Posted by When I joined my current company as VP of Sales, my first major decision was to buy Gong. We signed and implemented the tool within a week of my start date. Implementing Gong was widely considered to be a very good decision. It gave our organization tremendous visibility into our sales calls, it elevated our coaching culture, and it helped our reps to reference conversations quickly in order to make their jobs easier. I was lauded for having so much conviction in the decision, and our CEO even mentioned in my performance review that it was one of many highlights for our team last year.
At my last job, it was the same exact thing. I joined, and my first act was to bring in Gong. There, the tool was used across the entire revenue organization, inclusive of Customer Success. Even prior to that, it was the same thing. In that role, I had actually been an outside consultant for the company prior to joining them. It was ironically in my consulting capacity prior to me joining full-time that I made the recommendation to the CEO to implement the tool. Even though they were pre-revenue at the time, I believed they should have the “Ferrari” of the conversational intelligence space. So you could even argue that in some companies, I have implemented Gong in a negative amount of time.
I mentioned Gong in my book nearly six years ago and have continued to sing their praises in various interviews and podcasts ever since. Why is that? Well, I personally believe that Conversational Intelligence is the holy grail of any revenue organization. Everything you possibly want to know about how to make your go-to-market efforts better comes from the conversations your team is having with prospective clients and partners. And whether that takes form in Zoom calls, emails, text messages, or even face-to-face meetings, the technology exists now to capture it all. With a device called Plaud, you can record your in-person meetings and have them captured and transcribed just like a traditional virtual meeting.
The reason I always loved Gong was because they were so far ahead of their peers. They started providing forecasting tools, they began to provide qualification rubrics to see how your team was performing across SPICED and MEDDPICC, and for a long time, they offered the ability to track certain keywords. For example, if you wanted to know which of your competitors was mentioned the most by your clients, you could easily capture and depict that data and also see how it evolved over time. This was all very cool and much more interesting than simply recording the call.
So it pains me to say that I feel Gong has done such a painfully slow and poor job at innovating that I actually feel it is now fiscally irresponsible for any revenue leader to invest in the tool. I realize that is a pretty charged thing to say and I have no personal issue or vendetta with Gong – clearly I have liked them a lot over the years and recommended them to many people. But AI is completely changing the way teams operate, and unfortunately the tool that Gong offers today has basically been commoditized.
Almost a decade ago, I had an idea for a business. The idea was initially to take the data from all of your phone calls and use that data to automatically update the CRM for the sales rep. One of the most painstaking tasks for sales reps is doing glorified data entry: listening to something a client says, and entering it into the CRM to make sure the data is up to date. In addition to being a massive time suck, this process is also error prone. My little startup idea would give sales reps something like 25% of their time back and it would help sales leaders have better forecasting data. I was actually serious about the idea and brought a small group of people together to spend some time on it, but I never had the guts to pursue it.
That idea evolved. I realized that you could see the style of every sales rep and the disposition of every deal and start to understand what needs to happen in calls in order to make them successful, and what typically happens in calls that make them unsuccessful. In other words, imagine looking at the way people spoke on meetings throughout the sales process in successful deals and using that to give real-time coaching to reps as they progress through the same stages. I do not necessarily mean real-time call coaching (live on a call); I mean a system that can coach a rep on what to do in their meetings based on known outcomes. This could include something as simple as “When you are at X deal stage, you typically have had this specific stakeholder involved in the deal, but right now that stakeholder is not in the deal. The advice is to get this stakeholder involved before progressing to the next step.” You get the idea.
Everything I just described above is exactly what a company called Momentum does. Momentum was just acquired for a boatload of money by Salesforce. Momentum is also one of about a dozen companies that does this now. As for that CRM updating thing? There might be 100 companies who do that now. It is also something you can just build on your own in a short amount of time. As for Gong, they just had a major product update, and the simple procedure of just updating the CRM was one of the highlights of this major product update.
I’m not sitting here saying I’m a genius. I’m actually an idiot for not having the conviction to start that company. All I am saying is that there are clearly a lot of folks out there who are moving much, much faster in this space and getting to what actually matters. Recording the call is table stakes now. It is what you do with the data from the call that is interesting. Gong is only just now getting around to the ability to take data from a call transcript and map it to fields in your CRM. Meanwhile there are companies who started in a garage somewhere and got bought by Salesforce in the span of the last four years or so. That is telling.
Let me talk a little bit about what I am doing and continue to give some more concrete examples of why I think Gong is in big, big trouble. My team recently invested in a tool called Cassidy.ai. It is an AI Orchestration platform, very similar to Zapier, another tool I am a big fan of. I basically was of the mind that there were three pathways companies could take with AI:
- Continue to invest in point-solutions (like Gong).
