Ciente > Articles by: Ciente Editorial Team

AMD’s Helios AI Rack Challenges NVIDIA’s Monopoly with Open Infrastructure

Author: Ciente Editorial Team

AMD’s
News

AMD’s Helios AI Rack Challenges NVIDIA’s Monopoly with Open Infrastructure

AMD’s Helios AI Rack Challenges NVIDIA’s Monopoly with Open Infrastructure

AMD unveiled Helios, a 72-GPU rack-scale AI system built to rival NVIDIA’s top infrastructure. Here is why open standards and real hardware competition benefit the AI industry.

NVIDIA has dominated AI hardware for years. AMD just launched a direct challenge. At its Advancing AI event, CEO Lisa Su unveiled Helios- a fully integrated rack-scale system built to train and run giant AI models.

Helios packs 72 new Instinct MI455X GPUs, 6th-Gen Epyc CPUs, and high-speed Pensando networking into a single liquid-cooled frame. AMD claims the system delivers up to 30% more tokens per dollar than NVIDIA’s competing Vera Rubin NVL72 platform.

More importantly, major players like Meta, OpenAI, Microsoft, and Anthropic have already committed to test and deploy Helios at gigawatt scale.

Building a fast chip is no longer enough. Modern AI labs need entire data centers operating as unified supercomputers. NVIDIA mastered this end-to-end integration early. AMD is now matching that hardware scale while taking a distinct strategic path: open standards.

Helios relies on open frameworks like UALink and Ultra Ethernet rather than locking customers into proprietary networks. AMD is also aggressive on software. Anthropic even agreed to use its Claude model to help optimize AMD’s open ROCm platform.

This open strategy provides vital market nuance:

  1. Dual-sourcing hardware reduces single-vendor supply risks and cuts training costs for AI labs.
  2. Open standards prevent ecosystem lock-in to drive faster innovation across hardware tiers for developers.

AMD still faces a steep climb.

NVIDIA’s CUDA software ecosystem remains deeply entrenched among developers- translating new hardware into raw developer adoption takes time.

But Helios has proved that genuine competition has finally arrived in top-tier AI infrastructure. And that shift creates a more sustainable market for everyone building the future of AI.

Brand recall
Blogs

Why Brand Recall is a Structural Hiccup Most Brands Haven’t Tried to Fix

Why Brand Recall is a Structural Hiccup Most Brands Haven’t Tried to Fix

Over 83% of consumers are unable to recollect your brand within 24 hours of encountering it. Brand recall might be a bigger problem than marketers assume it to be.

A mere 17% of consumers can name a brand from an ad they came across just the previous day. Let the stat sink in for a sec.

We’re not talking about a campaign from a month ago- but from yesterday. Something your creative team spent weeks planning and executing. Where the budget ran into 6 figures. And 83% of your consumers simply forget you- within 24 hours.

What if this time it’s not the messaging, but a brand positioning gap, a more nuanced structural hiccup that affects brands irrespective of size and category.

It’s brand recall.

Marketers used to attribute brand recall to a positive feeling- the high numbers across brand health reports got nodded at during quarterly reviews. It didn’t entail any sense of urgency.

But something has shifted in marketing.

In an age where customers are comfortable researching independently, and AI mediates discovery, it’s easy for your audience to find your competitors first. And actually remember them at the time of need.

This makes brand recall imperative to the bottom line, and not a vanity metric.

What is Brand Recall, What It Measures and Why Marketers Conflate It with Another Vanity Metric

While brand recall isn’t the same as brand recognition, marketers have had a history of using the terms interchangeably. That’s the biggest mistake they can make.

Brand recognition is passive. The consumer sees your logo, color palette, jingle, and knows they’ve seen it before. That’s recall with training wheels. It depends on a prompt or cue, i.e., something external to activate the memory.

But brand recall is different.

Unaided recall is when a consumer thinks of coffee, and your brand appears without an external push. No logo in sight. No ad running. Just the mental shortcut that takes them from category to brand name in a fraction of a second.That’s the version worth building for.

