Kristoffer Kvam is the founder and CEO of SMOC.AI, which builds proactive AI sales agents that help companies engage customers and turn attention into leads and sales. He talks with Laurent Cochet about how a telco career led to the company, how its agents work across advertising and social channels, why enterprise sales move slowly, where sales automation is heading, and the mistakes he would warn other founders about.
From a telco career to B2B sales agents
Kvam has a background in both engineering and business, with an MBA and a master's degree in AI. The idea for the company came while he was heading the AI and marketing business at group level in a major telco operation. There he saw how much companies struggled to win attention and sales while large digital players such as Meta and Google controlled their customers' attention.
His first idea was a consumer business that would use AI and gamification to make a small mobile operator profitable. Along the founding journey, the company converted to B2B. The underlying problem stayed the same, he says: low engagement on social posts and emails, traffic that does not convert, and B2B leads that are very costly to get. By his account, customers see two to three times better performance with the company's technology.
The first problems he saw were at airlines and telcos, but the area where SMOC.AI is growing most is small and midsize businesses. With cold calling no longer working, he says, these companies want leads without having to learn every channel, and SMOC.AI abstracts that work away so they can go live quickly.
How proactive agents work
Kvam frames the problem in his own figures: by his account, a chief marketing officer now spends three times more on ads to acquire the same customer than ten years ago, while cost per click has increased five times over the same period. Chat services cut organic traffic, and the tools used to convert visitors remain static at a time when people expect interactive experiences.
SMOC.AI sets up agents in each relevant channel. Reinforcement learning agents steer advertising platforms toward more qualified leads, then work to convert those leads by interacting with visitors instead of showing a static page. For consumer brands this can take the form of a quiz; he mentions an airline among the customers that use one. Leads are then pushed into the customer's email, CRM or the SMOC.AI platform.
He explains proactive agents by contrast. A reactive agent is typically a customer service bot that answers questions. A proactive agent goes out, finds customers and converts them toward a lead or a sale, much like a shop assistant who asks what you need for a party and sells you a jacket and shoes as well as pants. Much digital selling is reactive search and price comparison, he says; his aim is to reach buyers earlier, when awareness and interest are forming.
Why enterprise sales take so long
The company grows through two models. For smaller businesses, a self-service model lets the AI give the advice and build the brief, in a way replacing the agency. For enterprises, SMOC.AI works mainly through resellers and partners, because enterprises often have more trusted relationships with agencies. Kvam also does some sales himself.
Enterprise resistance, in his experience, comes from a build-it-yourself mentality and from data protection questions, including where data is located and negotiating a full data processing agreement. A typical sale converts in two to three weeks, he says, but with enterprises it can easily take nine months. His recommended approach is a low-risk pilot that the enterprise can scale once it sees results.
He also points to the classic checklist of budget, authority, need and timing. Exciting the people with the need is easy, but the authority may not prioritize the project, while going straight to the authority can feel to the team like someone stepping into their shoes. His ideal is to engage both at the same time. Timing matters as well: the end of the year is easier than just before summer.
Automating the boring part of sales
Kvam expects the early stage of sales, qualifying prospects and finding the right fit, to be automated. Network-based selling, he believes, will become even more important, because people like to talk to other people.
The next target is follow-up, which he calls the very boring part of sales. Prospects who ignore an email, a call or a LinkedIn message are often just short of time, and following up and finding other roles in the account takes a lot of manual work. SMOC.AI is not working on this yet, but it is on his plate. He compares it to the AI tool that now collects his receipts for the 40 different kinds of software he buys every month, and expects this kind of automation over the next five years.
For the company, success means a salesperson who needs more leads can sign up and start generating leads, including campaigns, in five minutes. Advertising platforms are very confusing to navigate, he says, and the interface has to be simple enough to make it happen for everyone.
Lessons for founders
Kvam says one of his biggest mistakes was leaving sales and customer success to others early on. Customer conversations are now where he finds the value, and he acts on them directly: with AI generating much of the code, he counts himself among the most productive coders on the team, with agents running on ideas from those conversations.
He has also negotiated very big contracts with clients without delivering on expectations. As he describes it, a client may agree to pay 10,000 or 20,000 euros for a pilot on the strength of a fantastic vision, and then the pilot ends without results. Customers now care about value, which is why token-based, credit-based and consumption-based pricing is spreading. His advice: never give the product away for free except very early on, make price incremental to value, and avoid pricing too high, because churn will follow.
To his younger self, he would say not to raise too much money before product market fit, since investors push founders to hire too early and a small founding team can do a lot with AI. His company went too broad too early. The habit that has helped him most as CEO is spending time reflecting, whether on a walk or through meditation, because inspiration for change does not come while you are only acting.
Watch the full conversation
Watch the full conversation above. Kristoffer Kvam invites people to get in touch through the contact page on the SMOC.AI website or to connect on LinkedIn.
AI InterConnect hosts conversations with founders, investors and operators building with AI. Explore more episodes on the AI InterConnect podcast page.