Omri Shtayer on AI Agents and Data-Driven Growth

1. From a strategic perspective, how crucial is the depth and recency of data in the effectiveness of AI-powered insights for your organization's growth initiatives?

No organization should ignore the potential of AI to improve marketing effectiveness and productivity. That being said, marketing leaders shouldn't  expect these tools to deliver the results if they’re being starved for data – that’s when you run into the unfortunate syndrome of chatbots “making stuff up” in an attempt to answer a question. The data being fed to the AI tool  needs to be as complete and recent as possible, so that the AI tool stack is well informed and knowledgeable about your company and your market.

That’s particularly critical as we move from chatbot interactions to deploying AI Agents that we expect to do work on our behalf, more autonomously over time. We need them to execute tasks based on hard data and well-defined plans.

Think about data being like fuel for an aircraft. Skimp on fuel, and you won’t get high performance. You may not make it to your destination.

2.In what ways do you believe specialized AI tools can provide a competitive advantage in areas like SEO, sales, and market trend analysis?

Specialization in the software and the data is how we make AI more productive and impactful for marketing and sales. In addition to providing AI tools with high quality and up to date data, we want to focus it on the problem we aim to solve.

Similarweb’s mission is to provide marketers with more and better data, but the beauty of adding AI agents into the mix is that we don’t have to worry as much about overwhelming the users. Similarweb’s AI agents can cover a lot of ground very quickly, meaning, they can accomplish in minutes what otherwise would be time consuming research tasks.

With a conventional user interface for a data system, we work on developing individual screens and charts and dashboards to help users understand and navigate through the data. But given the limitations of our poor human brains, as consumers of that data it still takes time to look at each screen, understand what it means, and maybe download data into spreadsheets for additional analysis.

Instead of looking at one screen or one spreadsheet export at a time, they can look at all the data and make sense of it for us. We get the answers we’re looking for, all together, not in bits and pieces.

That doesn’t mean you let them do all the thinking for you. You and your colleagues still have to review the analysis the agent has produced and determine whether it is sound and makes sense for the organization. You might want to do some spot checking of the underlying data, just as you would for a new employee whose work you don’t entirely trust. However, the odds that the data will be correct and the output will be useful are a lot higher if you work with an AI agent that is specifically trained on marketing and the data that is important to marketers.

3.The "AI Meeting Prep Agent" is cited for reducing research time. How critical is the efficiency gained from such tools in optimizing the productivity of sales and business development teams within your enterprise?

The AI Meeting Prep Agent is a good example of focusing the technology on a common business task or challenge. A good salesperson will go into a meeting prepared to achieve the best outcome, and it’s something they do many times every week, or even every day. The AI agent functions as a capable assistant who finds out your goals for the meeting and produces a thorough briefing on the people you will be meeting with and the opportunities and challenges of their business, based on news reports and company data as well as digital signals.

The reception from Similarweb Sales Intelligence customers has been very enthusiastic, with customers sharing feedback about saving hours of work and going into meetings better prepared. We have a similar story with our AI Content Strategist, which crunches data on SEO and competing companies, identifies content gaps, and makes specific recommendations.

In addition to the agents announced at the end of May, we’re working on a lead generation agent, a competitive intelligence agent, a stock research agent, and an ad creative agent — while scanning the horizon for other opportunities to build or buy additional products. We also have several projects under way to create agents for our own internal use.

4.How can rapid AI-driven insights into emerging market shifts directly inform and accelerate your strategic decision-making processes?

Everyone who works in digital marketing knows how fast things change. Even before the current panic over changes in Google search (search clicks decreasing as summary AI Overviews appear in the results more often), digital marketers were continually forced to adapt to changes in the algorithm. The same concept applies for marketing on Facebook, LinkedIn, or X. All of this has to be considered before marketers begin to evaluate changes in the competitive landscape. For example, look at how fast Temu went from nothing to a major ecommerce player – becoming the #2 most visited ecommerce website in the US by August 2024 – and how fast it retreated from the US market in the face of new tariff policies in the past few months, throttling back paid advertising so that it dropped to #11 in May.

Change is constant in marketing, and often it feels like a full-time job to stay on top of. Even with access to good data, marketers could use help sorting through it. This is where AI agents can assist. They take us beyond providing dashboards of data to providing analysis and making recommendations.

5.Looking ahead, what specific opportunities do you foresee in adopting and scaling AI Agent technologies across diverse departments within a large enterprise, and how do you plan to address them?

I love that question because the answer is: we’re just getting started imagining all the possibilities. The whole MarTech industry is going to shift to competing over who can deliver the best agents. And by best, I don’t mean the ones with the flashiest demos, I mean the ones that deliver practical results.

Our customers are just getting started sifting through all the AI-related pitches that are coming their way to determine what’s real and what’s smoke and mirrors. As they get more comfortable with this agent-driven generation of AI technology, they will be better equipped to ask for what will be useful to their organizations.

Don’t get me wrong, we’re already seeing strong demand and a pipeline of desirable AI agents that will keep us busy for a long time to come. I think the “scaling” limit is less about scaling technology than it is about learning to work with it productively. Enterprises will need to scale their leadership in AI adoption, educating employees on embracing their new AI team members and putting them to work productively.

At the same time, skepticism is entirely warranted. Every MarTech product is going to claim to be an AI product as part of the same old tech hype cycle. Tech buyers will need to sort out what’s real. It’s my job to ensure Similarweb comes down on the right side of that.

Get in touch with our MarTech Experts.