Aided vs. Unaided Brand Recall: Why the Distinction Actually Matters to Your Strategy

There are two types of brand recall that show up in measurement, and both tell different stories.

Aided brand recall measures whether a consumer recognizes your brand when prompted, usually with a list of options. Helpful for tracking brand awareness and familiarity. Not very useful for predicting whether a consumer reaches for you first.

Unaided brand recall is the one that drives purchase behavior. When a buyer faces a category decision with no shortlist in front of them, they go with whoever comes to mind first.

That’s the moment unaided recall wins or loses revenue. And it’s the one most brand strategies underinvest in, because building it is slower and harder than running recognition campaigns.

Nielsen’s data makes the relationship between the two clear.

Podcast advertising reaches an aided recall rate of 71% among exposed audiences, compared to 50% for unexposed ones. Influencer marketing hits 79% aided recall. Branded content: 81%. These aren’t awareness numbers. These are signals of how deeply a channel embeds brand information.

The bottom line? The channel matters as much as the creative.

Why Brand Recall Is So Hard to Build Right Now

The Four-Touch Problem No Media Plan Is Solving For

Adobe’s consumer research found something that should rearrange every media planning conversation: consumers say they need to encounter a brand message at least four times within a 24-hour window for it to feel memorable.

Four times. In a single day.

Now think about how most campaigns actually work.

A consumer sees a pre-roll ad on YouTube Tuesday morning. Maybe a sponsored post Wednesday. A retargeting display ad sometime next week. The frequency is scattered across days, sometimes weeks, with zero coordination between formats.

By the time the second touch lands, the first one is long gone.

Brands have a simple solution to this. Not hammering the same audience with the same creative until they mute it.

But by concentrating exposure in a single place. That means coordinating channel timing so that a consumer moves through multiple touchpoints in a compressed window. Short-form video, display, social, and email hitting the same audience cluster within 24 to 48 hours.

That’s what makes messages stick. Creative quality alone isn’t enough.

The Brand Recall Problem AI Just Made Worse

Here’s the part most brand teams haven’t caught up with yet.

When a consumer forgets a brand name and wants to find it again, 31% now turn to an AI chatbot. That sounds like opportunity. The uncomfortable truth follows immediately: only half of those users get the correct brand name at least 50% of the time.

AI retrieves brands it can confidently surface from authoritative, well-structured, frequently cited sources. Brands that don’t show up in those sources don’t exist in AI-mediated discovery.

This creates a second recall gap that has nothing to do with creative. A consumer saw the campaign. They remembered they were interested. They asked an AI to help them find the brand again. The AI either surfaced the wrong brand or drew a blank. The sale went somewhere else.

Brand recall in 2026 requires two things working simultaneously: a memory that survives in the consumer’s mind, and a presence that survives in AI’s training data and real-time citations. Marketers who nail the first while ignoring the second are solving half the problem.

What Actually Drives Brand Recall (And What Just Eats Budget)

What Nielsen’s Brand Lift Data Says About Emerging Media Brand Recall

Nielsen analyzed brand lift across podcasts, influencer marketing, and branded content. Across all three channels, brand recall outranked every other driver of lift. Not creative quality. Not message relevance. Not share of voice.

Brand recall was the single biggest predictor of whether a campaign moved the needle, accounting for 38.7% of brand lift in emerging media. Baseline awareness came second at 37.5%.

This finding changes the framing of how to evaluate channel mix.

A channel isn’t just worth its audience size. It’s worth what it does to brand recall. Podcasts, influencer content, and branded partnerships go beyond traditional display and programmatic. They embed brand information within trusted, contextual content that audiences actively choose to consume.

That’s a different kind of attention than an ad interrupting a feed.

The Platform and Format Question for Brand Recall

Adobe’s data shows YouTube (46%) and cable/streaming TV (40%) drive the longest-lasting brand impressions. Instagram and TikTok follow. Podcasts and X are at the bottom.

Format matters just as much.

Short-form video sits at 52% for memorability. Static images at 30%. Consumers are 73% more likely to remember short-form video than static image ads. That gap widens significantly for Gen Z, who are also 111% more likely than older groups to recall influencer endorsements, and 94% more likely to recall interactive ads.

