At Kustomer, we ran seven positioning statements through the same five questions and killed four of them, including two I personally liked.
Positioning work usually gets treated as a naming exercise. Pick the smartest tagline, ship it, move on. That's how Kustomer ended up cycling through "proactive, not reactive," then "hello, not help," then "native AI CX platform," then "all-in-one solution," with none of it sticking. What was on the website didn't match the decks. What was in the decks didn't match what reps said on calls. What one rep said to a prospect didn't match what another rep said to the next one. The brand was diluting itself in real time, and nobody had done anything wrong on purpose. Every one of those messages had sounded right in the room where it got picked.
So Product Marketing built a framework instead: five questions, applied to every candidate message with the same rigor, regardless of how much anyone in the room liked it.
The five questions:
- Can our product truly support this positioning?
- Is it truly differentiated versus our key competitors?
- Will it resonate with enterprises?
- What does CX leadership think?
- What does Product Marketing think, after all of the above?
Seven messages went through that gauntlet: proactive-not-reactive, customers-not-tickets, X% cheaper than Zendesk, all-in-one platform, CX impact on revenue, native AI CX platform, and AI Agent differentiation. Here's what happened to four of them.
All-in-one platform died on the first question
This was the easiest kill and the most uncomfortable one, because the company had used this exact positioning before, during the Nova launch, and it wasn't wrong then. Kustomer was a CRM first, with native Voice and AI that wasn't bolted on. All-in-one was true.
It stopped being true. When I pressure-tested it against what the product actually did, the answer from my product counterpart was specific: no Workforce Management, no QA, weak back-office task coordination (you couldn't even build a checklist inside a Task and report on it), and social support that hadn't kept pace with Meta's API changes since the platform divestiture. Zendesk had WFM and QA. We didn't.
Another product lead put the real risk into words during the exercise: this positioning is "almost impossible to support because it's going to mean something different to every client/prospect." The moment we didn't include a feature or integration a prospect expected, the positioning statement itself became the liability. Every competitor in the category, Zendesk, Intercom, Freshdesk, Salesforce, was already claiming all-in-one. The phrase had become table stakes. Claiming it without being able to fully back it meant handing competitors an easy way to make us look like we were overselling.
I didn't want to kill this one. It was clean, it was simple, and it had worked before. It failed the first question, not the differentiation question, not the resonance question. The product itself couldn't hold it.
Native AI CX platform died on the differentiation question
This one passed the product question easily. Our AI Agents lived inside the same system as the customer data, which meant building a proactive agent didn't require an external API call the way it would on a bolt-on AI layer. My product counterpart liked this a lot, calling it "aspirational but doable" and something to build toward as a north star. Another colleague went further: nobody else was positioned to claim this, and Intercom was the closest competitor but still wasn't this.
Then someone on our CX leadership side said the sentence that ended it: "Consumers don't care. It doesn't matter that we're native AI. So what? Why would it matter to me?"
And there was a harder problem underneath the opinion. The claim used to be literally true, when AI was bundled into the core Kustomer platform price. It stopped being true the moment we moved AI to an add-on pricing model. We could only make this claim for legacy customers still on the old Nova pricing. For every new deal, we'd be selling something we couldn't deliver at the price point the positioning implied.
A message can be aspirational and still get killed if the thing making it aspirational, in this case being AI-native by default rather than by add-on, stopped being operationally true. The product could theoretically support the vision. The pricing model couldn't support the claim, right now, for the customers we needed to close.
X% cheaper than Zendesk never got past the first question, for a different reason
Our Deal Desk could execute this. Product couldn't own it. Help Scout, Freshdesk, and Zoho Desk were already running this exact play, at $20, $15, and $14 per agent per month against Zendesk's $49, and none of them were companies we considered real competitors. That told us something about who else plays this card.
The resonance question made the decision easy. Enterprises prioritize value over cost. Leading with price signals a budget alternative, not a strategic partner, and it risks a discounting war that commoditizes the product in the exact moment we're trying to raise our average deal size. This was a tactical weapon for Deal Desk to use inside a live competitive deal, not a message to put on a billboard.
Two messages survived, and one came back from the dead
Customers-not-tickets survived every question. My product counterpart called it "the most true to what we are," built around Timeline, the feature designed since day one to give agents a complete customer view instead of a stream of disconnected cases. It hit Zendesk's actual weakness, its ticket-centric model, without naming Zendesk. Gladly had used similar language but not consistently or dominantly, which left room to own it. The one open question, raised directly by CX leadership, was whether the message was really about agent experience dressed up as customer experience, and whether an economic buyer would pay for that distinction.
AI Agent differentiation is the interesting one, because we'd already deprioritized it after Nova. It came back because the market moved. Zendesk, Intercom, and Gorgias were all leading with a single generalist AI agent. Kustomer's AI Agent Studio let a company build a specialized team of agents, one for returns, one for order processing, agents that could hand off tasks and verify each other's answers to cut hallucinations. Nobody else in the competitive set was making that argument. We'd shelved a real differentiator because we didn't think the market was ready to hear it, and the market caught up to the argument before we brought it back out.
What the framework actually protects against
The five questions don't produce a single right answer. They produce a forcing function against the thing that killed our positioning in the first place: picking a message because it sounded good in a room, then discovering six months later that the website, the decks, and the sales floor were all saying something different.
Every one of the four killed messages had at least one champion in that room who wanted to ship it. All-in-one had history on its side. Native AI CX platform had an aspirational pull that made people want to build toward it. Cheaper-than-Zendesk had the seductive clarity of a number. What killed each of them wasn't a lack of enthusiasm. It was running the same five questions against all seven candidates instead of just the ones that felt exciting, and being willing to let the answers overrule the room.
Where we landed
Treat every customer like they're your only customer.
This concept serves as the umbrella theme, with supporting points reinforcing what that level of service actually looks like in practice. What does it mean to treat every customer like your only customer?
- Customer 360 → Native CRM: A complete view of each customer.
- Anticipate needs before they arise → Proactive Service with Workflows.
- Give accurate, consistent answers → AI Agents that don't hallucinate, powered by AI Agent Studio.
- Show empathy at scale → Brand Tone & Voice controls, AIR, and sentiment analysis (detect when a customer is angry).
- Respond faster, resolve smarter → Lower FRT and AHT with intelligent routing.
- Meet them where they are → Omnichannel support.
- See people, not tickets → Conversation Timeline gives agents full context.