Salesforce Challenges the "SaaSpocalypse" with Its New AI Agent "Albert"
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisis ton rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
Salesforce bets on "Agent Albert" to ease AI concerns
Marc Benioff, CEO of Salesforce, firmly opposes the worries that artificial intelligence agents could render traditional enterprise software obsolete. He views this technology as a major growth opportunity for the company.
Currently, Salesforce's AI tool, known as Agentforce, is capable of efficiently handling routine tasks. However, it shows weaknesses when it comes to processing more complex requests. To address this, Salesforce plans to launch a new product, "Agent Albert," by the end of the year. This new agent is designed to automatically analyze users and take actions independently.
A new metric to measure the impact of AI
In parallel, Salesforce has introduced a new measure called "Agentic Work Unit" (AWU) to assess the impact of AI. This metric is designed to quantify how AI concretely contributes to outcomes, such as the number of requests resolved.
Current challenges and Salesforce's response
Wall Street's concerns focus on the risk that AI could make traditional enterprise software obsolete, which could affect Salesforce's user-based pricing model. This fear is exacerbated by the possibility that companies might develop their own software, thereby reducing their reliance on Salesforce. Since the beginning of the year, Salesforce's stock has dropped by 28%.
In an interview with the Wall Street Journal, Marc Benioff dismissed the "SaaSpocalypse" theory, asserting that AI enhances the value of Salesforce. He emphasized that in-house solutions carry risks regarding data security and compliance, and that the growth opportunity has never been greater.
Mixed performance of Agentforce
Despite the promises of AI, the results of Agentforce raise questions. Since its launch at the end of 2024, only 23,000 of Salesforce's 150,000 customers have adopted this tool. While it performs well for simple tasks, as noted by the education company Pearson with a 40% increase in automatically resolved requests, it struggles with more complex tasks. For instance, jewelry manufacturer Pandora reported that Agentforce fails to provide reliable recommendations when requests are vague.
In its latest quarterly report, Salesforce recorded 2.4 billion AWU, marking a 57% increase. However, this is the first time this metric has been published, preventing any external comparisons.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.