The other finding worth paying attention to: multitasking kills brand recall.

51% of consumers say it actively hurts their ability to remember brand names. Mobile, despite being where so much advertising spend goes, is where recall is weakest. Consumers are 129% more likely to remember product details from desktop than mobile.

Brands that are over-invested in mobile display often do so without accounting for the attention environment those placements actually sit in.

How Brands Actually Build Brand Recall That Lasts

Emotion Is the Shortcut Brand Recall Relies On

Humor outperforms every other factor in making ads stick. 56% of consumers cite it as the primary reason an ad left a lasting impression. Entertainment follows at 43%. Problem-solving comes in at 38%.

The brands people remember are the ones that made them feel something.

Nike’s “Dream Crazy” campaign didn’t just sell shoes. It put Colin Kaepernick front and center and asked audiences to take a side. That’s a brand that understood emotional resonance doesn’t come from playing it safe. It comes from saying something that lands with conviction.

Heinz ran a guerrilla campaign in 2021 where they asked people across five continents to draw ketchup. The majority drew a Heinz bottle. Unprompted. Including the logo and name. That’s unaided brand recall in its purest form. Built over decades of consistent identity, smart advertising, and one campaign that turned it into a story the internet spread for free.

These campaigns succeed because the brand treated brand storytelling and emotional investment as a strategic priority, not a creative afterthought.

Brand Recall Requires Consistency as an Infrastructure.

Most brands treat consistency as a brand guideline document. Same colors, fonts, tone of voice. That’s the basic building block.

The brands with the highest recall scores prioritize consistency.

Every platform, format, campaign reinforces the same emotional territory. The Coca-Cola red isn’t just a color. It’s a conditioned response built over a century- appearing in the same context so reliably that the color alone triggers a brand’s memory.

Spotify Wrapped does this brilliantly every year. And this same concept has been executed with fresh data, wrapped in the same visual identity, and timed to the same moment in the year- every year. Hundreds of millions of people share it voluntarily.

The campaign exists because Spotify made personalization the consistent emotional core of the brand experience. Wrapped is just the annual proof point.

Consistency signifies the consumer knows what to expect from you, regardless of the channel they encounter you on.

When Brand Recall Fails: Winning Back Customers Who Forgot You

When recall breaks down, the re-engagement strategy usually determines whether the customer comes back or finds a competitor.

Adobe’s research shows incentives drive re-engagement more than content.

Price drop alerts work for 54% of consumers. Specific discount codes for 52%. Free shipping for 41%. These are transactional triggers, not brand-building ones. But they serve a specific purpose: they give the customer a reason to act on the memory before it fades entirely.

The more interesting finding is what happens when the recovery is purely passive.

48% of consumers who forget a brand think they might find it again through another ad. Half the audience is willing, yet do nothing because no one reached out to them with a relevant reason.

The mistake brands make in re-engagement is treating it like an awareness campaign.

Broad messaging to a cold audience doesn’t recover a forgotten relationship. Specific, relevant, timely outreach to someone who already expressed interest does. The audience already knows you, but they need a nudge that respects that context.

Brand Recall Makes Every Other Marketing Investment Pay Off

Every click, every open, every demo request moves faster when brand recall is strong. The consumer already knows the name. They’ve already formed an impression. The rest of the funnel moves quicker because the top of it did its job.

That’s the actual ROI case for brand recall investment. Not the recall score itself. The multiplier effect it has on everything downstream.

The brands building this well are ones who treat recall as a business problem. They coordinate frequency intelligentlyThey invest in channels where attention quality is high. They make emotional resonance a strategy. They measure what the data actually tells them, and they stay consistent long enough for the memory to form.

Brand recall isn’t built in a campaign. It’s built in the space between them.

Customer Experience in Financial Services Has an AI Problem Just Not the One Youd Expect
Blogs

Customer Experience in Financial Services Has an AI Problem, Just Not the One You’d Expect

Customer Experience in Financial Services Has an AI Problem, Just Not the One You’d Expect

Financial services brands spent years running AI pilots. The ones winning on customer experience aren’t piloting anymore. They’re deploying at scale.

