Wix CEO Avishai Abrahami on Base44, AI and Building for Small Businesses

Avishai Abrahami says his unusually late working schedule gives him uninterrupted time to think and plan. The Wix CEO also believes his absence earlier in the day encourages colleagues to resolve some issues themselves. He presents this as a personal routine, not a recommendation for everyone.

Harry Stebbings asks why Wix’s market valuation is close to the revenue figure he cites. Abrahami attributes pressure on software valuations partly to uncertainty about AI’s impact. Both the financial figures and their market explanations reflect the speakers’ discussion, not independently verified current valuations.

Stebbings asks how investors value Base44, Wix’s AI application-building acquisition. Abrahami argues that comparing it with other AI builders implies little value for Wix’s existing business. His hypothetical valuation and peer comparison are his opinion, not an independently established market price.

Abrahami argues that Salesforce’s value includes trust, security relationships and an ecosystem around customer data, not just the visible CRM interface. He believes those factors can make replacement difficult for large enterprises. The discussion does not establish that any particular customer has no viable alternative.

He contrasts this with Atlassian, where he believes developer customers may be more capable of building their own alternatives. Abrahami discloses that he owns Atlassian shares and says he admires the company; his concern is an investment opinion, not evidence of customer losses.

For Wix, Abrahami describes customers such as small restaurants and hairdressers who have businesses to run and may not want to build every system themselves. He says some will use AI coding tools, while he expects many to continue using packaged products.

He recounts an internal experiment to build hairdresser business logic using Base44. According to Abrahami, even experienced teams did not reproduce the full workflow quickly. The interview gives no independent evaluation of the test, its scope or whether other teams would achieve the same result.

Stebbings asks whether even partial customer movement shrinks Wix’s addressable market. Abrahami notes that Wix owns Base44 and argues it can serve customers who want more customization as well as those who prefer a ready-made product. This is his strategy for possible customer shifts, not a guarantee of retention.

He expects a mixed market over the next several years: some owners will use packaged services, some will build customized applications and others will combine the two. He says Wix is investing in Base44 partly to prepare for the latter possibilities.

On margins, Abrahami says Wix’s traditional product is substantially more profitable than Base44 and has better retention. He says the cost of generating and hosting a conventional website is lower, while building with AI involves more compute. The company does not supply detailed audited segment economics in the interview.

Abrahami says Base44’s costs are falling and describes a combination of fine-tuned models intended to provide good quality at lower cost. His comparison with leading models is a company claim; the interview does not provide a common benchmark or independent quality assessment.

The rationale for a specialized model, he says, is that Wix can learn from the tasks users attempt and the points where they struggle. Rather than reproduce every capability of a general model, the company wants better performance on application-building requests relevant to its product.

He says Wix also trains a model for website generation and uses customer feedback to improve it. When asked about cost savings, he gives a range and notes that the result depends on model and task. The comparison is not detailed enough to generalize across providers or workloads.

Stebbings asks whether the market is still prioritizing customer experience over cost optimization. Abrahami agrees that quality is particularly important in an early market and says he would choose a better result over a comparable cost reduction. This is his product judgment.

The CEO says Wix publishes Base44 information periodically but does not break out every metric continuously. He explains that customers and partners can move between products, which complicates the story. The interview does not resolve every question about segment reporting.

Wix encourages agencies and partners to try Base44 despite its lower margin, Abrahami says, because he expects a combination of visual tools and AI-assisted building to matter. His view that both Wix and Base44 will remain relevant is a forecast from their owner.

Abrahami says he tries not to tie his mood to daily stock movements. He believes Wix’s share price is affected by wider AI news as well as its own performance, and focuses instead on decisions he can influence. That explanation of market behavior is his interpretation.

A lower share price can make stock-financed acquisitions harder, Stebbings notes. Abrahami says Wix has access to cash, but the practical constraint is integrating and executing on Base44 while building another product. He believes taking on another similar acquisition at the same time would strain the organization.

He describes Base44 as a very small operation when Wix acquired it and says Wix supplied people to build a broader company around its founder. The host cites a purchase price and revenue figures; those numbers are part of the conversation and are not independently verified here.

Abrahami says the board focused on business logic, team building and how the product compared with alternatives. He recalls that the decision was easier than the host’s framing suggested. This is his account of the board’s discussion, not a complete transaction record.

Asked about cash, he cites roughly $400 million in annual free cash flow and says Wix is investing in Base44. The captions are unclear on the precise investment amount, so the interview does not support a reliable figure for that item. His financial statements are not independently audited here.

The host raises a hypothetical take-private transaction. Abrahami declines to speculate and says his focus is building the company. He emphasizes that investors may make different judgments about Wix and the effect of AI; the exchange is not investment advice.

Abrahami acknowledges that Wix’s share buyback had poor short-term timing in hindsight. He says the company had cash and was focused on existing projects rather than a large new acquisition. Whether the decision creates value depends on future results, which remain unknown.