- Build it all in-house on their own with a tool like n8n (requires some technical know-how).
- Implement an AI Orchestration tool like Cassidy.ai or Station.ai.
I went with Option 3. I felt it was the best way to get us away from investing in point solutions, providing us autonomy on building in-house, and eliminating the need for spending too much time on acquisition of the technical know-how.
One of the first things I did in Cassidy was build an Automated Jeff Kirchick Coaching Bot. I fed the system a transcript of my book, all the blogs I have ever written, and a compilation of materials I had prepared over the last year for my company in our team standups and trainings. Over time, I actually would manually coach calls of my sales reps and I would tell Cassidy what coaching I was giving them. So basically this little bot that I created was very smart about how I like to coach calls, and it was only continuing to get smarter and smarter as I fed it more coaching data. After every single phone call my sales reps have, they get an email with coaching from me. Except the coaching is not from “me.” It’s from some virtual, AI version of me I’ve never met before but who happens to sound a lot like me. And they love it.
What I just mentioned above technically should not be hard for Gong to do as well. Within Gong, you can provide call coaching (actually in a very nice UI, I might add). Gong should have records of all the coaching that the manager is giving the disciple. Gong could also (with very little effort) integrate LLMs into its product to allow a sales manager to upload their sales philosophies and collateral as additional “backing” for what the coaching model should know. Gong already sends a summary and synopsis of next steps after the call. Why can’t they do what I just described above, something I built in five minutes with Cassidy?
Our team is having our sales onsite this week. In preparation for the onsite, we wanted to have very broad-sweeping information about which of our competitors had been mentioned the most over the past year, but also what kinds of feedback we had been hearing, and any general takeaways or recommendations we should give to sales or product based on this feedback. Within Gong, you can ask about a specific call (i.e., “How many questions did the sales rep ask on this call?”) but unfortunately what you cannot do is ask it sophisticated questions about calls at-large. So while you can see what competitor has been mentioned the most, you cannot ask a question like “What do people like and dislike about X competitor based on all of these calls?”. But if you feed all your call data into an LLM (even just ChatGPT or Claude), obviously this is easy to do. So the people responsible for compiling this data did something in five minutes that somehow Gong simply cannot do. In those five minutes, we received board-level type of analysis about our entire competitive standing.
What I am working on right now is something fairly ambitious but it happens to be something I am pretty excited about. We are using a deal qualification system I made up called SPIDD. It’s kind of like if MEDDICC and SPICED had a baby, and that baby had a super lame name. Anyway, I’m in the midst of building a workflow where we automate the process of moving our call data into our CRM. However, I am also building a SPIDD scorecard where there is a 0-5 scoring rubric for each of the five letters, rendering a total score somewhere between 0 and 25. The more of the criteria you have crossed off, the healthier your deal is. This will allow us to run real-time reporting on deals based on health and to start to learn what makes our deals unhealthy. And all of the data will be completely objective based on the source of truth (the actually conversations) and not subject to the bias of the sales reps who could otherwise fill out a scorecard based on how they feel.
Anyone who has used Gong knows that they have their own SPICED and MEDDICC trackers. They don’t work very well in my opinion, but that’s neither here nor there. Those trackers do not allow you to build your own scoring rubrics that are automatically cross-referenced with the call data. Anyone who wants to do something like what I am building in Cassidy simply cannot do it within the confines of Gong. What they can do is take the call transcript from Gong and they can run that data through another tool to get themselves to the outcome, but in doing so, the only value coming from Gong is the transcript – the same transcript that can be retrieved from any tool. In my case, Cassidy has its own recorder, so I can keep everything all under one hood.
I have other use cases that go beyond this. I’ve been making a push to ask more open-ended questions. Again, I can go in Gong and ask how a team member does this on one specific call, but I cannot ask it about how the rep is trending with that coaching over a broad swath of time. I could go on and on. The point is, given how much you can build outside of tools like Gong, the value of the tool ultimately ends up being the transcript, which is commoditized.
Zoominfo knows this: they basically give their version of Gong (Chorus.ai) away for free now as part of their plans. They know that this is no longer the holy grail. What they did instead was pretty smart. They built a product called GTM Studio, which is kind of like their spinoff of Clay. Since GTM Studio is integrated with Chorus, you can use the two in tandem to build very cool workflows leveraging your conversational intelligence data, including some of the use cases I mentioned above.
None of this is meant to be an attack on anyone. I admit I could be wrong about a lot of this. I tried to buy myself some additional time to play with some of Gong’s new features but they did not seem too motivated to keep me as a customer. So I’m sharing my findings with limited data.
I find this space to be fascinating and hope it inspires some people I know to share with me some of the workflows they are building around conversational intelligence. It is truly the dawn of a new era.
Jeff Kirchick 