Financial services has never had a shortage of customer data. Banks know when you get paid. Insurers know your risk profile down to your postcode. Wealth managers know your portfolio, your goals, your risk appetite, sometimes your tax situation. No other industry sits on this much behavioral and financial intelligence about the people it serves.

And yet, for most customers, interacting with a financial services brand still feels like calling a number and waiting.

That gap, between the data banks hold and the experience they actually deliver, is the central CX failure of the industry. It isn’t a technology problem. It isn’t even primarily an AI problem. It’s a willingness problem. The infrastructure to deliver genuinely personal, fast, and relevant experiences has existed for years. Most institutions just haven’t built around it.

That’s starting to change. But slowly, and unevenly, and with a lot of “pilot” language that masks how far behind most brands still are.

What Customer Experience in Financial Services Actually Looks Like in 2026

Depends entirely on who you’re banking with.

The first scenario: A customer opens a neobank app => gets a real-time spending nudge that actually reflects their habits => resolves a dispute in three taps => gets a loan decision in under a minute.

The whole experience fits inside a phone screen and takes less time than making coffee.

The second scenario: A customer calls their high-street bank => waits eleven minutes => gets transferred twice => explains their problem three times to three different people.

The problem still isn’t resolved.

Both of those things happen in 2026. In the same market. Sometimes within the same institution, depending on which channel the customer chooses and which team picks up.

The divergence isn’t accidental. It reflects a structural reality.

The brands investing in customer experience in financial services as a genuine business priority, rather than a compliance checkbox or a PR talking point, have pulled far enough ahead that catching up now requires more than a technology refresh.

Why Customer Experience in Financial Services Keeps Falling Short

The Trust Problem That Makes Everything Harder

Financial services CX operates under a constraint most other industries don’t face.

Customers share more sensitive information with their bank than with almost anyone else in their lives. That creates a specific expectation. Not just that the experience is smooth, but that the data behind it gets used responsibly.

The moment personalization tips into surveillance territory, trust collapses. And in financial services, lost trust doesn’t just mean a bad review. It means switching, regulatory complaints, and in some cases serious reputational damage.

This is why the personalization problem in financial services is harder than it looks.

A retailer showing you ads based on your browsing history is annoying at worst. A bank referencing financial stress you didn’t explicitly share feels like a violation. The same data, different context, completely different emotional response.

The brands getting CX right in this space have figured out where that line sits. They use behavioral data to remove friction and surface relevant products at the right moment. They don’t weaponize it. That distinction isn’t just ethical. It’s commercial. Customers who trust their financial services brand spend more, churn less, and refer more.

Why Legacy Infrastructure Keeps Holding Customer Experience Back

Most large financial services institutions are running customer experience strategies on technology stacks that were never designed for them.

Core banking systems built in the 1980s and 1990s don’t talk to modern CX platforms natively.

Customer data sits in silos across retail, mortgage, insurance, and wealth management arms that were integrated on paper during a merger and never actually unified underneath. A customer who holds a current account, a mortgage, and a pension with the same institution might be treated as three separate people across three separate systems, because that’s how the data is structured.

The CX gap this creates is visible.

A mortgage advisor who doesn’t know the customer called customer service twice last week about a payment concern is having the wrong conversation. A renewal offer that doesn’t account for a life event the customer logged elsewhere in the same app is missing the point entirely.

Fixing this requires infrastructure investment that doesn’t appear in a single year’s P&L. That’s why most incumbents have deferred it. And why challengers who built clean from scratch have a structural CX advantage that’s harder to close than it appears from the outside.

How AI Is Reshaping Customer Experience in Financial Services

Agentic AI and What It Actually Changes for Financial Services CX

The pilot phase for AI in financial services CX is, for the leading brands, over.

The conversation has moved from “should we test this?” to “how do we scale what’s working?” And what’s working, increasingly, is agentic AI.

Not chatbots that answer FAQs. Agents that take action. Resolve a dispute without escalation. Adjust a payment date. Identify a fraud risk and pause a transaction before the customer notices anything wrong. Complete a claims process end to end without a human in the loop.

