I cannot seem to escape all the posts about how Agents (Digital Workers) will be replacing key parts of GTM. Jacco from Winning by Design does a great job making the case for AI in GTM. He goes so far as to say -
In 2025 AI Won't just Assist people it will replace them. 70% of GTM efforts will be handled by AI.
He is spot on about GTM inefficiencies -
The underlying cause is that human sellers are too expensive, lack the knowledge to provide value in buyer-seller interaction, and cannot keep up with ever-changing needs
He is also spot on about the promise of AI -
You can’t achieve exponential growth with a linear, people-based approach. The key to exponential growth lies in leveraging exponential systems powered by AI.
AI agents won't quit, they don't want pay raises, they will all deliver the same quality of work and they will work round the clock.
He talks about how the buyer now knows more about the problem, the market and sometimes the solution
This I am not so sure about. Let's break this down -
The buyer knows more about their problem is often true but not always. For example, we all know we have a GTM problem. Cost of sales is going up, revenue per rep is down etc. but is the problem a GTM problem or something else? Most organizations think something is wrong with their GTM and keep firing and hiring CROs.
Yes, there is a lot more information out there but here I have to agree with Yuval Harari when he says more information does not mean you have more knowledge. To find the truth requires a lot of research and effort and if you have a ton of information it will take a lot of time to find the signal from the noise. Today buyers can go and read up about every company and product but figuring out if the product can really meet your needs is tricky. How do you know what is reality and what is a sales pitch? - https://www.youtube.com/shorts/694_rtzVttM
The AI world is changing so fast and moving so fast, how is the buyer supposed to stay on top of everything. Last year it was ChatGPT and CoPilots, this year it is Agents. Then we say that agents can be autonomous or need a human in the loop. There are concerns about autonomous agents optimizing for the goal at the expense of other consideration and going rogue.
There is this thought experiment that Nick Bostrom proposes where an AI asked to build paperclips could bring an end to humanity. While fanciful, it does show that even simple and innocuous tasks about have serious unintended consequences - https://www.iflscience.com/how-an-ai-asked-to-produce-paperclips-could-end-up-wiping-out-humanity-68432
If you look at Salesforce stock price and all the buzz about AgentForce you might believe that it is a fundamental game changer, but is it? or is it Wall Street hype?
The point here is that I am not so sure that buyers can necessarily be more educated.
The big vision here is that we will have agents handling all aspects of the customer journey and these agents will collaborate with each other.
In a lot of ways, this is just the next iteration of Digital Transformation. Companies have been doing this with PLG (product led growth), for their 'scale' customers companies have had Digital CS and Digital Support and Digital Onboarding. Before GenAI, the only way to do this was through in product experiences or having customers log into Customer Portals or via email. It is not a surprise that a lot of these digital efforts had spotty results but with GenAI we can absolutely provide a much more superior experience.
While I absolutely buy into this vision, the last piece that I am very interested in is the product that will help us build all these agents. Jacco talks about 1mind.com and their superhumans. 1mind has something quite compelling to handle inbound interactions and does something really interesting by joining demos and handling technical questions (like an SE agent). It does seem to work on unstructured data and that will be limiting (I can be wrong about this).
That being said there are a few other companies that tackle different aspects of providing agents around GTM.
www.devrev.ai has built something extremely compelling for support agents that ties integration, analytics, workflows all in one engine. They call this SaaS 2.0 and I think have a good example of what a GenAI first architecture could look like.
To handle outbound, there are a lot of AISDR companies - AISDR.org, 11x.ai, Artisan.co, Relevance.com etc. Each of these in their own way, tackle the entire SDR/BDR outbound workflow.
Salesforce is building AgentForce and Microsoft came out with 11 agents.
Agents and digital workers are an inevitability. Just like Algorithms are the digital workers underlying the B2C platforms (Facebook, YouTube, Uber etc.) Agents will be the digital workers in B2B.
The biggest challenges we will face will be around good data, guardrails and liability. Just like a self driving car would not be possible without accurate maps and then training it with guardrails, the same will be needed for agents. From a liability standpoint, if my self driving car runs a red light or hits a pothole or gets in an accident, who is liable? Similarly when the agent goes rogue - gives my customer a big discount or gives the customer some wrong advice that brings their systems down etc. - who is liable?
The other reality check is that this will take time. Elon Musk promised self driving cars ten years ago and they have come a long way. The Full Self Driving capability in Telsa's are amazing and hard to believe but they are not autonomous. Similarly AI agents handling all aspects of GTM will take time.