This week I watched two of the most operationally sophisticated companies on the planet fail at the one thing that’s about to matter more than anything else. Not their product. Not their tech. Their service design.
And if Amazon and FedEx can’t get this right, your company is in more trouble than you think.
What Just Happened to B2B Software
Let me set the table first, because the urgency here is real and most CEOs are reading the wrong story about it.
Since January, B2B software stocks have lost roughly $2 trillion in market cap. The iShares software ETF (IGV) is down more than 21% year to date. Salesforce, Atlassian, Adobe, Snowflake: all of them have taken double-digit hits. (Bloomberg, SaaStr)
The headline everywhere is “AI agents are killing SaaS.” That headline is wrong, and the wrong story leads CEOs to the wrong conclusions.
The real story is quieter and more dangerous. SaaS isn’t being killed. It’s being starved. Enterprise IT budgets aren’t growing fast enough to absorb the tidal wave of AI spending, so the AI money is coming directly out of seat counts, expansion deals, and “let’s add another module” conversations. (SaaStr) When AI agents do the work of multiple humans, you need fewer humans. When you need fewer humans, you need fewer seats. The seat-based model that built Salesforce, Workday, and Atlassian is the one being eroded from underneath.
Here’s the part nobody is saying out loud. Once those seats are gone, those products go too. Not in a quarter. Not in a year. But once a CIO has stopped renewing, stopped expanding, and stopped paying attention, the relationship is already over. They just haven’t filed the paperwork yet.
So if your competitive position used to rest on “we have the best features,” you have a problem. Because features are now the easy part. AI can replicate most of what your product does in a weekend.
What it can’t replicate is how your customer experiences you.
The Question Most CEOs Can’t Answer
Here’s a test. Stop reading for ten seconds and try to describe, in detail, how a customer experiences your company from the moment they hear your name to the moment they renew or churn. Not your product. The whole experience. Every touchpoint. Every handoff. Every place something can go wrong and how you’d know if it did.
Most CEOs can describe their product features in their sleep. Ask them this question and the room goes quiet.
That silence is the most expensive thing on your balance sheet right now.
Historically, companies have treated service as the thing that happens after the product. Service was a cost center. Service was the call center. Service was whoever answered the phone when something broke. The product was the asset, and service was the overhead you tolerated to protect it.
That worldview is about to get a lot of CEOs fired.
Because in a market where any competitor can ship your features in weeks, the only thing distinguishing you is the entire experience around the product. And almost nobody is designing that experience as a system.
What Service Design Actually Is
Let’s clear something up, because most CEOs have never used the term and the ones who have usually mean something narrower than it actually is.
Service design is not customer support. It’s not the help desk. It’s not the satisfaction survey or the NPS score or the friendly tone in the chatbot.
Service design is the deliberate, end-to-end choreography of every way a customer engages with your company. It includes the product, yes, but it also includes how they find you, how they sign up, how they get value the first time, what happens when something breaks, how they get help, how they expand, and how they leave. It includes every handoff between every system and every team. It includes the moments you don’t think about because they happen to one customer in a thousand, until that customer is the one writing the LinkedIn post that costs you a deal.
Service design treats all of that as a single coherent system that has to be built, governed, and maintained on purpose. Not as the byproduct of whatever your support team and your product team and your billing team happened to ship this quarter.
Two real examples from this week make the difference impossible to miss.
Amazon: When Logistics Knows the Truth and the Service System Ignores It
Amazon is a logistics company. That’s the whole game. They know where every package is at every moment. They know when a box is damaged at a facility. They know which truck it’s on. They know when it should have arrived and didn’t.
I have a Subscribe and Save order I get every month. A few weeks ago, the box went into limbo at an Amazon warehouse a few miles from my home. Their own tracking showed it sitting there. For five days, I couldn’t talk to a human, couldn’t open a service ticket, couldn’t do anything except watch the dashboard and wait. Amazon’s policy: wait 5–7 days “in case the box shows up.”
It wasn’t going to show up. Amazon knew it wasn’t going to show up. Their data was screaming it.
Eventually I got the refund. But here’s the part that exposes the design failure: I had to repurchase every item at full list price. The Subscribe and Save discount, the whole reason I’m a recurring customer, didn’t apply to the replacement order. So Amazon’s “fix” punished the loyal customer for a failure inside Amazon’s own warehouse.
Amazon has world-class processes for damaged-on-arrival, missing-from-the-box, and wrong-item-shipped. Walk-in returns are everywhere. Refunds are nearly instant. They’ve designed the post-delivery service experience beautifully.
But the in-transit-failure experience? It doesn’t exist as a designed service. It exists as a policy. And the policy is “wait.” That’s not service design. That’s the absence of it.
The damning part is that Amazon has every piece of data needed to do this well. They could trigger a proactive refund the moment a box stalls past its expected dwell time. They could honor the original pricing automatically. They could send a one-line email that says “we see what happened, we already fixed it.” None of that requires new technology. It requires deciding that this slice of the customer journey is worth designing.
They haven’t decided that. Which means somewhere there’s a competitor who will.
FedEx: When Every Step Was Built by a Different Team
I needed to ship a package across the country. I’m a person, not a business. Here’s what the journey actually looked like.
