is non-negotiable. Brands need to show how data is being used and why it benefits the customer. ➡️ The best results come when AI is used to empower people; staff and customers alike rather than just reduce costs. As we step further into the “Engagement Era,” the message is simple. Technology isn’t the problem. Trust is. How do you see brands bridging the gap between efficiency and empathy in their AI use?

There's so much chatter about AI agents for customer support. Everyone's shipping them. But does anyone actually use them for anything important? I’ve tested most tools out there, and once you move beyond a single product, they become unusable. The issue goes beyond the fact that AI makes mistakes. Humans do too. What makes this dangerous is that AI makes mistakes with absolute conviction. Let’s take a simple product example. You have two FD products with different early withdrawal penalties. A customer asks, “What’s my early withdrawal penalty?” They leave out which FD they hold, because they no longer remember. Your AI has to pick one. It chooses an FD, answers in a perfectly confident tone, and anyone reading it would trust the response. The answer can still be wrong. That gap is what I mean: highly confident answers going out in situations where the details actually matter As time goes on and there is more data to train the models, things will get better but there is still that chance of hallucination. I know what you're thinking: "But customers expect fast support. Aren't you just being slow?" Customers do want fast support. But they want accurate support more. The moment a customer makes a financial decision based on wrong information from your AI, speed becomes irrelevant. They leave. They tell their community. And just like that you become untrustworthy at scale. Are you building for speed or for trust? Because it takes more than 100% automation if you wish to do both. 100% automation at scale means shipping AI everywhere, accepting that some answers are wrong but delivering them fast at scale. It means betting customer complaints won't outpace new signups. It’s hoping regulators don't notice. Trust means deploying AI only where context is clear, keeping humans in charge of anything with financial consequences. It's slower. But it actually builds the thing your customers chose you for. In the end, when your AI gives wrong information and a customer suffers harm, you’re liable. Not the AI. In the coming years, the companies that truly stand out will be the ones whose customers never have to worry about whether their financial information is accurate in the first place. Speed will always matter, but if we start treating it as equal to trust, we risk building products that look beautiful in launch decks but disappoint the people using them when it really counts. I’d love to see the next wave of fintech focus on this: technology that feels not just impressive, but genuinely dependable. #AI #Customer #Fintech

Customers want you to know them better, but they also want you to know less about them. As we get started on 2026, those contrary expectations will only get stronger. Now we’ve hit January, I'm thinking about what’s really going to change over the next twelve months. The technology will continue to evolve, obviously, but I think the more interesting change will be in the trust equation itself. Customers have grown to expect Netflix-level personalisation while simultaneously growing more sceptical about what happens to their data. They've been reading the headlines, experiencing the spam, and they're (rightly, in my view) warier than they once were. The firms that succeed in balancing these expectations will be the ones that make customers feel genuinely in control of how their data is used. It’s not going to be enough to just use the data in a clever way (have we seen too many ‘wrapped’ posts now?!) in 2026. In my experience, this means building transparency and customer control into the product itself: don't bury privacy settings in a menu and make opting out easy. Counterintuitively, the easier you make it to leave, the more willing customers are to stay. It also means showing your working. Has the AI recommended something? Explain why. Used customer data to make a decision? Show them what you learned and what value they got in return. Every data use should answer an implicit question: what did the customer gain from this? I've written before about cognitive offloading in AI deployment and the same principle applies here. AI should handle the transactional while humans handle the emotional. But there's a third dimension now: customers need to feel that AI is working in their interest, and isn’t something that is being done to them. The moment that belief changes, their trust is lost. In regulated industries, we're somewhat ahead of the curve; compliance frameworks have forced us to build trust mechanisms that will become standard elsewhere. But meeting requirements and earning trust are different things, and customer expectations are evolving faster than ever before. This is fundamentally a leadership challenge. It requires aligning the CFO, CRO, and the COO around a shared understanding: customer trust hits the bottom line. LTV, churn, ARPU... all of these sit downstream of whether customers believe you're using their data in their interests. Are we ready to re-earn customer trust in 2026?

A few years ago we ran an uber style referral model experiment that outperformed everything else we tried. Not in volume, but in quality It was pretty simple... If a visitor filled a landing page form in and they switched energy providers, they got a £20 voucher But immediately after the form was submitted, we sent a text with a unique link: "Share this with a friend. If they switch to a deal, you both get £20" The original lead could earn £40 if they and the referral switched. The intro got £20 These referrals were never our biggest volume driver, but they were consistently the best performers At the time, we called it referral based lead nurturing, but it was more like a filter based on confidence. Sharing that link meant putting your own reputation on the line Because the people who shared it only sent it to people they believed would switch That single behavioural filter removed more low intent leads than anything Right now lead gen economics are shifting, particularly in B2C, and mainly due to AI AI is being used in two very different ways... 1. as a tool to push low intent records at mass scale through automated outreach, programmatic ads, and hyper targeted retargeting 2. as a decision engine that depends on clean data to provide credit decisions, dynamic pricing, personalised offers, churn prediction etc As AI makes outreach cheaper and easier, everyone’s doing it - Automated calls, automated emails, ads running 24/7 etc Ad costs haven’t necessarily changed, but the organic layer (us, the messy human part) has been stripped out and replaced with hyper precision targeting, instant A/B testing, and algorithmically perfect messaging On paper it works - Campaigns run faster. Targeting is sharper But, when I see an ad that says 'Simon - Provero' or mirrors my situation with perfect accuracy, I don’t think 'this is exactly what I need' I think, 'I'm being sold to' and that's if it resonates at all The efficiency kills the magic Trust collapses when automation is obvious. People can feel when they’re being targeted, and the more precise it gets, the more it triggers ambivalence instead of interest But when a friend recommends something, or a brand I like, I pay attention. The difference isn’t the message. It’s the source The referral mechanic worked because it restored that layer of trust. The lead didn’t come from an algorithm, it came from a trusted person Engineering intent into the sign up process follows the same logic as real time data verification They both exist to protect systems from bad data You’re not just collecting data. You’re collecting proof that someone is real, reachable, and genuinely interested Maybe the future isn’t about how AI helps us scale data collection It's about the constraints we put in place Verified data People with intent