Show notes
In this episode of Marketing Spark, Mark Evans sits down with Vahagn Sargsyan, VP of Marketing at Krisp, a pioneering voice AI company.
Vahagn shares his unique journey from full-stack developer to marketing leader and dives into the strategies behind Krisp's emergence as a category-defining brand. He explains how Krisp went from a simple noise-cancellation tool to a comprehensive AI-powered communication platform used by millions, including enterprise clients like call centers and Discord.
This conversation explores agile marketing, product-led growth, community-driven innovation, and how aligning closely with customers can fuel brand loyalty and expansion.
From the conversation
From full stack developer to marketing leader
Vahagn Sargsyan, VP of Marketing at Krisp, began his career as a full stack developer working with PHP, JavaScript and jQuery. After about three years he moved toward marketing, which he describes as a mix of psychology and systems thinking. Around 2011 he started as an SEO specialist, then moved into performance marketing and later into leadership roles.
Before Krisp he worked at Rocket Internet, his first exposure to the startup world. Vahagn found he enjoyed early stage environments where nothing is set up yet and everything has to be created from the ground up. He joined Krisp in 2019 as marketing director, when the company had no marketing engine, no revenue and no clear product category, but a clear vision.
Vahagn says his engineering background still helps in two ways. It gives him a more structured way of planning and managing projects, and it makes conversations with product managers and engineering heads easier.
"Because if you are working in a tech company as a marketer, you need to have a good enough understanding of how the product works, how the code works, how the features are designed, and so on."
What Krisp does today
Krisp started in 2018 as a simple noise cancellation feature. Vahagn describes it now as a voice AI platform with three business lines. The first is an AI meeting assistant that transcribes, summarizes and records meetings.
The second is Krisp for call centers, which includes noise cancellation, accent conversion and a newer live interpretation product. The third is Krisp as a technology that other voice AI companies build into their own products. Discord is one example, where Krisp runs in the background as a noise cancellation tool.
Building a category with no existing demand
Vahagn says his earlier jobs involved capturing demand that already existed. At Krisp nobody was searching for noise cancellation, so the team had to explain that the problem existed, explain why it mattered, and then position the product as the answer. He calls this a fundamentally harder and more expensive kind of marketing.
The first step was launching on Product Hunt, where Krisp was named product of the year in 2019. Gamers and freelancers signed up, and early adopters shared the product on Twitter and with friends. That word of mouth built awareness before any formal category work.
Users then included gamers, freelancers and call center agents, so the team wanted one umbrella category for all of them. G2 had no category for noise cancellation, so Krisp worked with G2 and industry analysts to define the software, the players and the problem. Timing also helped, because COVID arrived a few months after launch.
"Krisp went viral because it was solving a problem that was a new problem, actually. Probably before that, no one was thinking about background noise being a real problem."
Community and direct contact with customers
Vahagn says the early adopters were often tech-savvy decision makers and even influencers. Krisp kept close contact with them, and their requests for new features and products became the main source of ideas for what to build next.
The marketing team was small at the time, just Vahagn and one cofounder focused on marketing. They answered comments on Product Hunt, AppSumo and other platforms. Users knew they were talking to decision makers, which made them more comfortable sharing ideas. The cofounders also handled support tickets themselves.
Vahagn advises early stage startups, and cofounders in particular, to stay in direct contact with customers.
"Because if the customers feel that their comment, their feedback can be turned into the next feature or next product for the company, they feel much better."
Two growth motions under one brand
Each business line needs its own kind of marketing. The AI meeting assistant is a consumer product, so Krisp relies mainly on product-led growth, building the best product it can and gaining users and awareness through virality in the product itself. Enterprise and call center sales are sales-led and marketing-assisted. Marketing creates demand and leads, and the sales team closes them.
The brand holds the two together. Krisp positions itself as the voice AI platform that improves communication and makes people more productive, which Vahagn says applies to both consumers and call centers.
Internally, Krisp moved away from centralized teams. It now has smaller teams dedicated to one business line across engineering, product, marketing and sales, while leadership stays centralized. The technology foundation is still one shared voice AI platform. Customers pick the functions that suit them, such as meeting transcription for an SMB or accent conversion for a call center.
"One of the best decisions that we've made at Krisp in the past couple of years was to separate how these different business lines operate."
Balancing brand, SEO, LLMs and PR
Vahagn describes a waterfall for planning. A company vision covers the next three to five years, which becomes annual goals such as revenue and ARR targets, which are then broken into OKRs for short-term work. Fast channels like performance marketing and advertising serve the short-term goals. Brand building and SEO are treated as long-term investments, and SEO has been a priority from day one.
On the rise of LLMs, Vahagn says a lot has changed on the surface but the core job is the same. Marketers still need to create high-quality, firsthand content that is useful to customers. Krisp now uses AI tools to help produce that content. The team tracks LLM traffic through the new referral section in Google Analytics and through features that Ahrefs and similar tools have added. He also sees PR as a major help in getting listed by LLMs, because they look at what third-party sites say about a company.
On timing, Vahagn agrees with Mark that PR does not suit the earliest stage. A startup first needs a user base, customer feedback and case studies to build a compelling story. After a couple of years, when there is budget and stories to tell, it makes sense to test PR.
"I would say the right time for doing PR is when you have a strong point of view or interesting story or strategic narrative to communicate to the world."