I went to the FedEx website to figure out my options. The site assumed I already knew what I wanted. No simple intake. No “tell us the size, weight, and whether you have a box.” I poked around for a while, eventually figured out a price, and drove to the FedEx shipping store.
Walking in, I couldn’t tell where to go. The store was organized for people who already knew the system. After some wandering, I found the self-serve area for my type of shipment, found the right box, and packed my item.
Then I tried to buy the shipping label at the kiosk. You can’t. The self-serve area doesn’t sell labels.
So I bought the label on my phone, on FedEx’s website, while standing in their store. The website gave me a printable PDF. The kiosk told me to scan a QR code to print my label. My order didn’t have a QR code.
I ended up in line at the counter. Waited. Asked the staff to print the label. They had no way to receive it from me except by email. I emailed the PDF to a FedEx employee, standing five feet from her, so she could print a label for the package I had already paid to ship.
This is FedEx. FedEx. The company whose entire brand is precision logistics.
Every single step in that journey was built by a competent team. The website team built the website. The kiosk team built the kiosk. The retail team built the store layout. The label team built the label system. Each one probably hit their internal metrics. None of them designed the customer’s journey across all of those systems, because nobody owns the journey.
That’s the failure mode. Service design isn’t missing because the people are bad. It’s missing because the system was never designed as a system. It was assembled from departments.
Why This Is Suddenly an Existential Issue
For most of the last twenty years, you could get away with this. Customers had high switching costs, integration was painful, contracts were long, and your product features were defensible enough to paper over the friction in everything else.
That world is gone.
When AI can replicate your features, your product is no longer the moat. When budgets are getting harvested for AI, your seat-based revenue is no longer the moat. When customers can swap vendors faster than ever, your contract is no longer the moat.
What’s left?
The total experience of doing business with you. The way you anticipate failure modes before they reach the customer. The way you make the hard moments feel easy. The way every team across your company shows up as one company instead of seven uncoordinated departments. The thousand small decisions that make a customer think “I trust these people to take care of me” instead of “I’m one bad ticket away from leaving.”
That experience is not a feature. It can’t be ripped off in a weekend. It can’t be replicated by an AI agent because it isn’t a single thing, it’s the result of a system that took years of deliberate design to build. And it shows up in every interaction, every renewal conversation, every reference call.
That is your moat now. It might be the only one you have left.
What You Can Actually Do About This
This is where most articles wave their hands. I’m going to be specific.
The CEO sets the tone for how the business delivers on its mission. Service design starts with you deciding it’s a strategic priority, not a customer support project. Here’s what that looks like in practice.
Map the journey honestly. Pick one customer segment. Walk through every touchpoint from first contact to renewal. Don’t let your teams describe what’s supposed to happen, walk through what actually happens. Note every handoff, every system change, every place a customer has to repeat information, every place something can fail silently. Most CEOs are stunned by what this exercise surfaces. The map you build is almost never the journey you thought you had.
Find the failure modes nobody owns. In the Amazon story, in-transit limbo had no owner. In the FedEx story, the gap between the kiosk and the counter had no owner. Every company has these. They’re the gaps between team boundaries where customers fall through. Find yours. Assign owners.
Make service a designed system, not a department. This is the part most CEOs miss. Service design isn’t something the customer success team does. It’s something the whole company does, coordinated by someone whose job it is to see the system end to end. That role probably doesn’t exist in your org chart. That’s the problem.
Govern it like you govern the product. You have a product roadmap. You have a product review cadence. You have metrics that tell you whether the product is healthy. Do you have any of that for the service experience? If service is going to be your moat, it needs the same operating discipline as the product. Same rigor. Same attention from the leadership team. Same clear ownership.
Decide what “good” looks like, explicitly. Every team in your company has an opinion about what good service is. Most of those opinions contradict each other, and none of them are written down. When good is implicit, every team optimizes for their own version of it, and the customer experiences the contradictions. When good is explicit, every team can work toward the same outcome without being micromanaged. That’s where speed comes from.
Speed through clarity. Every role knows what good looks like. Every team designs toward the same experience. That’s not a customer service initiative. That’s a competitive moat.
The CEOs Who Move First Win This
In a year, the gap between the companies that took this seriously and the ones that didn’t is going to be visible from space. The ones who treated service as overhead are going to find their customers quietly switching to competitors who designed the experience on purpose. The ones who made service a designed system are going to look like they got lucky.
They won’t have gotten lucky. They’ll have done the work that AI can’t do for them.
Stop investing in features. Start investing in how you deliver. The next decade of B2B winners will be sorted by which CEOs heard that and acted, and which ones kept building products inside companies that couldn’t deliver them properly.
If you’re starting to look at your own company through this lens and you’re not sure where to begin, I want to hear from you. I’ve built an intelligent service design methodology specifically for B2B CEOs who need to turn this from an idea into a fully functional operating system. It lives right within your AI, forming and shaping how your team works right when your customers need it the most.
→ I’m working directly with a small group of leaders right now to shape it. Interested in trying it out? Send me a message. Tell me what you’re seeing in your own customer journey.
Your competitors are quietly building this moat right now. The question is whether you start before they finish.
