We had six months to create a product category that didn't exist, land the Gartner MQ leader spot, and exceed pipeline targets. Here's the honest account of how we did it — and what we got wrong first.
Kuber Sharma led the GTM strategy for Salesforce Data Cloud Zero Copy — a product category that didn't exist before they named it. The essay covers the three things that make category creation succeed: a technical truth simple enough to become a movement ("your data doesn't need to move to be used"), analyst relations as architecture not afterthought, and internal alignment as the hardest GTM problem of all. The Zero Copy Partner Network, launched April 2024, validated the category with IBM, Google, Snowflake, and Databricks. The lesson: a category wins when competitors start using your language against you.
Most product categories don't need to be created. You find an existing category, decide where you want to sit in it, and execute. Category creation — real category creation, where you're asking buyers to think about a problem they haven't named yet — is rarer and harder and riskier than almost any other GTM challenge.
In early 2022, we were facing it at Salesforce. We had a product — Salesforce Data Cloud — that did something genuinely new: it let enterprises use their data across systems without having to copy it. No ETL pipelines. No data duplication. No governance nightmare of keeping multiple copies of customer records in sync. Your data stayed where it was, and Salesforce could act on it directly.
The problem was that "no copying" was a technical property, not a value proposition. Buyers don't buy technical properties. They buy outcomes. And we hadn't yet found the language that connected this architectural choice to the business problems these buyers were actually losing sleep over.
What followed was six months of testing, failing, realigning, and eventually winning. Here's what I learned.
Our initial positioning centered on the technology. "Zero copy architecture for enterprise data." Clean, accurate, defensible. And almost completely ineffective in sales conversations.
The problem was that "architecture" is a decision data engineers care about and business decision-makers don't. The VP of Sales doesn't wake up thinking about ETL pipelines. The CMO doesn't lose sleep over data warehouse synchronization costs. They wake up thinking about why their customer data is fragmented, why their agents have incomplete information when a customer calls, why their AI initiatives keep stalling because the data isn't where the model needs it to be.
The architectural truth — zero copy — was the right answer to a question nobody was asking out loud. Our job was to surface the question.
The turning point came from a customer conversation — as it usually does. A financial services CIO described their data problem to us in a way we hadn't heard before. He said: "We have all the data we need. We just can't use it in the same place at the same time without creating a compliance nightmare."
That sentence contained the value proposition we'd been looking for. The problem wasn't that enterprises lacked data. It was that data's usefulness was locked behind the cost and risk of moving it. Zero copy didn't just eliminate a technical step. It eliminated the reason why valuable data stayed siloed in the first place.
We rewrote the positioning around data activation — the idea that your existing data investments were underperforming not because the data was bad, but because the infrastructure required to use it was too expensive and too slow. Zero copy was the unlock. Not the technology as such, but what it made possible: unified customer profiles, AI models trained on complete data, agents with full context.
Pipeline response changed immediately. The same data engineers who had nodded politely at "zero copy architecture" started calling their business leaders into follow-up meetings.
One of the mistakes I see PMMs make consistently with new categories is treating analyst relations as a communications function — something you do to get coverage, to influence the Magic Quadrant placement, to have quotes for press releases. That's not wrong, but it's incomplete.
The Gartner Magic Quadrant for data integration is a document that CDOs and CIOs use to make vendor shortlists. If your category doesn't map cleanly onto a Gartner framework, you're invisible to a significant portion of your addressable market. That's not a PR problem. That's a positioning architecture problem.
Getting Salesforce Data Cloud to a Gartner MQ Leader position required us to do something harder than brief analysts on our features: we had to make a credible case that the category itself was real and was emerging. That meant showing analyst firms evidence of market adoption, customer case studies, partner ecosystems, and competitive behavior that validated our category framing.
The moment you know a category is real is when your competitors start using your language. When we saw major data cloud competitors beginning to use "zero copy" to describe their own capabilities — sometimes accurately, sometimes not — we knew we'd won the naming battle. A category name that competitors adopt is a category name that buyers will search for.
Every category creation story I've read focuses on the external work: the messaging, the analyst relations, the customer evidence, the partner ecosystem. Almost none of them talk about the hardest part, which is getting your own organization to believe in the category before the market does.
At the beginning of the Zero Copy initiative, we had skeptics inside Salesforce. Not malicious ones — smart people who had seen enterprise software narratives come and go and who asked reasonable questions. Is the market ready for this? Are we creating a category that's too early? Are we differentiating on something that competitors can copy in 18 months?
These are good questions. The honest answers are: maybe, possibly, and yes respectively. Category creation is always a bet. You're betting that the market is moving in your direction, that you can move faster than your competitors at defining what the category means, and that by the time competition arrives you'll have enough of a moat — in customer relationships, partner depth, ecosystem adoption — that the category belongs to you even if the technology gets commoditized.
We won that internal argument not by having better answers to the skeptical questions, but by moving faster than the questions could accumulate. The Zero Copy Partner Network — launched April 2024, featuring IBM, Google Cloud, Snowflake, Databricks, and others — was the moment that answered most internal skepticism. When you can point to IBM and Google using your category language and your architectural approach in their own customer conversations, it's hard to argue that the category isn't real.
We found the customer language too late. The CIO conversation that unlocked our positioning happened in month four. With better win/loss discipline and more structured customer discovery in months one and two, we might have found that framing earlier and accelerated the entire timeline.
We also underestimated how much field enablement mattered. Our sellers are smart, but "zero copy" is not an intuitive concept to explain in a thirty-minute discovery call. We built excellent technical content. We were slower to build the two-minute conversational version — the one a sales rep could use to open a door with a business buyer who'd never heard of data federation. That gap cost us pipeline in the middle quarters before we closed it.
The last thing: category creation requires patience from leadership that most GTM organizations aren't structured to provide. The metrics that matter for a new category — category share of voice, analyst positioning, partner ecosystem depth, competitive language adoption — are slower to move and harder to attribute than pipeline and revenue. If you're measuring yourself only on quarterly numbers, you'll kill the category investment before it has time to work.
"A category wins when your competitors start using your language. That's not imitation. That's confirmation."
Zero Copy is now a standard term in the enterprise data industry. Databricks uses it. Snowflake uses it. AWS uses it. That didn't happen because we marketed harder than them. It happened because we named something real before they did, moved fast to establish the definition, and built a partner ecosystem that gave the category commercial proof.
That's how categories work. You don't win them by being right. You win them by being first to be useful.
Related reading: The full body of Zero Copy thought leadership — including bylined articles on Salesforce.com, the Zero Copy Partner Network press release, and webinars with Google Cloud, IBM, and Snowflake — is collected on my Media page.
Kuber Sharma led the GTM strategy and market narrative for Salesforce Data Cloud Zero Copy from its inception through the Zero Copy Partner Network launch. He now leads platform product marketing at UiPath. He writes Positioned, a newsletter on AI-era product marketing strategy for enterprise PMMs.