This changes customer experience in financial services in a specific way. It collapses the time between a customer feeling a problem and the problem getting resolved.

For most of financial services history, that gap was measured in days. Phone calls, callbacks, escalations, manual reviews. Agentic AI compresses it to seconds. And the emotional impact of a problem resolved before it becomes a problem is fundamentally different from the emotional impact of a problem eventually fixed after multiple contacts.

Barclays deployed AI to reduce customer wait times and surface proactive alerts.

NatWest’s Cora assistant now handles more complex queries than it did twelve months ago, because the underlying model has been trained on enough resolved interactions to handle edge cases that would previously have required human intervention. These aren’t experiments anymore. They’re operational infrastructure.

Personalization at Scale in Financial Services Customer Experience

Personalization in financial services used to mean putting a customer’s name in an email subject line. Understanding the psychology of personalization and how customers actually respond to relevance is why that bar has moved considerably.

The brands setting the standard now deliver personalization that reflects actual behavior, actual timing, and actual financial context.

A customer whose spending patterns suggest they’re approaching overdraft gets a nudge before it happens, not a fee after. A customer who just received a large deposit gets a relevant savings product surfaced within 24 hours, not three weeks later in a generic marketing email they ignore.

This requires three things working simultaneously.

A unified data layer that pulls customer behavior from every channel into one place, the kind of setup covered in how to use data analytics to improve customer experience. A model that knows which signals matter and which are noise. And a delivery mechanism fast enough to act on the signal while it’s still relevant.

Most large financial services brands have one or two of those. Very few have all three working together. The ones that do are the ones whose NPS scores look different from their competitors.

The Human Element Still Matters in Financial Services Customer Experience

AI resolves the fast, repeatable, high-volume interactions well. It handles the things that don’t require judgment.

But financial services is full of interactions that do require judgment.

A customer going through a divorce needs someone who can hear what’s unsaid. A small business owner facing a cash flow crisis needs more than an automated payment deferral; they need a conversation. A first-time buyer navigating a mortgage application isn’t just submitting documents; they’re making the biggest financial decision of their life, and they need to feel like someone is actually looking after them.

The mistake some brands make is treating AI as a headcount replacement rather than a quality upgrade for human interactions. The customer conversations that matter most, the ones that determine whether a customer stays for twenty years or leaves at the next renewal, still require human judgment, human empathy, and human accountability.

The right model isn’t AI instead of people. It’s AI handling everything it can handle well, so human advisors spend their time on the interactions where they’re genuinely irreplaceable, which is really why customer success is important to the bottom line. That’s not a cost-cutting framing. It’s a CX quality framing. And the distinction shows in outcomes

What Regulation Actually Does to Customer Experience in Financial Services

Nobody discusses this part enough.

Regulation shapes financial services CX in ways that have no equivalent in other industries. FCA consumer duty requirements in the UK, for instance, place a direct obligation on firms to deliver good outcomes for customers. That isn’t just a compliance requirement. It’s a CX mandate written into law.

This creates an interesting dynamic. Firms that treat consumer duty as a compliance exercise build systems that technically meet the standard. Firms that treat it as a CX framework build systems that actually improve outcomes. The first group spends money to avoid regulatory action. The second group spends money and gets better retention, lower complaints, and a defensible commercial case for the investment.

AI helps here in a specific way. It makes it possible to monitor outcomes at scale, the same discipline behind cx analytics and how to measure and improve the customer experience more broadly. To identify which customer segments are getting worse outcomes than others. To flag when a product is being sold to customers it isn’t well suited for. To catch problems before they become complaints, or complaints before they become regulatory incidents. That’s compliance infrastructure and CX infrastructure at the same time.

Where Customer Experience in Financial Services Has to Go Next

The brands that pull ahead on CX in financial services over the next three years won’t be the ones that launch the most impressive pilot. They’ll be the ones that make the unglamorous investments that pilots don’t require.

Unified data infrastructure. Real-time signal processing. AI models trained on actual customer interaction data rather than generic benchmarks. Human teams structured around the interactions that actually require judgment. And a genuine organizational commitment to treating customer experience as a revenue driver rather than a cost center.

