Another year, another Dreamforce, and yet another game-changing announcement from Salesforce that is set to change the world of work as we know it. Of course, this happens every year. The Salesforce hype machine spools up, and product announcements such as Chatter, Einstein GPT, and Data Cloud, are all meant to change the way we work forever.
However, fact from fiction in the world of Salesforce is hard. The cloud giant always wants to ensure it’s at the forefront of any new technological advancements, and as such, needs to reinvent itself each and every year.
Over the past decade or so, Salesforce announcements have been hit or miss, but mostly miss. Products such as the IoT Cloud, Work.com, NFT Cloud, amongst many others have been all but forgotten. But when you compare this to the first decade Salesforce were in business, it pales in comparison to releases such as Service Cloud, the AppExchange, Force.com, Salesforce Lightning, and Chatter.
One exception from this is Data Cloud, which by all accounts, seems to have been an extremely successful launch of an organically built product that is also providing the foundation for Salesforce’s future AI and Data strategy.
One thing that is for sure with Salesforce growth numbers declining quarter over quarter, is that Salesforce need to stimulate huge chunks of revenue from somewhere. Artificial Intelligence is surely the best bet they have at becoming a dominant leader in the next wave of technology.
Is Agentforce Different?
Putting aside the fact that Salesforce has been an AI leader and researcher for the best part of a decade, last year’s Dreamforce announcement hasn’t seemed to stimulate the growth that Salesforce was hoping for. The Einstein Copilot Studio, which includes Copilot, Prompt Builder, and Model Builder, are let down by a few factors. In my opinion, it’s let down by a lack of AI readiness for enterprises, the relatively rudimentary output of LLMs, and a lack of fleshed-out use cases.
In Salesforce’s keynote yesterday, they released that over 82 billion flows are launched each week. In comparison, there are only 122,000 prompts executed. Now, of course, Flow has been around for many years, but this is less than one prompt per Salesforce customer per week on average.
Back at the tail end of 2022 with the release of ChatGPT, many of us were hailing in the next age of AI and anticipating that any day now, our whole working lives would change. Two years later, it’s clear that we are pretty far away from seeing any meaningful impact of AI, especially one that is going to affect global GDP.
However, this latest release from Salesforce feels different. Although it’s the first time many of us have heard about AI Agents, this concept has been discussed in AI circles for decades.
Marc Benioff has made a few statements in this year’s Dreamforce news hype cycle that feel a bit different from the usual. But mainly, the fact that Salesforce is hard pivoting in their agent strategy, as well as taking a swipe at Microsoft and how their Copilot product has disappointed customers. It’s an out-of-character move that shows the AI war is really heating up.
Whatever Salesforce’s marketing strategy for this year’s announcement is, an agent-based strategy certainly feels more like the revolution we have all been promised. Although we will no doubt see tools such as Salesforce Copilot develop over years to come, agents that are able to automatically handle cases for service agents, as well as handle questions and book meetings for sales users instantly bring ROI to a Salesforce org.
Is the Future of Salesforce in the Hands of Agents?
As always, there are still many looming questions. Are Salesforce customers ready for this? Are there orgs and teams in a position where Agents will be able to work effectively? Are we dealing with some level of vaporware or is this the real deal?
Although Salesforce and the V word do get mentioned in the same sentence quite often, this probably isn’t the case with agents. It’s been widely reported that Agentforce has partly been influenced by the Airkit.AI acquisition, who claim to solve 90% of customer queries via its platform. Salesforce also has set up dedicated launchpads at Dreamforce for customers to start building their own agents.
Of course, Salesforce’s complexity, technical debt, and growing platform (that remains somewhat disconnected) will be a concern for customers.
Whilst these are very real concerns that were further highlighted in our first annual Salesforce developer report earlier in the year, you also can’t deny the genius in Salesforce’s strategy over the years.
Their platform has matured to such a point, that they have essentially every functionality you could ever want out of a platform, albeit, sometimes a bit disjointed. For example:
- The ability to bring your entire organization’s data into the Salesforce ecosystem via Data Cloud, using a zero-copy architecture with partners such as IBM Cloud, Snowflake, Google Cloud, Workday, AWS, and more.
- Using Salesforce’s automation suite such as Flow, Apex, and the Einstein Copilot Studio including Prompt Builder to automate time-consuming tasks.
- Then, use a multitude of channels such as Phone, Slack, WhatsApp, Email, SMS, and live chat to interact with your customers on the platform of their choosing.
- Now with Agents, the time-consuming and expensive part of having a human interact on these channels, potentially replaced forever.
With this potentially game-changing innovation, Salesforce are also planning to bring in usage-based pricing, teasing a $2 cost per conversation.
Will this be the innovation that Salesforce has been trying to crack in recent years, and turn Salesforce back into the growth story of the past two decades?
Final Thoughts
Whatever the future holds for Salesforce, Agents look to be a big part of the world’s future AI story, with ecosystem leaders such as Copado simultaneously launching their own Agents.
What is true is that workforces are overwhelmed and companies are overstretched trying to chase targets that are increasingly hard to hit in the current macroeconomic environment. This is coupled with the fact that customers’ expectations on the level of service they receive is higher than ever.
We will all wait with bated breath on whether Salesforce has cracked the code on AI. But putting sales agents to the side, whoever cracks customer service AI first will have a huge head start on the windfall of AI spending that will occur in the very near future.
Comments: