Why Yoav Izhar-Prato keeps changing course

At a client summit in San Francisco in May 2015, Yoav Izhar-Prato gave a talk about transformation and changed his clothes onstage. There are sober ways to explain the shifting economics of digital advertising. A wardrobe change has the advantage of being harder to sleep through. In the photograph, he stands with his arms spread, wearing a Golden State Warriors jacket, looking rather pleased with the demonstration.

The setting was the Golden Gate Club in the Presidio. More than 250 marketers had gathered for Kenshoo’s K8 summit. His subject was the need to adapt as their industry moved around them. Eleven years later, the company is called Skai, he remains its co-founder, chief executive and chairman, and adaptation is still the assignment. The clothes were the easy part.

A useful problem on a small napkin

His account of the beginning involves rain, a coffee shop and a napkin. Alon Sheafer and Nir Cohen, his future co-founders, were trying to grow a comparison-shopping engine. Acquiring users became inefficient as the operation expanded. They needed a better return on the money spent bringing people to the site, just as search advertising was becoming an increasingly consequential business.

They built a tool to bring in traffic more efficiently. It worked well enough to suggest another business: sell the means of acquiring customers to other companies. Kenshoo was founded in 2006. A practical difficulty had become a product. The first version was bootstrapped; Sequoia subsequently became an investor. For all the romance that a napkin lends a founding story, the interesting object here is the problem written on it. Someone was spending money and wanted it to work harder.

Izhar-Prato came to that problem with a career already spread across countries and businesses. He had managed ECI Telecom’s Thailand operation, worked in marketing, business development and sales at CellPay, consulted for Motorola and founded Beijing PrePay. His route into advertising software ran through telecommunications and mobile payments. The recurring work was getting technology into a market, then making a business around it.

The question that keeps coming back

By 2011, he was already discussing a business that reached beyond search. Social advertising offered targeting; search revealed intent; display helped create awareness. The challenge was to put those different strengths to work together. A marketer could buy in several places and still struggle to understand how those purchases contributed to a sale. Kenshoo’s proposition was to help manage that complexity.

He also faced a familiar question for anyone selling software in a fast-moving market: how long before the product becomes a commodity? His answer placed the burden on continued movement and demonstrable customer value. There is something bracing about a founder accepting that the question must be asked every day. Yesterday’s useful product carries no automatic entitlement to tomorrow’s budget.

“My goal is to always remain humble and over-deliver.”

Yoav Izhar-Prato, 2011

One example he offered was wonderfully unglamorous. A tool called Watchdog monitored landing-page URLs. It helped a retailer identify $120,000 of wasted advertising spend caused by broken links in its first week. Advertising technology can occupy itself with elaborate theories of attention. Here, the immediate victory was that somebody noticed the door was locked while the shop was paying people to come in.

When the storefront becomes the advertising business

Retail media gave his company another version of the same challenge. A retailer already has shoppers, product information and transactions. Selling advertising around that activity creates a different opportunity from buying a search term or placing a social ad. It also creates another set of systems for a brand’s marketing team to operate.

In a 2023 essay, Izhar-Prato examined what retailers would need to do to grow those advertising businesses. His concerns were concrete: more useful data for advertisers, better access to insights and less friction when buying across platforms. A proliferation of retail advertising businesses would be of limited use if most budgets continued to pool around a handful of accessible networks.

That perspective makes the middle of the market interesting. Retailers want advertisers. Advertisers want results without having to learn a different operating procedure for every destination. The company helping them work across those destinations has a reason to exist for as long as the connections remain difficult. It is a business built in the spaces between other businesses.

His own explanation of Skai’s work has repeatedly returned to that connective role. The individual channels can change. The practical question persists: how does a team see enough of the picture to decide where its next advertising dollar should go?

The name had to catch up

On June 8, 2021, Kenshoo announced its new name. Skai combined references to Kenshoo, Signals Analytics and actionable intelligence. The acquisition of Signals Analytics brought market intelligence into a business already associated with advertising execution. The ambition widened to encompass more of the decisions made before, during and after a campaign.

Gil Sadeh, formerly the co-founder and chief executive of Signals Analytics and now Skai’s president, became part of that continuing story. The combination promised to connect information about markets and consumers with the systems used to act on it. For Izhar-Prato, leading the combined business meant explaining why a company known for one job should be trusted with adjacent ones.

A new name can be an expensive distraction. In this case, it announced a change in scope that customers would have to assess in their own work. Better planning, execution and measurement are useful ambitions. Their value emerges when a person with a budget can make a better decision. The napkin’s original demand for a better return had acquired considerably more machinery.

Giving the data a conversational partner

On April 2, 2025, Skai introduced Celeste AI, a generative AI marketing agent. Early exploration took place in a closed beta with organizations including Acosta Group, Advantage Solutions, Tinuiti and WaterWipes. The proposed interaction was straightforward: ask questions about business performance and receive analysis and recommendations grounded in connected marketing data.

For Izhar-Prato, this extended the responsibility he associated with the company’s founding. Improving campaign performance had grown into helping marketers interpret a much larger volume of information. A question about shifting a budget across channels requires context that a single campaign dashboard may struggle to provide. Celeste was intended to make that context easier to interrogate.

In September 2026, Agent Connect opened another route. Marketers and developers could connect their own AI agents to Skai’s infrastructure. The launch included reporting and campaign-action capabilities, with proposed campaign changes subject to human review and approval. Customers could bring an agent into the system rather than accept only the interface the vendor supplied.

An October 20 partnership with Power Digital made the idea tangible. Its Omega AI operating system would connect through Agent Connect to unified retail-media data. Skai would supply connections and standardized information; Power Digital could concentrate on the intelligence and workflows it wanted to build. For a founder who began with a tool that solved someone’s acquisition problem, the familiar opportunity was underneath the fashionable vocabulary.

The cost of changing direction

There is a harder record beside these announcements. Skai cut approximately 80 jobs in December 2024. In June 2026, it cut roughly 100 more, about a fifth of its workforce, in Israel and elsewhere. The company connected the latter restructuring to changes in how it worked, including its adoption of AI.

Those decisions belong in any account of a founder who talks about people and adaptation. New capabilities and fewer jobs can arrive in the same period. The employees leaving are part of the company’s history as surely as the products being launched. A claim about more efficient work has consequences beyond the customer demonstration.

Izhar-Prato’s public language often emphasizes the people around him. In a post marking Skai’s twentieth year, he described a gathering of more than 100 colleagues among the redwoods between Palo Alto and Santa Cruz. He used the growth rings of a tree to describe successive years of the business. It was an affectionate image of accumulated experience. The restructuring adds a difficult question about how that experience is carried forward, and by whom.

A different kind of depth

At ShopAble 2026, his closing presentation used an underwater image. Diving is a personal interest he has also discussed publicly, and he brought it into his account of approaching unfamiliar territory. His invitation was to choose exploration deliberately. After years of asking clients to change how they work, he was still looking for a way to make the invitation feel personal.

There is continuity between that underwater backdrop and the Warriors jacket in 2015. He reaches for a physical experience to explain an abstract business demand. Put something on. Go somewhere unfamiliar. Try the thing. Marketing software supplies plenty of acronyms; a person standing in front of a room still needs to give the room a reason to care.

In October 2026, Izhar-Prato leads a business approaching its third decade, with a different name and a growing collection of AI offerings. Its oldest question remains surprisingly compact: can this help somebody make a better decision? The career around that question has crossed borders, products and advertising channels. The next answer will have to work in practice. Even an accomplished wardrobe change only buys so much attention.