That last part is harder than it sounds. Customer experience in financial services has spent decades being managed as a function that prevents bad things from happening. Complaints down. Wait times down. Escalations down. Those are defensive metrics.

The brands reframing CX as an offensive capability, something that drives acquisition, expansion, and retention, are starting to measure different things, often built on a stronger voice of the customer to understand what actually drives loyalty. Lifetime value. Advocacy rates. Product depth per customer. Revenue per interaction channel.

The shift in measurement tells you everything about the shift in intent. CX as damage control is a cost center. CX as growth infrastructure is a competitive weapon. The financial services brands figuring that out right now are the ones whose customer experience feels genuinely different from everyone else’s.

That gap will widen. Quickly.

Google
News

White House Accuses China’s Moonshot AI of Training on Banned NVIDIA Chips

White House Accuses China’s Moonshot AI of Training on Banned NVIDIA Chips

White House officials claim Chinese startup Moonshot AI used restricted NVIDIA GB300 chips routed through Thailand to train its powerful Kimi K3 model.

The White House is aiming at one of China’s hottest AI startups.

Top White House tech official Michael Kratsios publicly accused Beijing-based Moonshot AI of using banned NVIDIA hardware to train its flagship model, Kimi K3.

According to Kratsios, Moonshot acquired servers packed with high-end NVIDIA GB300 Blackwell chips. The company reportedly routed training workloads through infrastructure in Thailand to bypass strict US export controls. Kratsios also claimed Moonshot covertly distilled Anthropic’s Claude models to boost Kimi K3’s performance.

These allegations follow a stunning debut for Kimi K3 last week. The open-weight model shocked Silicon Valley by nearly matching top American systems like OpenAI’s GPT-5.6 and Anthropic’s Claude Fable 5 on major benchmark tests.

From a policy standpoint, Washington’s frustration makes total sense. US export rules explicitly prohibit Chinese companies from buying top-tier Blackwell chips anywhere in the world. Officials want to protect American intellectual property and maintain a decisive technological lead.

Yet, looking past the political heat reveals a fascinating tech reality.

Moonshot’s rapid breakthrough highlights the sheer momentum of global AI innovation. Strict trade rules may slow physical hardware access, but resourceful engineering teams consistently find clever software solutions to build world-class products.

Neither Moonshot nor NVIDIA has officially commented on the White House statements.

Amid all the concerns, the confrontation marks a pivotal moment. Geopolitical trade barriers are actively reshaping the tech industry, but brilliant developers keep pushing frontier AI forward regardless of borders.

Google
News

EU Slaps Google with $1 billion DMA Fine, Marking a Turning Point for Digital Competition

EU Slaps Google with $1 billion DMA Fine, Marking a Turning Point for Digital Competition

The EU fined Google €890 million over Search self-preferencing and Play Store steering rules. Why does this decision matter for consumers and developers?

European regulators just sent another massive signal to Silicon Valley. On Thursday, the European Commission hit Google’s parent company, Alphabet, with an €890 million ($1 billion) fine under the Digital Markets Act (DMA).

The EU split the penalty into two clear slices:

  1. Search Preference (€460 million): Google routinely gives its own shopping, flight, and hotel features prime visual real estate right at the top of search results.
  2. App Store Steering (€430 million): Google stops app developers from telling users about cheaper deals outside the Google Play Store.

From a regulatory perspective, you can totally see where Brussels is coming from. EU officials want a genuinely level playing field. They believe search engines should point you to the best options on the web, not just internal Google tools. They also want app creators to speak directly to consumers about better pricing options.

At the same time, Google makes a fair point about user experience. Built-in maps, flight trackers, and instant shopping widgets make Search fast, clean, and ridiculously useful. Stripping out those direct answer boxes forces everyone to click through multiple third-party links just to check a flight time or compare prices.

Google already plans to appeal, calling the decision a step backward for product design. Still, this landmark fine marks a fascinating moment in tech history. Regulators are actively reshaping digital platforms to boost consumer choice. Meanwhile, developers gain fresh freedom to market their services directly. Ultimately, this friction forces big tech to work even harder to earn user loyalty.