He views buybacks as one way to return value and offset dilution from issuing shares. Stebbings raises concern about stock-based compensation, while Abrahami argues employee ownership helps attract and retain talent. Their comments are opinions about capital allocation, not recommendations to investors.

Abrahami lists acquisitions, buybacks and additional engineering investment as possible uses of cash. Each has limits: acquisitions require integration, repurchases use capital and hiring changes expenses. He says management must balance investors’ interests with obligations to customers and employees.

On retention, Abrahami says there is no magic solution when stock prices fall and employees leave. He says some departures can create opportunities for others, while stressing that he works hard to retain top talent. This is his management philosophy, not a statement about specific employees.

He reports roughly 3,500 people at Wix including about 400 at Base44, and says customer support is a large function. The figures are company-reported snapshots and may change. Serving customers in many countries makes a very small support team difficult, in his view.

Asked whether Wix uses external AI customer-support products, Abrahami says the company tested several options and ultimately continued building internally. His account concerns Wix’s requirements and experience; it does not independently establish that those products cannot work for other customers.

He says it is hard to predict the future size of the team. AI is powerful, he argues, but detailed work often exposes limitations that broad demonstrations conceal. He resists promising that customer support will disappear, while remaining open to further improvement.

Abrahami’s more ambitious hope is that AI will help small businesses run operations, reach customers and sell additional services. He predicts meaningful gains soon, using a gym or personal trainer as examples. These benefits and timing are his expectations, not verified customer outcomes.

The host asks whether that optimism conflicts with his caution about model quality. Abrahami distinguishes useful, well-scoped workflows from uncritical reliance on a model’s answer. He recalls challenging an AI-generated safety protocol and finding that the suggested requirements changed after questioning.

He sees potential in combining models with specific tools and workflows for scheduling, customer conversations and other business tasks. He says small businesses tend to look for practical help with daily problems. His comments reflect his observations, not a representative survey of all owners.

Abrahami says students may worry about a rapidly changing job market, although he does not expect current language models to replace every white-collar role. He also allows that further advances could change the picture. Both assessments are uncertain forecasts.

Stebbings discloses an investment in legal AI company Legora and asks whether AI will improve in legal work as quickly as in coding. Abrahami thinks text-heavy legal tasks are promising, but distinguishes them from scientific research that may require other tools and methods.

The conversation turns to medical uses of AI. Abrahami discloses an investment in K Health, and the speakers express differing views about model and clinician performance. They provide no independent clinical evidence in the interview; their opinions are not medical advice or proof that an AI system is better than a doctor.

Abrahami argues that general language models alone may be insufficient for certain scientific research tasks and that specialized models or simulations may be needed. This is his assessment of current technology and a conditional view of possible future architectures.

Turning to relationships, he shares a story about his wife communicating that a bad mood was not caused by him. For him, being aware of one’s own emotions and not automatically placing them on a partner is valuable. He presents it as personal experience, not a universal formula.

The two discuss whether financial security changes decisions. Abrahami says less fear can support more measured choices, while the host notes that greater wealth can also raise expectations. They describe personal experiences rather than a general law about wealth or judgment.

For Abrahami, the central benefit of having enough money is freedom to choose where and how to work. He says recognizing that he has chosen to lead Wix helps him commit to the role even during difficult periods.

Asked how to build resilience, he says setbacks are inevitable and often unpredictable. His approach is to ask whether he is doing the best he can with what he can control, rather than assuming a quiet period will continue indefinitely.

He says meditation and other personal practices have helped him manage pressure. He also reminds himself that happiness is not determined solely by wealth. These are his personal experiences, not evidence that a particular technique will work for everyone.

Abrahami recalls struggling to leave work stress behind at home. He says he now postpones some thinking, sometimes writing down an issue to revisit later, so that he can pay attention to his partner. His late-night work routine is part of that personal approach.

Asked about parenting, he emphasizes being present and engaged during time spent with children. He acknowledges that this can be easier to say than to practice. The advice reflects his experience rather than a substitute for understanding each family’s needs.

Abrahami says he has revised his expectations about AI: he now thinks replacing people in many roles may take longer than he once feared. He still considers the technology powerful and acknowledges that capabilities can change. This is his current judgment, not a timeline.

He praises Figma’s product understanding and the attachment he sees among its users. The comparison is his opinion about another company’s product, not an independently measured competitive ranking.

His largest concern is execution: helping a team work efficiently and pursue ambitious goals. He imagines Base44 might grow substantially over two years, with many engineers, but says headcount is difficult to predict and depends in part on AI progress.

Abrahami thanks people who shared their time and advice when he was starting out. He sees mentorship as a generous gift of a valuable resource: attention from people who have already learned through their own mistakes.

Asked to compare Wix, Monday and Base44, Abrahami says he values the opportunity to have been involved with several products and sees potential in each. He praises the people leading them while declining to predict which company will have the higher valuation. These are personal reflections and business opinions.