Wikitree Selects DeeperDive From Taboola, Gen AI Answer Engine Built for the Open Web, to Connect Readers with Timely, Contextual Answers for Topics They Care About

Taboola (TBLA) announced that Wikitree will deploy DeeperDive, a genAI answer engine integrated into publisher websites, to boost readership and engagement and open additional revenue streams (including high-intent ads within AI results). The deal is positioned as a step in Taboola’s agentic roadmap and reinforces its push to embed AI-powered discovery directly on publisher properties. Overall, the announcement is incremental but directionally supportive for Taboola’s monetization and platform differentiation.

Analysis

This is best viewed as a proof-of-distribution event, not a revenue event. For TBLA, the important question is whether AI can migrate the company from low-multiple traffic monetization into a higher-quality software-like layer with recurring publisher relationships and better gross margin mix. If that happens, the market could start valuing TBLA on product adoption and retention rather than pure ad-cycle beta.

The second-order risk is cannibalization of the very inventory TBLA is trying to monetize. An on-site answer layer can improve dwell time, but it can also reduce page refreshes and click volume, which means engagement metrics may rise while monetizable impressions fall. That creates a winner/loser split: publishers with proprietary archives and strong brands can gain leverage, while commoditized open-web traffic brokers and ad-tech names tied to raw clicks face margin pressure.

Near term, this should trade mainly on sentiment. The next 1-3 months catalyst is not another logo, but evidence that the product is repeatable: multiple publisher wins, disclosed RPM uplift, or any commentary on inference costs versus incremental ad yield. The contrarian miss is that AI UX on publisher sites may be economically attractive in demos but still mediocre at scale if it shifts traffic from high-volume pages into lower-impression conversational sessions; if the next two quarters do not show measurable EBITDA contribution, the market should fade the AI narrative premium.

Key Decisions for Investors

• TBLA: keep as a small tactical long only on weakness, not on the headline. Time horizon 1-3 months; thesis requires follow-on publisher wins and any quantified uplift in revenue per session. Falsify on the next print if management cannot show monetization or if AI-related costs offset engagement gains.
• TBLA vs. ad-tech basket: long TBLA / short a basket of click-dependent ad-tech names such as PUBM or MGNI for a relative-value expression over the next quarter. The idea is that TBLA has the earliest optionality on AI-native publisher monetization, while the short side is more exposed to AI-driven click cannibalization. Cover if the product fails to scale beyond pilot logos.
• Do not trade XIACY on this item. There is no direct fundamental read-through; treat any move as sympathy only unless TBLA proves this model is being embedded across OEM or large-platform distribution.
• Set an alert for TBLA next earnings: if management does not disclose a meaningful pipeline of DeeperDive deployments or any margin-neutral monetization data, reduce or exit the thesis. If they do quantify adoption, the stock can re-rate faster than fundamentals because expectations remain low.