Scalling partner programs
Blogs

Challenges of scaling partner programs

Challenges of scaling partner programs

There’s a moment every growing company arrives at. Direct sales have plateaued. The CAC keeps climbing while the CLV refuses to budge. Headcount is expensive and slow. And then, in some leadership meeting, someone says the magic word.

Partners.

And why not? The numbers are seductive. By 2025, roughly 75% of global B2B transactions will flow through channel partners. Mature programs reportedly drive 2x revenue growth and up to 28% of total company revenue. The pitch writes itself: build an ecosystem, borrow other people’s trust, other people’s markets, other people’s sales teams. Scale without the payroll.

So you launch the program. Tiered framework. Deal registration portal. A shiny PRM. Onboarding modules. A partner manager or two.

And then, somewhere around partner number thirty, the whole thing starts to groan.

This piece is about that groan. Not the launch problem, which everyone talks about, but the scaling problem, which almost no one diagnoses correctly. Because scaling a partner program isn’t a bigger version of running one. It’s a different animal entirely.

The thing that doesn’t scale is the thing that made it work.

Here’s something that most channel content skips: partnerships run on trust, and trust does not scale linearly.

Your first five partners worked because someone, usually a founder or a head of partnerships, knew them. There were real conversations. Context was shared over calls and dinners and the occasional difficult negotiation. The relationship carried the program.

And when leadership sees those first five work, they do what organizations always do: they try to systematize it. Turn the relationship into a workflow. Turn judgment into a dashboard. Turn the head of partnerships into a “process.”

This is where scaling breaks, and it breaks quietly. 65% of partnerships fail, and 73% of marketers say managing partners is a major challenge. Those aren’t launch statistics. Those are scale statistics. Programs don’t usually die at zero partners. They die at scale, when the human infrastructure that made the first handful work can no longer stretch across the hundredth.

So the real question of scaling isn’t “how do we get more partners?” It’s “what do we do about the fact that the thing making this work cannot be copy-pasted?”

Let’s unpack where it actually breaks.

Why partner programs fail to scale.

1. The partner manager becomes the bottleneck.

Every partner wants attention. Early on, you can give it. One manager, ten partners, real relationships.

Now multiply. A hundred partners, fragmented data across spreadsheets and a CRM and a portal, manual deal registration, generic enablement, and slow quarterly review cycles. The partner manager who used to be a relationship becomes a queue. The good partners wait. The bad partners take up the oxygen. And the person you hired to grow the ecosystem spends their week doing deskwork and chasing registrations.

You didn’t scale the program. You scaled the admin.

2. You rebranded your sales training and called it enablement.

Here’s the trap almost everyone falls into. You take your internal sales playbook, slap a partner logo on it, and ship it.

But as Greg Portnoy puts it, partner enablement is not sales enablement. Your sales team works for you. Your partners don’t. Your reps have time for a two-week ramp. Your partners have their own quota, their own products, and roughly nine minutes of attention for yours.

Generic enablement assumes the partner cares as much as you do. They don’t, and they shouldn’t. They have a portfolio. You are one line in it. If your product isn’t the easiest thing on their desk to sell, it becomes the hardest thing to prioritize, and it quietly slides down the list.

Certification, done right, is worth it. Certified partners reportedly earn 6x more revenue than the ones who skip training. But that’s exactly the point that gets missed: enablement that works is built for the partner’s reality, not for your org chart.

3. Channel conflict is not a bug. It’s what scale creates.

With five partners, everyone has their lane. With a hundred, lanes overlap. Two partners chase the same account. A partner chases an account your own direct team is already working. Overlapping territories, unclear ownership, inconsistent incentives, no shared visibility – the exact things that don’t matter at small scale become existential at large scale.

And here’s the part leaders hate to hear: channel conflict isn’t a failure of the rules. It’s the predictable result of putting more self-interested actors in the same market and expecting them to defer to a portal. You cannot out-policy a zero-sum incentive. If two parties both get paid for the same deal, or worse, only one does, no deal-registration timestamp is going to make that feel fair.

4. The dashboard delusion.

This is the big one, and it deserves its own reckoning.

The industry’s answer to scale is software. PRM spend is projected to hit $45B in 2025. The promise is intoxicating: automate the workflows, remove the humans from the repetitive steps, let the dashboard tell you which partners are performing and which should be “reevaluated.”

And tools genuinely help. Nobody should run a hundred partners on spreadsheets and goodwill.

But here’s where the logic quietly betrays you. A dashboard measures what a partner did. It does not measure whether you can trust them. It cannot see the affiliate padding their numbers, the reseller controlling the terms, the SI who says the right things on the QBR and does nothing in the field. The metric tells you the partnership is healthy right up until the moment it isn’t.

We convince ourselves that because B2B is “rational and logical,” a good enough dashboard can replace judgment. It can’t. Partnerships are as messy and as human as anything in business. The software scales the workflow. It does not scale the trust. And when a program mistakes the first for the second, it grows fast and rots from the inside.

The principal-agent problem, now at scale.

If you’ve read anything I’ve written on partner marketing, you know the villain by name: the principal-agent problem. You (the principal) want an outcome. Your partner (the agent) acts on your behalf but has their own interests, and knows things you don’t.

At five partners, you can manage this with attention. You notice when incentives drift. You have the conversation.

At scale, the information asymmetry explodes. Every new partner is a new set of incentives you can’t fully see and can’t fully align. Some will be honest. Some will optimize for whatever your program rewards, even when that’s bad for you – over-registering deals, cherry-picking easy accounts, coasting on MDF. This isn’t cynicism. It’s structural. More agents means more asymmetry, and more asymmetry means more room for the gap between what’s good for them and what’s good for you to widen unnoticed.

This is why programs “built for optics, not outcomes” stall. They design tiers and portals and swag, then blame “partner performance” when revenue stalls, instead of the program architecture itself. The architecture was the problem. It scaled the incentive structure without ever solving the asymmetry underneath it.

So, what actually helps? (No easy answers.)

If you came here for a five-step framework, you already know I’m not going to give you one. Those advices expire quickly because they aren’t based on the first principles of partnership. Here’s what will actually shift your thinking.

Design for coopetition, not obedience. The healthiest partnerships aren’t built on the fiction of pure cooperation, where the partner exists to serve you. They’re built on a mutual push and pull – both parties growing a bigger market and competing for a piece of it. Think Samsung and Apple: rivals who still build each other’s components. A partner who has a real, independent reason to win alongside you doesn’t need to be policed. Scale that mindset instead of scaling surveillance.

Manage information asymmetry deliberately. At scale, the question is not “how much can we share?” but “what do we share, with whom, and when, without handing over leverage or creating duplication?” Which partners get deep product access? Which get deal-level data? Which get the arms-length version? Answering this on purpose, instead of defaulting everyone to the same tier, is the closest thing to a real scaling lever there is.

Segment by trust, not just by revenue. Tiers usually measure what a partner sold. The more useful axis is how much of your business you can safely place in their hands. Those are not the same number, and pretending they are is how programs get burned by their “top” partners.

Keep a human in the loop where trust lives. Automate registration, reporting, and content delivery – please do. But the judgment calls (who to invest in, who to quietly wind down, where a conflict is really coming from) are exactly the parts you cannot hand to a dashboard. Ruthless with the admin, human with the relationship. That’s the balance scaling demands.

Ask the question nobody asks. Before adding partner number one hundred, ask what they actually need to succeed – not what makes your program look impressive. Most programs are built as a monument to the company that built them. The ones that scale are built around the reality of the people selling for them.

The real challenge.

Scaling a partner program is not an operations problem you can buy your way out of, and it’s not a motivation problem you can incentive your way out of. It’s a trust problem wearing an operations costume.

The companies that scale partners well aren’t the ones with the best PRM or the most tiers. They’re the ones honest enough to admit that the relationship, and not the software, was always the product. Everything else is just infrastructure for a thing that was human the whole time.

The dashboard will tell you the numbers are going up.

Whether you can trust the ecosystem those numbers are built on – that part, you still have to earn.